Texture based weed classification has played an important role in agricultural applications. In the recent years weed classification based on wavelet transform is an effective method. But the feature extraction is main issue for proper classification of weed species. In this paper, the issue of statistical and texture
classification based on wavelet transform has been analysed. The efficient texture feature extraction
methods are developed for weed discrimination. Three group feature vector can be constructed by the mean
and standard deviation of the wavelet statistical features (WSF), Texture feature as Contrast, Cluster
Shade, Cluster Prominence and Local Homogeneity (WCSPH) and Energy, Correlation, Cluster Shade,
Cluster Prominence and Entropy features (WECSPE) which are derived from the sub-bands of the wavelet
decomposition and are used for classification. Experimental results show that Rbio33 Wavelet with
WECSPE texture feature obtaining high degree of success rate in classification.
Pest Control in Agricultural Plantations Using Image ProcessingIOSR Journals
Abstract: Monocropped plantations are unique to India and a handful of countries throughout the globe. Essentially, the FOREST approach of growing coffee along with in India has enabled the plantation to fight many outbreaks of pests and diseases. Mono cropped Plantations are under constant threat of pest and disease incidence because it favours the build up of pest population. To cope with these problems, an automatic pest detection algorithm using image processing techniques in MATLAB has been proposed in this paper. Image acquisition devices are used to acquire images of plantations at regular intervals. These images are then subjected to pre-processing, transformation and clustering.
Agriculture is the backbone of human sustenance on this world. Now a days with growing population we need the productivity of the agriculture to be increased a lot to meet the demands. In olden days they used natural methods to increase the productivity, such as using the cow dung as a fertilizer in the fields. That resulted increase in the productivity enough to meet the requirements of the population. But later people started thinking of earning more profits by getting more outcome. So, there came a revolution called “Green Revolution”. In this paper we implemented image processing using MATLAB to detect the weed areas in an image we took from the fields.
Weed and crop segmentation and classification using area thresholdingeSAT Journals
Abstract In the agricultural industry, the weed and crop identification and classification are major technical and economical importance. Two classification algorithms are focused in this paper. And the better classification algorithm has been selected to classify weed and crop from the images. There are three main parts of proposed system are segmentation, classification and error calculation. The developed algorithm based on area thresholding has been tested on weeds and various locations. Forty one sample images have been tested and the result of some weed coverage rate is illustrated. Moreover, the misclassification rate is also computed. An algorithm has been done to automate the tasks of segmentation and classification. The overall process is implemented in MATLAB. Keywords - Objects segmentation, Image processing, Plant classification, Area Thresholding
Image processing to identify foliar infectionTHANMAY JS
Most important and basic recognition requirement for measuring percentage of Leaf Infection was expressed by “Stoyer’s experiment to determine Foliar Infections”. Other than this tedious and destructive method, a new and simple method for plant leaf infection measurement using image processing technique is provided in this research work. Leaves with different shape and size of same plant were selected for study. Conventional Experiments were performed to test the performance of the estimating system by comparing the estimated leaf infected area. The ranges of relative errors of estimation were analyzed for better future of research and development in this area. The results illuminated more accurate of the estimating leaf infected area. So it can be concluded that the method developed in this study obtained the sufficient accuracy for the possible non-destructive leaf Infection area measurement.
Pest Control in Agricultural Plantations Using Image ProcessingIOSR Journals
Abstract: Monocropped plantations are unique to India and a handful of countries throughout the globe. Essentially, the FOREST approach of growing coffee along with in India has enabled the plantation to fight many outbreaks of pests and diseases. Mono cropped Plantations are under constant threat of pest and disease incidence because it favours the build up of pest population. To cope with these problems, an automatic pest detection algorithm using image processing techniques in MATLAB has been proposed in this paper. Image acquisition devices are used to acquire images of plantations at regular intervals. These images are then subjected to pre-processing, transformation and clustering.
Agriculture is the backbone of human sustenance on this world. Now a days with growing population we need the productivity of the agriculture to be increased a lot to meet the demands. In olden days they used natural methods to increase the productivity, such as using the cow dung as a fertilizer in the fields. That resulted increase in the productivity enough to meet the requirements of the population. But later people started thinking of earning more profits by getting more outcome. So, there came a revolution called “Green Revolution”. In this paper we implemented image processing using MATLAB to detect the weed areas in an image we took from the fields.
Weed and crop segmentation and classification using area thresholdingeSAT Journals
Abstract In the agricultural industry, the weed and crop identification and classification are major technical and economical importance. Two classification algorithms are focused in this paper. And the better classification algorithm has been selected to classify weed and crop from the images. There are three main parts of proposed system are segmentation, classification and error calculation. The developed algorithm based on area thresholding has been tested on weeds and various locations. Forty one sample images have been tested and the result of some weed coverage rate is illustrated. Moreover, the misclassification rate is also computed. An algorithm has been done to automate the tasks of segmentation and classification. The overall process is implemented in MATLAB. Keywords - Objects segmentation, Image processing, Plant classification, Area Thresholding
Image processing to identify foliar infectionTHANMAY JS
Most important and basic recognition requirement for measuring percentage of Leaf Infection was expressed by “Stoyer’s experiment to determine Foliar Infections”. Other than this tedious and destructive method, a new and simple method for plant leaf infection measurement using image processing technique is provided in this research work. Leaves with different shape and size of same plant were selected for study. Conventional Experiments were performed to test the performance of the estimating system by comparing the estimated leaf infected area. The ranges of relative errors of estimation were analyzed for better future of research and development in this area. The results illuminated more accurate of the estimating leaf infected area. So it can be concluded that the method developed in this study obtained the sufficient accuracy for the possible non-destructive leaf Infection area measurement.
An Effective Tea Leaf Recognition Algorithm for Plant Classification Using Ra...IJMER
A leaf is an organ of a vascular plant, as identified in botanical terms, and in particular in plant morphology. Naturally a leaf is a thin, flattened organ bear above ground and it is mainly used for photosynthesis. Recognition of plants has become an active area of research as most of the plant species are at the risk of extinction. Most of the leaves cannot be recognized easily since some are not flat (e.g. succulent leaves and conifers), some does not grow above ground (e.g. bulb scales), and some does not undergo photosynthetic function (e.g. cataphylls, spines, and cotyledons).In this paper, we mainly focused on tea leaves to identify the leaf type for improving tea leaf classification. Tea leaf images are loaded from digital cameras or scanners in the system. This proposed approach consists of three phases such as preprocessing, feature extraction and classification to process the loaded image. The tea leaf images can be identified accurately in the preprocessing phase by fuzzy denoising using Dual Tree Discrete Wavelet Transform (DT-DWT) in order to remove the noisy features and boundary enhancement to obtain the shape of leaf accurately. In the feature extraction phase, Digital Morphological Features (DMFs) are derived to improve the classification accuracy. Radial Basis Function (RBF) is used for efficient classification. The RBF is trained by 60 tea leaves to classify them into 6 types. Experimental results proved that the proposed method classifies the tea leaves with more accuracy in less time. Thus, the proposed method achieves more accuracy in retrieving the leaf type.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
Feature extraction of Jabon (Anthocephalus sp) leaf disease using discrete wa...TELKOMNIKA JOURNAL
Jabon (Anthocephalus cadamba (Roxb.) Miq) is one type of forest plants that have very rapid growth until the process of the harvest. One inhibitor is a disease that attacks the leaves in the form of spots and blight that can cause death during the growth process of this tree. The purpose of this process is to detect the object of diseases that attack the leaves of jabon at the time in the nursery. Images of affected jabon leaf disease segmented by reducing the RGB color cylinders to separate the disease object from the background. Reduced channel G-R provides information in the form of disease areas contained in the image of Jabon leaf. Furthermore, the characteristics of leaf disease can be detected well using DWT in the 3-level decomposition process with SVM classification results that can separate both classes of spots and blight by 84.672%.
