Tips & Tricks
Like this document? Why not share!
Welcome to International Journal of...
by IJERD Editor
Pollination based optimization for ...
3 3 energy efficient topology
1 selvarasu nanjappan 3
4 image segmentation through cluste...
by IJERA Editor
Email sent successfully!
Show related SlideShares at end
Aug 12, 2012
Art & Photos
Comment goes here.
12 hours ago
Are you sure you want to
Your message goes here
Be the first to comment
Be the first to like this
Number of Embeds
No notes for slide
1. ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology Volume 1, Issue 5, July 2012 A Survey on Clustering Based Image Segmentation Santanu Bhowmik, Viki DattaAbstract – In computer vision, segmentation refers to II. CLUSTERINGthe process of partitioning a digital image intomultiple segments (Sets of pixels, also known as super Clustering is a process of organizing the objectspixels). This paper is a survey on various clustering into groups based on its attributes. A cluster istechniques to achieve image segmentation. In order to therefore a collection of objects which are “similar”increase the efficiency of the searching process, only a between them and are “dissimilar” to the objectspart of the database need to be searched. For this belonging to other clusters. An image can besearching process clustering techniques can be grouped based on keyword (metadata) or itsrecommended. Clustering can be termed here as a content (description).grouping of similar images in the database.Clustering is done based on different attributes of an In keyword based clustering, a keyword is a formimage such as size, color, texture etc. The purpose of of font which describes about the image keywordclustering is to get meaningful result, effective storage of an image refers to its different features. Theand fast retrieval in various areas. similar featured images are grouped to form a cluster by assigning value to each feature.Key Words – Clustering, Image segmentation, K- In content based clustering “, , ” ameans, N-cut, Spectral Clustering. content refers to shapes, textures or any other information that can be inherited from the image itself. The tools, techniques and algorithms that areI. INTRODUCTION used originate from fields such as statistics, patternClustering in image segmentation is defined as the recognition, signal processing etc. Clustering basedprocess of identifying groups of similar image on the optimization of an overall measure is aprimitive . Clustering techniques can be fundamental approach explored since the earlyclassified into supervised clustering-demands days of pattern recognition. The most popularhuman interaction to decide the clustering criteria method for pattern recognition is K-meansand the unsupervised clustering- decides the clustering.clustering criteria by itself. Supervised clusteringincludes hierarchical approaches such as relevance In K-means clustering a centroid vector isfeedback techniques “, ” and unsupervised computed for every cluster. The centroid must beclustering includes density based clustering chosen such that it should minimize the totalmethods. These clustering techniques are done to distance within the clusters.perform image segmentation. Segmentation is the Qprocess of partitioning a digital image into multiplesegments based on pixels. It is a critical andessential component of image analysis system. The Smain process is to represent the image in a clear T VV Qway. The result of image segmentation is acollection of segments which combine to form theentire image . Real world image segmentationproblems actually have multiple objectives such asminimize overall deviation, maximize connectivity, Pminimize the features or minimize the error rate of Uthe classifier etc .Image segmentation is a multiple objective Figure-1 Rproblem. It involves several processes such aspattern representation , feature selection, featureextraction and pattern proximity. Considering all Figure-1 shows the preferred centroid (V) for thethese objectives is a difficult problem, causing a triangle. The points namely S, T, U are the midpoint forgap between the natures of images. To bridge this corresponding edges.gap multi-objective optimization approach is anappropriate method “, , ”. 280 All Rights Reserved © 2012 IJARCET
ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology Volume 1, Issue 5, July 2012Both supervised and unsupervised clustering IV. CLUSTERING TECHNIQUEStechniques are used in image segmentation. In An image may contain more than one object and tosupervised clustering method, grouping is done segment the image in line with object features toaccording to user feedback. In unsupervised extract meaningful object has become a challengeclustering, the images with high features to the researches in the field. Segmentation can besimilarities to the query may be very different in achieved through clustering.terms of semantics . This is known as semantic This paper critically reviews and summarizesgap. To overcome this novel image retrieval different clustering techniques.scheme called as cluster based retrieval of imagesby unsupervised learning (CLUE) can be used .This works based on a hypothesis: semantically IV.1. Relevance feedback:similar images tend to be clustered in some feature A relevance feedback approach allows aspace. user to interact with the retrieval algorithm byA variety of clustering techniques have been providing the information of which images userintroduced to make the segmentation more thinks are relevant to the queryeffective. The clustering techniques which are “,,”.Keyword based image retrieval isincluded in this paper are relevance feedback , performed by matching keyword according to userlog based clustering , hierarchical clustering input and the images in the database., graph based, retrieval-dictionary based, filter Some images may not have appropriate keywordsbased clustering etc. to describe them and therefore the image search will become complex. One of the solution in order to overcome this problem is “relevance feedback”III. SEGMENTATION technique  that utilize user feedback and henceImage segmentation is the important process of reduces possible errors and redundancy “, ”.image analysis and image understanding . It is This technique uses a Bayesian classifier “,defined as the process of partitioning the digital ” which deals with positive and negativeimage into different sub regions of homogeneity. feedback. Content based clustering methods cannotThe objective of image segmentation is to cluster adapt to user changes, addition of new topics due topixels into salient image regions i.e., regions its static nature. To improve the performance ofcorresponding to individual surfaces, objects or information retrieval log-based clusteringnatural parts of objects. approaches are brought into the application.A segmentation might be used for objectrecognition “, ” image compression, image IV.2. Log –Based Clustering:editing, etc. The quality of the segmentationdepends upon the digital image . In the case of Images can be clustered based on thesimple images the segmentation process is clear retrieval system logs maintained by an informationand effective due to small pixels variations, retrieval process . The session keys are createdwhereas in the case of complex images, the utility and accessed for retrieval. Through this the sessionfor subsequent processing becomes questionable. clusters are created. Each session cluster generatesImage segmentation is one of the best known log –based document and similarity of imageproblems in computer vision. Graph based methods couple is retrieved. Log –based vector is created forwere earlier considered to be too insufficient in each session vector based on the log-basedpractice. Recent advances in technology and documents . Now, the session cluster isalgorithm “, ” have negated this replaced with this vector. The unaccessedassumption. Histogram “, , ” based document creates its own vector.methods are very effective while compared to other A hybrid matrix is generated with at least oneimage segmentation methods because they individual document vector and one log-basedtypically require only one pass through the pixels. clustered vector. At last the hybrid matrix isIn this method a histogram is computed from all of clustered. This technique is difficult to perform inthe pixels in the image and the peaks and valleys in the case of multidimensional images. To overcomethe histogram are used to locate the clustering of this hierarchical clustering is adopted.the image. Intensity can be used as the measure.This process is repeated with smaller and smaller IV.3. Hierarchical Clustering:clusters until no more clusters are formed. Thisapproach can be quickly adapted to multiple frames One of the well- known technologies inwhich is done in multiple fashion. information retrieval is hierarchical clustering . It is the process of integrating different images andSegmentation can also be done based on spatialcoherence . This includes two steps: Dividing building them as a cluster in the form of a tree andor merging existing regions from the image and then developing step by step in order to form a small cluster.growing regions from seed points. 281 All Rights Reserved © 2012 IJARCET
ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology Volume 1, Issue 5, July 2012The steps involved in this process are as follows: terminates when the bound on the number ofthe images from various databases are divided into clusters is reached or the Ncut value exceeds someX-sorts. The classification will be calculated by threshold T.modifying the cluster centers, sorts of the imagesand stored in the form of matrix m*m continuously 200which also includes dissimilarity values. At first itcalculates the similarities between the queried Vimage and the retrieved image in the image 70database. Secondly, it identifies the similarities 130between two closest images(In m*m matrix)and C1integrate them to form a cluster. Finally all the C2similarities are grouped to form a single cluster. 50 20 75 55IV.4. Retrieval Dictionary Based Clustering: A rough classification retrieval system is C7 C3 C8 C4formed. This is formed by calculating the distancebetween two learned patterns and these learnedpatterns are classified into different clusters 30 45followed by a retrieval stage. The main drawbackaddressed in this system is the determination of the C5 C6distance.To overcome this problem a retrieval system is Figure-2developed by retrieval dictionary based clustering. This method has a retrieval dictionary Figure-2 shows Ncut Nodes organized as groups.generation unit that classifies learned patterns intoplural clusters and creates a retrieval dictionary The recursive Ncut partition is essentially ausing the clusters. Here, the image is retrieved hierarchical divisive clustering process thatbased on the distance between two spheres with produces a tree . For example, Figure 2 showsdifferent radii. Each radius is a similarity measure a tree generated by four Recursive Ncuts. The firstbetween central cluster and an input image. An Ncut divides V into C1 and C2. Since C2 is largerimage which is similar to the query image will be than C1, the second Ncut partitions C2 into C3 andretrieved using retrieval dictionary. C4.Next, C3 is further divided because it is larger than C1 and C4. The fourth Ncut is applied to C1, and gives the final five clusters (or leaves): C4, C5,IV.5. K-Means Algorithm: C6, C7 and C8.The above example suggest trees as In K-means algorithm data vectors are a natural organization of clusters . Nonetheless,grouped into predefined number of clusters  the tree organization here may mislead a user. At the beginning the centroids of the because there is no guarantee of anypredefined clusters are initialized randomly. The correspondence between the tree and the semanticdimensions of the centroids are same as the structure of images. Furthermore, organizing imagedimension of the data vectors. Each pixel is clusters into a tree structure will significantlyassigned to the cluster based on the closeness , complicate the user interface.which is determined by the Euclidian distancemeasure. After all the pixels are clustered, themean of each cluster is recalculated. This process is V. CONCLUSOINrepeated until no significant changes result for each To summarize, a comprehensive surveycluster mean or for some fixed number of highlighting different clustering techniques usediterations. for image segmentation have been presented. Clustering concepts and image segmentationIV.6. Ncut Algorithm: concepts have been analyzed. Through clustering algorithms, image segmentation can be done in an Ncut method attempts to organize nodes effective way. Spectral clustering technique can beinto groups so that the within the group similarity is used for image clustering because images thathigh, and/or between the groups similarity is low. cannot be seen can be placed into clusters veryThis method is empirically shown to be relatively easily than other traditional methods . Inrobust in image segmentation . This method general, clustering is a hard problem. Clusteringcan be recursively applied to get more than two techniques helps to increase the efficiency of theclusters. In this method each time the sub graph image retrieval process.with maximum number of nodes is partitioned(random selection for tie breaking). The process 282 All Rights Reserved © 2012 IJARCET
ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology Volume 1, Issue 5, July 2012VI. REFERENCES  P. Felzenszwalb and D. Huttenlocher. Image segmentation using local variation. In IEEE Conference on Computer Vision Puzicha, J., Hofmann, T. and Buhmann, J. M.,”Histogram and Pattern Recognition, pages 98–104, 1998.Clustering for Unsupervised Image Segmentation”., Computer  Zhang, Q.,Goldman,S. A., Yu,W.,and Fritts, J.E..”ContentVision and Pattern Recognition , Vol.2.IEEEpress,602-608, based image retrieval using multiple instance learning”.In2000. Proc.ICML,2002. Zhou XS, Huang TS. Relevance feedback in image retrieval:  C. Rother, T. Minka, A. Balke and V. Kolmogorov,A comprehensive review. Multimedia Syst;8:536-544, 2003. “Cosegmentation of Image Pairs by Histogram Matching- Chundi, P., Dayal, U., Sayal, M., Hsu, M: Incorporating a Global-Constraint into MRFs”, proc. IEEEUS20077181678,2007. Conf. Computer Vision and Pattern Recognition, pp. 993-1000, C.Harris and M.Stephens, “A Combined Corner and Edge 2006.Detection,”Proc.Fourth Alvey Vision Conf.,pp.147-151,1988.  Jun Zhang and Jinglu Hu”Image segmentation based on 2D Saha, S., and Bandyopadhyay, S.: A new symmetry based Otsu method with Histogram analysis”, Int’l Conf. on Computermultiobjective clustering technique for automatic evolution of science and software engg, IEEE, 978-0-7695-3336-0/08, 2008.clusters. Pattern Recognition,2010.  