International Journal of Computational Engineering Research(IJCER) ijceronline
nternational Journal of Computational Engineering Research (IJCER) is dedicated to protecting personal information and will make every reasonable effort to handle collected information appropriately. All information collected, as well as related requests, will be handled as carefully and efficiently as possible in accordance with IJCER standards for integrity and objectivity.
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...Tarun Kumar
In this computing era, image processing has
spread its wings in human life upto the extent that image
has become an integral part of their life. There are various
applications of image processing in the field of commerce,
engineering, graphic design, journalism, architecture and
historical research. In this research work, Image
processing is considered for the analysis of plant leaf
diseases. Plant leaf diseases can be detected based on the
disease symptoms. Here, dataset of disease affected leaves
is considered for experimentation. This dataset contains
the plant leaves suffered from the
AlternariaAlternata,Cercospora Leaf Spot, Anthracnose
andBacterial Blight along with some healthy leaf images.
For this analysis, an autonomous approach of modified
SVM-CS is introduces. Here, concept of cuckoo search is
considered to optimize the classification parameters. These
parameters further help to find more accurate solutions.
This autonomous approach also extracts the healthy
portion and disease affected leaf portion along with the
accuracy of results.
PROFILE DOSE AND PDD ANALYSIS IN SMALL PHOTON FIELD WITH PTW PINPOINT CHAMBER...AM Publications
: Profile dose and PDD of a small field 6MV have been measured using a PTW PinPoint Chamber 0.015 cc detector. The aim of this study is to analyze the profile dose and PDD of small field using PTW PinPoint Chamber detector 0.015 cc. From the results of this study is expected optimal dose measurement accuracy can be achieved in clinical treatment. The analysis includes calculating the symmetry, penumbra value and PDD. Linear accelerator Electa Precise Treatment SystemTM and PTW PinPoint Chamber 0.015cc are the material used in this research. Profile Dose and PDD were measured with an SSD technique in 100 cm with a various of field sizes and depths. The results showed that the symmetry values for depth variations would decrease in ranges (0.55-6.27)% when the depth increase, and for field size variations will decrease in range (0.39-6.27)%. when the filed size increase. The result of the penumbra analysis shows that its value is more than 4 mm for all variations of depth and field size. PDD results showed that maximum dose for all fields size is achieved in maksimum depth (1.4 – 1.7) cm.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Classification of Macronutrient Deficiencies in Maize Plant Using Machine Lea...IJECEIAES
Detection of nutritional deficiencies in plants is vital for improving crop productivity. Timely identification of nutrient deficiency through visual symptoms in the plants can help farmers take quick corrective action by appropriate nutrient management strategies. The application of computer vision and machine learning techniques offers new prospects in nondestructive field-based analysis for nutrient deficiency. Color and shape are important parameters in feature extraction. In this work, two different techniques are used for image segmentation and feature extraction to generate two different feature sets from the same image sets. These are then used for classification using different machine learning techniques. The experimental results are analyzed and compared in terms of classification accuracy to find the best algorithm for the two feature sets.
Weed and crop segmentation and classification using area thresholdingeSAT Publishing House
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Detection of urban tree canopy from very high resolution imagery using an ob...IJECEIAES
Tree that grows within a town, city and suburban areas, collection of these trees makes the urban forest. These urban forest and urban trees have impact on urban water, pollution and heat. Nowadays we are experiencing drastic climatic changes because of cutting of trees for our growth and increasing population which leads to expansion of roads, towers, and airports. Individual tree crown detection is necessary to map the forest along with feasible planning for urban areas. In this study, using WorldView-2imagery, trees in specific area are detected with object-based image analysis (OBAI) approach. Therefore with improvement in spatial and spectral resolution of an image, extracted from WorldView-2 carried out urban features with better accuracy. The aim of this research is to illustrate how object-based method can be applied to the available data to accurately find out vegetation, which can be further sub-classified to obtain area under tree canopy. The result thus obtained gives area under tree canopy with an accuracy of 92.43 % and a Kappa coefficient of 0.80.
Hybrid features and ensembles of convolution neural networks for weed detectionIJECEIAES
Weeds compete with plants for sunlight, nutrients and water. Conventional weed management involves spraying of herbicides to the entire crop which increases the cost of cultivation, decreasing the quality of the crop, in turn affecting human health. Precise automatic spraying of the herbicides on weeds has been in research and use. This paper discusses automatic weed detection using hybrid features which is generated by extracting the deep features from convolutional neural network (CNN) along with the texture and color features. The color and texture features are extracted by color moments, gray level co-occurrence matrix (GLCM) and Gabor wavelet transform. The proposed hybrid features are classified by Bayesian optimized support vector machine (BO-SVM) classifier. The experimental results read that the proposed hybrid features yield a maximum accuracy of 95.83%, higher precision, sensitivity and F-score. A performance analysis of the proposed hybrid features with BO-SVM classifier in terms of the evaluation parameters is made using the images from crop weed field image dataset.
An Effective Tea Leaf Recognition Algorithm for Plant Classification Using Ra...IJMER
A leaf is an organ of a vascular plant, as identified in botanical terms, and in particular in plant morphology. Naturally a leaf is a thin, flattened organ bear above ground and it is mainly used for photosynthesis. Recognition of plants has become an active area of research as most of the plant species are at the risk of extinction. Most of the leaves cannot be recognized easily since some are not flat (e.g. succulent leaves and conifers), some does not grow above ground (e.g. bulb scales), and some does not undergo photosynthetic function (e.g. cataphylls, spines, and cotyledons).In this paper, we mainly focused on tea leaves to identify the leaf type for improving tea leaf classification. Tea leaf images are loaded from digital cameras or scanners in the system. This proposed approach consists of three phases such as preprocessing, feature extraction and classification to process the loaded image. The tea leaf images can be identified accurately in the preprocessing phase by fuzzy denoising using Dual Tree Discrete Wavelet Transform (DT-DWT) in order to remove the noisy features and boundary enhancement to obtain the shape of leaf accurately. In the feature extraction phase, Digital Morphological Features (DMFs) are derived to improve the classification accuracy. Radial Basis Function (RBF) is used for efficient classification. The RBF is trained by 60 tea leaves to classify them into 6 types. Experimental results proved that the proposed method classifies the tea leaves with more accuracy in less time. Thus, the proposed method achieves more accuracy in retrieving the leaf type.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
Feature extraction of Jabon (Anthocephalus sp) leaf disease using discrete wa...TELKOMNIKA JOURNAL
Jabon (Anthocephalus cadamba (Roxb.) Miq) is one type of forest plants that have very rapid growth until the process of the harvest. One inhibitor is a disease that attacks the leaves in the form of spots and blight that can cause death during the growth process of this tree. The purpose of this process is to detect the object of diseases that attack the leaves of jabon at the time in the nursery. Images of affected jabon leaf disease segmented by reducing the RGB color cylinders to separate the disease object from the background. Reduced channel G-R provides information in the form of disease areas contained in the image of Jabon leaf. Furthermore, the characteristics of leaf disease can be detected well using DWT in the 3-level decomposition process with SVM classification results that can separate both classes of spots and blight by 84.672%.
International Journal of Computational Engineering Research(IJCER) ijceronline
nternational Journal of Computational Engineering Research (IJCER) is dedicated to protecting personal information and will make every reasonable effort to handle collected information appropriately. All information collected, as well as related requests, will be handled as carefully and efficiently as possible in accordance with IJCER standards for integrity and objectivity.
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...Tarun Kumar
In this computing era, image processing has
spread its wings in human life upto the extent that image
has become an integral part of their life. There are various
applications of image processing in the field of commerce,
engineering, graphic design, journalism, architecture and
historical research. In this research work, Image
processing is considered for the analysis of plant leaf
diseases. Plant leaf diseases can be detected based on the
disease symptoms. Here, dataset of disease affected leaves
is considered for experimentation. This dataset contains
the plant leaves suffered from the
AlternariaAlternata,Cercospora Leaf Spot, Anthracnose
andBacterial Blight along with some healthy leaf images.
For this analysis, an autonomous approach of modified
SVM-CS is introduces. Here, concept of cuckoo search is
considered to optimize the classification parameters. These
parameters further help to find more accurate solutions.
This autonomous approach also extracts the healthy
portion and disease affected leaf portion along with the
accuracy of results.
PROFILE DOSE AND PDD ANALYSIS IN SMALL PHOTON FIELD WITH PTW PINPOINT CHAMBER...AM Publications
: Profile dose and PDD of a small field 6MV have been measured using a PTW PinPoint Chamber 0.015 cc detector. The aim of this study is to analyze the profile dose and PDD of small field using PTW PinPoint Chamber detector 0.015 cc. From the results of this study is expected optimal dose measurement accuracy can be achieved in clinical treatment. The analysis includes calculating the symmetry, penumbra value and PDD. Linear accelerator Electa Precise Treatment SystemTM and PTW PinPoint Chamber 0.015cc are the material used in this research. Profile Dose and PDD were measured with an SSD technique in 100 cm with a various of field sizes and depths. The results showed that the symmetry values for depth variations would decrease in ranges (0.55-6.27)% when the depth increase, and for field size variations will decrease in range (0.39-6.27)%. when the filed size increase. The result of the penumbra analysis shows that its value is more than 4 mm for all variations of depth and field size. PDD results showed that maximum dose for all fields size is achieved in maksimum depth (1.4 – 1.7) cm.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Classification of Macronutrient Deficiencies in Maize Plant Using Machine Lea...IJECEIAES
Detection of nutritional deficiencies in plants is vital for improving crop productivity. Timely identification of nutrient deficiency through visual symptoms in the plants can help farmers take quick corrective action by appropriate nutrient management strategies. The application of computer vision and machine learning techniques offers new prospects in nondestructive field-based analysis for nutrient deficiency. Color and shape are important parameters in feature extraction. In this work, two different techniques are used for image segmentation and feature extraction to generate two different feature sets from the same image sets. These are then used for classification using different machine learning techniques. The experimental results are analyzed and compared in terms of classification accuracy to find the best algorithm for the two feature sets.
Weed and crop segmentation and classification using area thresholdingeSAT Publishing House
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Detection of urban tree canopy from very high resolution imagery using an ob...IJECEIAES
Tree that grows within a town, city and suburban areas, collection of these trees makes the urban forest. These urban forest and urban trees have impact on urban water, pollution and heat. Nowadays we are experiencing drastic climatic changes because of cutting of trees for our growth and increasing population which leads to expansion of roads, towers, and airports. Individual tree crown detection is necessary to map the forest along with feasible planning for urban areas. In this study, using WorldView-2imagery, trees in specific area are detected with object-based image analysis (OBAI) approach. Therefore with improvement in spatial and spectral resolution of an image, extracted from WorldView-2 carried out urban features with better accuracy. The aim of this research is to illustrate how object-based method can be applied to the available data to accurately find out vegetation, which can be further sub-classified to obtain area under tree canopy. The result thus obtained gives area under tree canopy with an accuracy of 92.43 % and a Kappa coefficient of 0.80.
Hybrid features and ensembles of convolution neural networks for weed detectionIJECEIAES
Weeds compete with plants for sunlight, nutrients and water. Conventional weed management involves spraying of herbicides to the entire crop which increases the cost of cultivation, decreasing the quality of the crop, in turn affecting human health. Precise automatic spraying of the herbicides on weeds has been in research and use. This paper discusses automatic weed detection using hybrid features which is generated by extracting the deep features from convolutional neural network (CNN) along with the texture and color features. The color and texture features are extracted by color moments, gray level co-occurrence matrix (GLCM) and Gabor wavelet transform. The proposed hybrid features are classified by Bayesian optimized support vector machine (BO-SVM) classifier. The experimental results read that the proposed hybrid features yield a maximum accuracy of 95.83%, higher precision, sensitivity and F-score. A performance analysis of the proposed hybrid features with BO-SVM classifier in terms of the evaluation parameters is made using the images from crop weed field image dataset.
Identification and Classification of Leaf Diseases in Turmeric PlantsIJERA Editor
Plant disease identification is the most important sector in agriculture. Turmeric is one of the important
rhizomatous crops grown in India. The turmeric leaf is highly exposed to diseases like rhizome rot, leaf spot,
and leaf blotch. The identification of plant diseases requires close monitoring and hence this paper adopts
technologies to manage turmeric plant diseases caused by fungi to enable production of high quality crop yields.
Various image processing and machine learning techniques are used to identify and classify the diseases in
turmeric leaf. The dataset with 800 leaf images of different categories were pre-processed and segmented to
promote efficient feature extraction. Machine learning algorithms like support vector machine, decision tree and
naïve bayes were applied to train the model. The performance of the model was evaluated using 10 fold cross
validation and the results are reported.
Identification and Classification of Leaf Diseases in Turmeric PlantsIJERA Editor
Plant disease identification is the most important sector in agriculture. Turmeric is one of the important
rhizomatous crops grown in India. The turmeric leaf is highly exposed to diseases like rhizome rot, leaf spot,
and leaf blotch. The identification of plant diseases requires close monitoring and hence this paper adopts
technologies to manage turmeric plant diseases caused by fungi to enable production of high quality crop yields.
Various image processing and machine learning techniques are used to identify and classify the diseases in
turmeric leaf. The dataset with 800 leaf images of different categories were pre-processed and segmented to
promote efficient feature extraction. Machine learning algorithms like support vector machine, decision tree and
naïve bayes were applied to train the model. The performance of the model was evaluated using 10 fold cross
validation and the results are reported.
Digital image processing methods for estimating leaf area of cucumber plantsnooriasukmaningtyas
Increasingly emerging technologies in agriculture such as computer vision, artificial intelligence technology, not only make it possible to increase production. To minimize the negative impact on climate and the environment but also to conserve resources. A key task of these technologies is to monitor the growth of plants online with a high accuracy rate and in non-destructive manners. It is known that leaf area (LA) is one of the most important growth indexes in plant growth monitoring system. Unfortunately, to estimate the LA in natural outdoor scenes (the presence of occlusion or overlap area) with a high accuracy rate is not easy and it still remains a big challenge in eco-physiological studies. In this paper, two accurate and non-destructive approaches for estimating the LA were proposed with top-view and side-view images, respectively. The proposed approaches successfully extract the skeleton of cucumber plants in red, green, and blue (RGB) images and estimate the LA of cucumber plants with high precision. The results were validated by comparing with manual measurements. The experimental results of our proposed algorithms achieve 97.64% accuracy in leaf segmentation, and the relative error in LA estimation varies from 3.76% to 13.00%, which could meet the requirements of plant growth monitoring systems.
Edge detection of herbal plants is a set of mathematical methods which aim at identifying points in a digital image at which the image brightness changes sharply and has discontinuities. They are defined as the set of curved line segments termed edges. Effective edge detection for microscopic image of herbal plant is proposed through this paper which compares the edge detected images and then performs further segmentation. Comparison between Sobel operator, Prewitt, Canny and Robert cross operators is performed. Our method after efficient edge detection performs Gabor filter and K-means clustering to procure a better image. It is then subjected to further segmentation. Experimental methods in our proposed algorithm show that our method achieves a better edge detection as compared to other edge detector operators. Our proposed algorithm provides the maximum PSNR value of 43.684 amongst the other commercial edge detection operators.
precise weed management is very useful under large land holdings which reduces cost of cultivation to a greater extent. remote sensing plays a major role in site specific weed management
Abstract In this paper illustrates the improvement of a low cost machine vision system using webcams and image processing algorithms for defect detection and sorting of tomatoes The sorting decision was based on three features extracted by the different image processing algorithms. This methodology based on the color features, which used for detecting the BER from good tomatoes. Two methods were developed for decision based sorting. The color image threshold method with shape factor was found efficient for differentiating good and defective tomatoes. The overall accuracy of defect detection attained was 94 and 96.5% respectively. Comparison of the results is also presented in this paper. Keywords: Dither Image, Stem Image, Histogram, Tomato.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)MdTanvirMahtab2
This presentation is about the working procedure of Shahjalal Fertilizer Company Limited (SFCL). A Govt. owned Company of Bangladesh Chemical Industries Corporation under Ministry of Industries.
Overview of the fundamental roles in Hydropower generation and the components involved in wider Electrical Engineering.
This paper presents the design and construction of hydroelectric dams from the hydrologist’s survey of the valley before construction, all aspects and involved disciplines, fluid dynamics, structural engineering, generation and mains frequency regulation to the very transmission of power through the network in the United Kingdom.
Author: Robbie Edward Sayers
Collaborators and co editors: Charlie Sims and Connor Healey.
(C) 2024 Robbie E. Sayers
About
Indigenized remote control interface card suitable for MAFI system CCR equipment. Compatible for IDM8000 CCR. Backplane mounted serial and TCP/Ethernet communication module for CCR remote access. IDM 8000 CCR remote control on serial and TCP protocol.
• Remote control: Parallel or serial interface.
• Compatible with MAFI CCR system.
• Compatible with IDM8000 CCR.
• Compatible with Backplane mount serial communication.
• Compatible with commercial and Defence aviation CCR system.
• Remote control system for accessing CCR and allied system over serial or TCP.
• Indigenized local Support/presence in India.
• Easy in configuration using DIP switches.
Technical Specifications
Indigenized remote control interface card suitable for MAFI system CCR equipment. Compatible for IDM8000 CCR. Backplane mounted serial and TCP/Ethernet communication module for CCR remote access. IDM 8000 CCR remote control on serial and TCP protocol.
Key Features
Indigenized remote control interface card suitable for MAFI system CCR equipment. Compatible for IDM8000 CCR. Backplane mounted serial and TCP/Ethernet communication module for CCR remote access. IDM 8000 CCR remote control on serial and TCP protocol.
• Remote control: Parallel or serial interface
• Compatible with MAFI CCR system
• Copatiable with IDM8000 CCR
• Compatible with Backplane mount serial communication.
• Compatible with commercial and Defence aviation CCR system.
• Remote control system for accessing CCR and allied system over serial or TCP.
• Indigenized local Support/presence in India.
Application
• Remote control: Parallel or serial interface.
• Compatible with MAFI CCR system.
• Compatible with IDM8000 CCR.
• Compatible with Backplane mount serial communication.
• Compatible with commercial and Defence aviation CCR system.
• Remote control system for accessing CCR and allied system over serial or TCP.
• Indigenized local Support/presence in India.
• Easy in configuration using DIP switches.
Hierarchical Digital Twin of a Naval Power SystemKerry Sado
A hierarchical digital twin of a Naval DC power system has been developed and experimentally verified. Similar to other state-of-the-art digital twins, this technology creates a digital replica of the physical system executed in real-time or faster, which can modify hardware controls. However, its advantage stems from distributing computational efforts by utilizing a hierarchical structure composed of lower-level digital twin blocks and a higher-level system digital twin. Each digital twin block is associated with a physical subsystem of the hardware and communicates with a singular system digital twin, which creates a system-level response. By extracting information from each level of the hierarchy, power system controls of the hardware were reconfigured autonomously. This hierarchical digital twin development offers several advantages over other digital twins, particularly in the field of naval power systems. The hierarchical structure allows for greater computational efficiency and scalability while the ability to autonomously reconfigure hardware controls offers increased flexibility and responsiveness. The hierarchical decomposition and models utilized were well aligned with the physical twin, as indicated by the maximum deviations between the developed digital twin hierarchy and the hardware.
Sachpazis:Terzaghi Bearing Capacity Estimation in simple terms with Calculati...Dr.Costas Sachpazis
Terzaghi's soil bearing capacity theory, developed by Karl Terzaghi, is a fundamental principle in geotechnical engineering used to determine the bearing capacity of shallow foundations. This theory provides a method to calculate the ultimate bearing capacity of soil, which is the maximum load per unit area that the soil can support without undergoing shear failure. The Calculation HTML Code included.
CW RADAR, FMCW RADAR, FMCW ALTIMETER, AND THEIR PARAMETERSveerababupersonal22
It consists of cw radar and fmcw radar ,range measurement,if amplifier and fmcw altimeterThe CW radar operates using continuous wave transmission, while the FMCW radar employs frequency-modulated continuous wave technology. Range measurement is a crucial aspect of radar systems, providing information about the distance to a target. The IF amplifier plays a key role in signal processing, amplifying intermediate frequency signals for further analysis. The FMCW altimeter utilizes frequency-modulated continuous wave technology to accurately measure altitude above a reference point.
6th International Conference on Machine Learning & Applications (CMLA 2024)ClaraZara1
6th International Conference on Machine Learning & Applications (CMLA 2024) will provide an excellent international forum for sharing knowledge and results in theory, methodology and applications of on Machine Learning & Applications.
Student information management system project report ii.pdfKamal Acharya
Our project explains about the student management. This project mainly explains the various actions related to student details. This project shows some ease in adding, editing and deleting the student details. It also provides a less time consuming process for viewing, adding, editing and deleting the marks of the students.
Student information management system project report ii.pdf
A STUDY ON WEED DISCRIMINATION THROUGH WAVELET TRANSFORM, TEXTURE FEATURE EXTRACTION AND CLASSIFICATION
1. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
DOI:10.5121/ijcsit.2015.7304 41
A STUDY ON WEED DISCRIMINATION THROUGH
WAVELET TRANSFORM, TEXTURE FEATURE
EXTRACTION AND CLASSIFICATION
D.Ashok Kumar1
and P.Prema2
1
Government Arts College, Trichy, TamilNadu, India.
2
Agricultural College and Research Institute, Madurai-625104. TamilNadu, India
ABSTRACT
Texture based weed classification has played an important role in agricultural applications. In the recent
years weed classification based on wavelet transform is an effective method. But the feature extraction is
main issue for proper classification of weed species. In this paper, the issue of statistical and texture
classification based on wavelet transform has been analysed. The efficient texture feature extraction
methods are developed for weed discrimination. Three group feature vector can be constructed by the mean
and standard deviation of the wavelet statistical features (WSF), Texture feature as Contrast, Cluster
Shade, Cluster Prominence and Local Homogeneity (WCSPH) and Energy, Correlation, Cluster Shade,
Cluster Prominence and Entropy features (WECSPE) which are derived from the sub-bands of the wavelet
decomposition and are used for classification. Experimental results show that Rbio33 Wavelet with
WECSPE texture feature obtaining high degree of success rate in classification.
KEYWORDS
Pre-processing, Wavelets, Texture Features, Neural Network.
1. INTRODUCTION
Unwanted plants growing in a place at the wrong time and wrong place in the field can be defined
as Weed. The weeds compete with crop for nutrient, water, light and space”. Controlling of weed
in the agricultural field is more expensive and time-consuming activities. Weed control is a
necessary management practice for all cropping systems. Weeds are controlled by manual
removal by labourer and applying herbicide uniformly within the field.
Actually the weed infestation rates vary from low to high densities. Based on the weed infection
some dosage of herbicide can be applied in the particular location an position.
Mostly all existing weed detection methods the image has been processed in two steps. In the first
step, segmentation of vegetation against the background (soil and/or harvest residues) and in
second step, detection of the vegetation pixels that represent weeds.
The vegetation can be extracted by combination of color planes such as HIS, RGB and YCbCr
method. Green vegetation segmentation can be achieved using vegetative indices techniques such
as excess green index (ExG), [1,2] the excess red index (ExR), [3], the excess green minus excess
red index (ExGR), Offset Excess Green(OEG)[4], vegetative index (VEG) [5],HIS color model
combined with classification methods such as Bayes networks and clustering [6]. The author [7]
has explained several vegetation segmentation methods for greenness extraction.
2. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
42
Our approach Global Histogram Equalization with excess green index (ExG) pre-processing
techniques was designed to cope with the variability of natural daylight illumination. The
vegetation pixels have been identified, color, position, shape, texture features of weed has
extracted for classification. These features are depends on the way the photographic images are
taken and the weed species involved. Development of feasible site-specific weed control
technology makes the above property.
Wavelet transform is widely used in machine vision in an image processing technique for object
detection and classification. For real-time specific weed leaf discrimination that is employing haar
wavelet decomposition[8] and 200 highest coefficients was extracted, that is used to classify the
broad and narrow weed leaves using a classifier k-Nearest Neighbor (k-NN) and classification
accuracy is 94% for broad weed and 92% for narrow weed and 97% for little or no weed. The
overall accuracy for weed discrimination is 94.3%.
The author [9] reported that Symlet wavelet transforms along with stepwise linear discriminant
analysis (SWLDA) and support vector machines (SVMs) for crop and weed classification for real
time selective herbicides systems . Classification accuracy obtained using the symlet wavelet
family was 98.1%. To recognize radish and weeds from the field an artificial neural network
(ANN) was used and the successful recognition rate in the farms was 93.3% for radish and 93.8%
for weeds [10].
For automatic detection is based on machine vision or spectroscopy [11]. Geometrical, textural or
other statistical features are used to detect the weeds [3] particularly in machine vision methods
shape features can be used to discriminate between corn and weeds. Other features such as mean
of color information [12, 13] and non-morphological analysis based on the texture of the image
[13] are used for weed discrimination. For texture analysis [4] wavelet transform provide better
multi resolution analytical tool. Higher accuracy rate can be achieved in weed classification.
Texture image feature provides vital characteristics for surface and object identification from
image[14]. Major component of image processing is texture analysis and texture features are used
in many applications such as remote sensing, quality inspection, medical imaging, etc. So the in
this paper texture features are used for weed discrimination. Compared to other features color and
shape, wavelet based textures feature provide high degree of accuracy in weed discrimination for
site specific real-time for site-specific spraying of infested areas [15].
Our objective of the weed discrimination study was to develop weed discrimination image
processing algorithms that are based on texture feature of the image to detect broad and narrow
leafed weed with high accuracy without any prior assumptions about environmental factors such
as lighting conditions. Our proposed pre-processing technique global histogram equalization
method is used to normalize the histograms of the images to illuminate the lighting effect. The
vegetation extraction method Excess Green method (2G-R-B) is used to enhance the green pixels
of the image. After pre-processing, segmentation techniques have applied to retrieve the weed
portion of the image. The retrieved weed RGB image was given as input to the wavelet transform.
We evaluated and compared four different wavelet families such as symlet , Daubechies ,
biorthogonal and reverse biorthogonal wavelet family for weed classification. Using Gray level
co-occurrence matrix (GLCM) meaningful statistical and texture features are extracted from the
sub-band images and Neural Network techniques are used for classification. Our proposed
segmentation method and the reverse biorthogonal 3.3 wavelet families with texture features
such as Energy , Correlation, Cluster Shade, Cluster Prominence and Entropy (WECSPE) gave
the lesser execution time 19.20ms and higher accuracy of 98.9% for weed discrimination from
the agricultural field.
3. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
43
2. MATERIALS AND METHODS
2.1. Image Acquisition
The images used for this study belongs to broad and narrow leaf weed. All the weed images were
captured using Nikon digital camera under various lighting conditions (sunny,cloudy)in the field
of Agricultural college and Research institute, Madurai. The RGB color images were captured
under perspective projection and stored in the JPG format. Figure 1 shows the sample images of a
Broadleaves and Narrow weed.
Figure 1: Input images for weed
2.2. Pre-processing
The initial goal of the weed detection task was to divide the different pixels of the images into
two classes as background and plant. Most of the input images have additional parameters such as
lighting effects, background information and unnecessary details that may cause
misclassification. Hence it is important to remove unnecessary parameters for fast processing and
to improve the quality of the image. The proposed pre-processing method is represented in the
Figure 2.
Figure 2 : Flow chart for Weed and Background Separation
.
4. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
44
In the pre-processing stage, Global Histogram Equalization (GHE) was used to improve the
quality of the image by lengthening the intensity of the dynamic range using the histogram of the
whole image. Excess green index is used to extract the greenness of the image. The next goal is to
isolate the green parts from the background. For this purpose several thresholding methods were
applied. For automatic weed discrimination we computed the average gray level was set as
threshold [16]. The binary image can be obtained based on the threshold. The pixel values greater
than the threshold are labelled as white, i.e. they contain a high degree of greenness. The outputs
of the pre-processing stage are represented in Figure 3.
Figure 3 : Result of Pre-processing techniques
Sample Input Image Global Histogram Equalization
Image
Excess Green output
Binary image of
weed(white)
Binary image of
Background(White)
RGB Image for weed
2.3. Wavelets transform
Wavelet is a widely used multi-resolution tool in image processing [4,9,15]. Wavelet transform is
the best method to represent both time and frequency content of the image. There are number of
ways to separate the low(smooth variations in color) and the high frequency components(the
edges which give details). The input image can be decomposed using discrete wavelet
transform.To extend the wavelet transform to two dimensions, the input image is down sample in
the horizontal and vertical directions. The output produces four sub bands at each decomposition
denoting the horizontal frequency, vertical frequency represented as high-high (HH), high-low
(HL), low-high (LH) and low-low (LL) image sub bands. Continuously applying the same
scheme to the low-low sub band a multiresolution decomposition can be achieved. Figure 3
shows the normal layout of such decomposition. At each scale the sub bands are sensitive to
frequencies and the LH, HL and HH sub bands are sensitive to vertical, horizontal and diagonal
frequencies respectively. Figure 4 shows a DWT decomposition of the narrow weed image.
Figure 4: Three level decomposition of Discrete Wavelet Transform
5. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
45
The Input images were decomposed using different types of wavelet families. The 2D RGB
image was represented as I, the decomposition of the wavelet transform first level can be
expressed as I = A1+D1, where A1 and D1 represent the approximation and details coefficients
respectively. In the image is decomposed up to multilevel the Equation can be written as
I=Aj+Dj+Dj-1+Dj-2+…..D2+D1 ---------- (1)
Where j represents number of wavelet decomposition. For feature extraction, only approximation
coefficients are used, the detail coefficients consist of noise. Each image was decomposed up to
three levels j=3. Figure 5 shows the result of applying wavelet transform (Reverse
Biorthogonal3.3) up to three levels on a sample image. In this study three different level of
decomposition based on the four different wavelet function namely
Daubechies(db4,db10),Symlets(sym4,sym8),Biorthogonal(bior3.3,bior6.8),Reverse Biorthogonal
(rbio3.3,rbio6.8) were used.
Figure 5 : Reverse Biorthogonal3.3 Wavelet Transform (a). Three level wavelet decomposition (b)
Reconstructed image
2.4. Feature Extraction
Texture is defined as an entity having group of mutually related pixels within an image. This
group of pixels is represented as texture elements [4,18].The aim of the weed discrimination
analysis was to classify the objects derived by image segmentation into defined number of classes
according to their specific features. The wavelet energy texture features were efficiently used for
texture classification and segmentation. First order statistical features mean, standard deviation
and second order statistical features like cluster shade and cluster prominence are calculated from
GLCM and used for classification of broad and narrow leaf weed [3].Five texture features such as
Energy, Entropy, Inertia and local homogeneity were extracted from coefficient of Discrete
wavelet transform are used texture classification of crop and weed. In real time automatic spray
control system [17], hundred highest average feature coefficients were selected from each level of
decomposition and Euclidian distance method was used to recognize and differentiate broad and
narrow weed. Sum of all the approximation coefficients at each level of decomposition are
considered as feature vector for classification of Broad and Narrow leaf weed [9].
The proposed work introduces a new texture based feature extraction method for weed
discrimination called Multi Resolution Combined Statistical and Texture Feature (MRCST).
MRCST is a combination of first order, second order statistical and texture properties. Using
GLCM [17] number of texture features are extracted from the input image and stored in a
database for weed discrimination. Texture classification involves learning and classification. In
the learning phase, the original image is decomposed using Discrete Wavelet Transform (DWT)
with three level of decomposition of different wavelet family. Features such as mean µ and
standard deviation σ are calculated from each of these wavelet sub-bands and are used to
6. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
46
construct a global feature vectors as Wavelet Mean Standard deviation G=[Gµ,Gσ],where Gµ=[µ1
, µ2,…. µS ], Gσ =[σ1, σ2, σS],From the GLCM texture feature as Contrast ,Cluster Shade,
Cluster Prominence and Local Homogeneity (WCSPH) and Energy , Correlation ,Cluster Shade,
Cluster Prominence and Entropy features (WECSPE) and stored in the Features set for the
purpose of weed classification.
We use the following notation: G is the number of gray levels used. P(i,j) is the Grey level co-
occurrence matrix. µ is the mean value of P. µx, µy are the means of Px and Py. Px(i) is the ith
entry in the marginal-probability matrix obtained by summing the rows of P(i, j):
Four Texture features were extracted from coefficient of DWT. The texture features of Contrast,
Cluster shade, Cluster Prominence and Local Homogeneity are calculated using the following
formula
2.5. Feature Classification
Various classification techniques are used for weed discrimination and disease identification.
Most of the classification is based on population variance [5], weed coverage rate (2D-WCR) [6],
histogram analysis [7]. All the above methods, the value of image pixels are summed up and set
the threshold value for weed image classification. This may cause misclassification of real time
weed images, because there are no specific criteria for selecting a threshold value. To overcome
the above problem automated classification methods have also been used. The classification
methods are support vector machines (SVMs) [17], and neural networks (NNs) and active shape
model [19] used for weed discrimination and seed identification. In this paper Back propagation
Neural Network was used to classify the input weed as broad and narrow leaf weed. Figure 6
represent the texture feature extraction and classification process.
7. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
47
Fig 6 : Process flow of texture feature extraction and classification
2.6. Neural Network
For weed recognition and classification a back propagation neural network was used. The node or
elements are composed, which is operating in parallel. These elements are stimulated by
biological nervous systems. The network function is determined largely by the connections
between elements. A neural network can be trained to perform a particular function by adjusting
the values of the connections (weights) between elements. The network is adjusted, based on a
comparison of the output and the target, until the network output matches the target image.
Feature of 50 images of broad and narrow leaf weeds were used to build ANN model. Set of
Feature values are assigned to the input layer of neural network. The number hidden layer will be
selected based on the following formula.
No. of hidden layer = (number of attribute + number of classes)/2.
The expected output in the training file was {1} for Broad weed and {0} for narrow weed.
2.7. Result and Discussion
Multilevel wavelet decomposition is a technique used in weed discrimination. In this crop/weed
discrimination technique is based on wavelet transforms comprised of Daubechies, Mayer and
Biorthogonal and compared it to a Gabor filtering. From this wavelet algorithm comparison
Daubechies and Mayer gave the best results (84.6% & 84.1%) and the Biorthogonal is the worst
one. Biorthogonal wavelet gave the result 76.4% and Gabor filter contributed the result as 73.7%
[15] . For weed classification the author[17] used wavelet based algorithm as Symlets,
Daubechies, Biorthogonal and Reverse Biorthogonal. Among those the Biorthogonal wavelet
with energy features gave the best result (97.64%).
In this paper the proposed algorithm classified real-time site specific weed discrimination based
on wavelet families Daubechies(db4,db10), Symlets(sym4,sym8), and Reverse Biorthogonal
(rbio3.3,rbio6.8) with three group of texture features as Mean and standard deviation , texture
feature as Contrast ,Cluster Shade, Cluster Prominence and Local Homogeneity (WCSPH) and
Energy , Correlation ,Cluster Shade, Cluster Prominence and Entropy features (WECSPE). To
build the database 50 images are used for each class (broad and narrow weed. Three levels of
decomposition are applied to all the type of wavelet families. And three group of features
8. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
48
sets(MWS, WCSPH, WECSPE) are extracted from approximation and details coefficients of
from different wavelet families and all the decomposition level and are stored. All the coefficients
(detail and approximation) contains some useful information, it will necessary requires for proper
classification and higher accuracy. Feature sets are assigned as input to the Neural Network.
Number of Hidden layer is chosen based on the number of inputs. Neural network techniques
were applied separately to all the three features set of different wavelet families. Confusion
matrix is used to find the Broad and narrow weed classification accuracy.
Figure 7 shows the level wise execution time for different wavelet family. The Average elapse
time for different wavelet family up to three level of decomposition represented in Table 1. From
the Table 1, level 1 execution time for all the wavelet family is high compared to the level2 and
level3.The level1 execution time for Reverse Biorthogonal 3.3 wavelet is 55.88ms and Symlet8
family is 56.82ms , Daubechies4 wavelet is 58.16ms and Biorthogonal 6.8 wavelet is 63.02ms.
In level2 the execution for Reverse Biorthogonal3.3 wavelet is 14.60ms and Symlet8 family is
13.18ms, Daubechies4 wavelet is 11.32ms and Biorthogonal 6.8 wavelet is 15.77ms. The elapsed
time for level3 for the above wavelet family is 6.30ms, 6.85ms, 5.84ms and 6.20ms.
Figure 7: Level wise Average processing time (s) for wavelet algorithm
From Table 1 and Figure 8 shows the elapsed time for three level wavelet decomposition of weed
discrimination. In Daubechies (Db4) wavelet the elapsed time is 18.83ms and Reverse
Biorthogonal (rbio3.3) wavelet is 19.20ms which is less compared to the other wavelet
transform.Symlet4 has an highest elapsed time as 70.75ms and Biorthogonal wavelet as
26.85ms.The elapsed time for three level decomposition for Biorthogonal wavelet6.8 is 26.85ms
and Reverse biorthogonal6.8 is 21.27ms.Compare to all the wavelet Daubechies(Db4) has lowest
elapsed time as 18.83ms.
Table 1: Average Elapsed time for Wavelet Family
Wavelet Family Execution Time(ms)
Sym4 70.75
Sym8 19.21
Db4 18.83
Db8 21.92
Bior33 26.85
Bior68 21.66
Rbio33 19.20
Rbio68 21.27
9. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
49
Figure 8 : Execution time for Wavelet Family Decomposition upto three level
Broad and narrow weed texture feature set for wavelet families are classified with neural network.
The Figure 9 and Table 2 show the classification of broad and narrow weeds using different types
of wavelets at three levels of decompositions.
Table 2: Accuracy of Wavelet Texture Features
WMS :Wavelet statistical features(WMS), WCSPH : Contrast ,Cluster Shade, Cluster
Prominence and Local Homogeneity Texture feature WECSPE : Energy , Correlation, ,Cluster
Shade, Cluster Prominence and Entropy features
Wavelet
Family
WMS WCSPH WESPEnt
Sym4 76.67 91.67 96.33
Sym8 73.67 93.33 96.67
Db4 77.34 95.33 96.33
Db10 75.67 90.42 83.33
Bior33 76.87 90.67 96.67
Bior68 76.87 92.33 93.33
Rbio33 83.67 98.7 98.9
Rbio68 76.67 93.33 96.67
In Figure 9 the classification accuracy for the texture for Mean and standard deviation for the
wavelet family Sym4, Sym8, Db4, Db10, Bior33, Bior68, Rbio33 and Rbio68 are 76.67%,
73.67%, 77.34%,75.67%,76.87%,76.87% ,83.67 and 76.67. The wavelet Reverse Biorthogonal
wavelet3.3 wavelet have high accuracy as 83.67%. For the texture feature Contrast ,Cluster
Shade, Cluster Prominence and Local Homogeneity (WCSPH) the classification accuracy the
wavelet family are represented in Figure 9. Compare to the other wavelet the highest accuracy
86.67% achieve from the Reverse Biorthogonal3.3. It is obvious form the Table 2 and Figure. 9
that the classification through Reverse Biorthogonal3.3 wavelet (Rbio33) with Energy,
Correlation, ,Cluster Shade, Cluster Prominence and Entropy features (WESPEnt) texture features
10. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
50
is more efficient due to highest accuracy 98.9% and less execution time for three level of
decomposition.
Figure 9: Accuracy of Wavelet Texture Features as WMS, WCSPH and WECSPE
Comparison with Existing Algorithm
The classification accuracy of Broad and narrow weed was compared to the existing
algorithm the author [19] has proposed a real-time system had an accuracy of about 81% due to
the classification based on the highest intensity value taken from the histogram. The author [4]
developed a method using Offset Excess Green , Db4 wavelet transform and Multiresolution
combined statistical an spatial frequency features for the broad and narrow weed classification.
The overall classification accuracy is this method is 98.55% . Using Global Histogram
Equalization, Symlet wavelet and approximation coefficient of each level (four level) feature
values and SVM classification method, the accuracy of 98.1% can be obtained [9]. Our proposed
algorithm use Global Histogram equalisation, Offset excess green and thresholding techniques are
used for segmentation and three level wavelet decomposition of different wavelet families , and
three group of texture features are extracted detail and approximation coefficient of three levels
and Neural network and confusion matrix are used for classification. From the result of the
proposed algorithm Reverse Biorthogonal 3.3 wavelet families with texture features as Energy ,
Correlation, ,Cluster Shade, Cluster Prominence and Entropy features (WECSPE) gave the
accuracy of 98.9 %.The results are shown in Fig 10. Compared with the existing algorithm for
weed discrimination, our approach gave best results for weed discrimination and useful for site
specific herbicide application.
Fig. 10. The Comparison result for proposed algorithm and the existing algorithm
11. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
51
3. CONCLUSIONS AND FUTURE SCOPE
In this study, we proposed an accurate and efficient method for real-time specific weed
discrimination using Global Histograms Equalization, Excess Green Index pre-processing
techniques and multilevel (three level) wavelet decomposition to extract the prominent texture
features from images and Neural Network classification techniques. To increase the accuracy and
efficiency, we developed efficient pre-processing methods and feature extraction techniques. We
verified four different types of wavelet families Daubechies(db4,db10),Symlets(sym4,sym8) and
Reverse Biorthogonal (rbio3.3,rbio6.8) with three group of texture features are extracted and used
for Neural Network classification techniques. Our proposed approach using Reverse Biorthogonal
(rbio3.3) wavelet family with Energy, Correlation, Cluster Shade, Cluster Prominence and
Entropy (WECSPE) features yielded a classification accuracy of 98.9%. Compared to the existing
algorithm our proposed algorithm indicates that Rbio3.3 is a promising wavelet due to its high
accuracy, less error rate and elapsed time and it is suited for real-time specific weed
discrimination. To improve the classification accuracy curvelet can be used.
REFERENCES
[1] Woebbecke, D., Meyer, G., VonBargen, K., Mortensen, D.,(1995),”Color indices for weed
identification under various soil, residue, and lighting conditions”, Trans. ASAE, vol 38 (1), pp 271–
281.
[2] Ribeiro, A., Fernández-Quintanilla, C., Barroso, J., García-Alegre, M.C., (2005).”Development of an
image analysis system for estimation of weed”, Proceedings 5th
European Conf. On Precision
Agriculture (5ECPA) , pp. 169–174.
[3] Meyer, G.E., von Bargen, K., Woebbecke, D.M., Mortensen, D.A.,(1994),”Shape features for
identifying young weeds using image analysis”, ASAE paper 94-3019, American Society of
Agricultural Engineers, St. Joseph,MI, USA.
[4] R.S. Sabeenian and V. Palanisamy,(2010),”Crop and weed discrimination in agricultural field using
MRCSF”. Int. J. Signal and Imaging Systems Engineering, Vol. 3, No. 1, pp.61-69.
[5] Hague, T., Tillet, N., Wheeler, H.,(2006).”Automated crop and weed monitoring in widely spaced
cereals”, Precis. Agr, Vol 7, pp 21–32.
[6] Hemming, J., Rath, T.,(2001),”Computer-vision based weed identification under field conditions
using controlled lighting”,J. Agr. Eng. Res. Vol 78 (3), pp 233–243.
[7] Meyer, G., Neto, J.,(2008),”Verification of color vegetation indices for automated crop imaging
applications”,Comput. Electron. Agr. Vol 63, pp 282–293.
[8] Irshad Ahmad, Muhammad Hameed Siddiqi ,Iram Fatima ,Sungyoung Lee, Young-Koo Lee,(2011)
“Weed Classification Based on Haar Wavelet Transform via k-Nearest Neighbor (k-NN) for Real-
Time Automatic Sprayer Control System”,ICUIMC ’11, February 21–23, Seoul, Korea.
[9] Muhammad Hameed Siddiqi, Seok-Won Lee And Adil Mehmood Khan,(2014) ”Weed Image
Classification using Wavelet Transform,Stepwise Linear Discriminant Analysis, and Support Vector
Machines for an Automatic Spray Control System”, JOURNAL OF INFORMATION SCIENCE AND
ENGINEERING , Vol 30, pp 1253-1270.
[10] S. I. Cho; D. S. Lee; J. Y. Jeong, (2002),”Weed–plant Discrimination byMachine Vision and
Artificial Neural Network”, Biosystems Engineering, Vol. 83 (3), pp 275–280.
[11] Wang, N., Zhang, N., Dowell, F.E., Sun, Y., Peterson, D.E.,(2001),”Design of an optical weed sensor
using plant spectral characteristics”. Trans. ASAE, Vol. 44 (2), pp 409–419.
[12] Cho, S.I., Lee, D.S., Jeong, J.Y.,(2002),”Weed-plant discrimination by machine vision and artificial
neural network”,.Biosyst. Eng. Vol 83(3), pp 275–280.
[13] Lamm, R.D., Slaughter, D.C., Giles, D.K.,(2002),”Precision weed control system for cotton”,Trans.
ASAE, Vol 45 (1),pp 231–238.
[14] Haralick, R.M. (1979) ‘Statistical and structural approaches to texture’, IEEE Proceedings, Vol. 67,
pp.786–804.
[15] J. Bossu, Ch. Gée, G. Jones, F. Truchetet ,(2009),”Wavelet transform to discriminate between crop
and weed in perspective agronomic images”, computers and electronics in agriculture”,Vol
65,pp133–143.
12. International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015
52
[16] M. Guijarro, G. Pajares, I. Riomoros, P.J. Herrera, X.P. Burgos-Artizzue, A.
Ribeiroe,(2011),”Automatic segmentation of relevant textures in agricultural images”,Computers and
Electronics in Agriculture, Vol 75,pp 75–83.
[17] Siddiqi, M.H.,S.B.T Sulaiman,I,Faye and I.Ahmed(2009), “A Real Time Specific Weed
Discrimination System using Multi-Level Wavelet Decomposition”, International Journal of
Agricultural and Biology,Vol 11, pp 559-565.
[18] K.K.Ghazali ,M.F.Mansor,M.M.Mustafa and A.Hussain,(2007),”Feature Extraction Techniques using
Discrete Wavelet transform for Image classification”, IEEE 5th
Student conference on Research and
Development.
[19] C.Anucti and P.PBekaert,” SIFT-CCH:Increasing the SIFT distinctness by color co-occurrence
histogram” in proceeding of the 5th International symposium on Image and Signal Processing and
Analysis,2007,pp.130-135.
Authors
P.Prema received her B.Sc(Computer Science) degree in V.V.Vanniya Perumal College,Virudhunagar, in
1995 and MCA degree in Thiagarajar College of Engineering,Madurai and M.Phil (Computer Science) in
Manonmaniam Sundaranar University,Tirunelveli. She has 15 years of teaching experience and she is
currently working as Assistant Professor (Computer science) in Agricultural College an Research Institute,
Madurai, TamilNadu, India. Her current research interest includes Image Processing, Wavelet, Curvelet and
Neural Network.