Hadjidemetriou, E., Grossberg,M. D.,and Shirakawa, S., and Nagao, T., “Evolutionary Image Nayar,S.K.2004.”Multiresoultion Histogram and their use forSegmentation Based on Multiobjective Clustering”.Congress on recognition”.IEEE Trans.Pattern Analysis and MachineEvolutionary Computation(CEC ’09),Trondheim,Norway,2466- Intelligence 26,7,831-847,2004.2473,2009.  Zabih, R.; Kolmogorov, “Spatially coherent clustering Guliashki, V., Toshev, H., and Korsemov,C., Survey of using graph cuts”, V.;Cornell Univ., Ithaca, NY, USA, ISSN:Evolutionary Algorithms used in multiobjective optimization. 1063-6919, On page(s): II-437 - II-444 Vol.2.Problems of Engineering Cybernetics and Robotics, Bulgarian  I. J. Cox, M. L. Miller, T. P. Minka, T. V. Papathomas andAcademy of Sciences,2009. P.Yianilos, “The Bayesian Image Retrieval System, PicHunter: Jones,D.F.,Mirrazavi, S.K. and Tamiz,M.,Multi-objective Theory, Implementation and Psychophysical Experiments”,meta- heuristics:An Overview of the Current State-of the art. IEEE Trans. Image Processing, vol. 9, no. 1, pp. 20-37, 2000.European Journal of Operational Research, 137,1-9,2002.  Y. Rui, T. S. Huang, M. Ortega and S. Mehrotra, Coullo, C.A.C.,”A Comprehensive Survey of Evolutionary- “Relevance Feedback: A Power Tool for Interactive Content-Based Multiobjective Optimization Techniques”. Knowledge Based Image Retrieval”, IEEE Trans. Circuits and VideoAnd Information Systems,1,129-156,1998. Technology, vol. 8, no. 5, pp. 644-655, 1998 Yixin Chen, James Z.Wang, Robert Krovetz, “Content  Rui Y, Thomas, S. Huang. “Content-based image retrievalBased Image Retrieval by Clustering”,Proc of the 5th ACM with relevance feedback in MARS”, In proceedings of IEEESIGMM I’ntl workshop on Multimedia information retrieval international conference on image processing, pp. 815-818,,New York,ACM press,pp-193-200,2003. 1997. Huiyu Zhou, Abdul H. Sadka, Mohammad R. Swash, Jawid  Zhou XS, Huang TS. Relevance feedback in imageAzizi and Abubakar S. Umar., “Content Based Image Retrieval retrieval: A comprehensive review. Multimedia Syst; 8: 536-and Clustering: A Brief Survey” school of Engineering and 544, 2003.Design, Brunel University, Uxbridge, UB8 3PH, UK  Irani, A.A.Z. Belaton, “A K-means Based Generic D. Melas and S. Wilson, “Double Markov Random Fields Segmentation System B.Dept. of Comput. Sci., Univ. Sainsand Bayesian Image Segmentation,” IEEE Trans. Signal Malaysia, Nibong Tebal, Malaysia Print ISBN: 978-0-7695-Processing, vol. 50, no. 2, pp. 357-365, Feb. 2002. 3789-4 On page(s): 300 – 307, 2009. Wang JZ, Li J, Wiederhold G. Simplicity: Semantics-  Li Wenchao Zhou Yong Xia Shixiong China Univ. of Min.sensitive integrated matching for picture libraries. IEEE Trans & Technol., Xuzhou, “A Novel Clustering Algorithm Based onpattern Analysis Machine Intell;23:947-963, 2001. Hierarchical and K-means Clustering” Print ISBN: 978-7- Jin J, Kurniawati R, Xu G, Bai X. Using browsing to 81124-055-9, On page(s): 605, 2009.improve content-based image retrieval. J Visual Common Image  Isa, N.A.M.; Salamah, S.A.; Ngah, U.K.; Sch. of Electr. &Represent;12:123-135, 2001. Electron. Eng., Univ. Sains Malaysia, Nibong Tebal, Malaysia , Huang Min,Sun bo,Xi Jianqing”An Optimized image “Adaptive fuzzy moving K-means clustering algorithm forretrieval method based on Hierarchical clustering and genetic image segmentation” ISSN: 0098-3063 ,On page(s): 2145 –algorithm”I’ntl forum on Information technology and 2153, 2009.applications,978-0-7695-3600-2/09-IEEE,2009.  Feng SUN, Jin-Peng HE, “A Normalized Cuts Based Image K. Barnard and D. Forsyth, “Learing the Semantics of Segmentation Method”, Dept. Automation Harbin EngineeringWords and Pictures”, proc. 8th Int’l Conf. on Computer Vision, University, Harbin, China, 2009 Second Internationalvol. 2, pp. 408-415,2001. Conference on Information and Computer Science,2009. O.Stehling,M.A. Nascmento,X.A.Falcgo “An Adaptive and  Jianbo Shi Malik, J.Robotics Inst., Carnegie Mellon Univ.,efficient clustering based approach for Content Based Image Pittsburgh, PA ,” Normalized cuts and image segmentation”,Retrieval in image databases”,Proc of I’ntl database engg and ISSN: 0162-8828, On page(s): 888 – 905,2000.application symposium,pp-356-365,2001.  Emma Regentova, Dongsheng Yao, and Shahram Latifi, Shi. J, and Malik. J, “Normalized cuts and image “Image segmentation using Ncut in the wavelet domain”,segmentation”, IEEE Conf. Computer Vision and Pattern International journal of image and graphics,world scientificRecognition, IEEE Computer Publishing company,vol.6(4),pp.569-582,2006.  LI XiaoBin and TIAN Zheng”Multiscale stochastic J. Winn and J.Shotton, “The Layout Consistent Random hierarchical image segmentation by spectral clustering”,SciField for Recognizing and Segmenting Partially Occluded China Ser F-Inf Sci, vol.50 |no.2|pg.no-198-211,2007.Objects”, proc. IEEE CS Conf. Computer Vision and Pattern  Cox, I. J., Miller, M. L., Minka, T. P., Papathomas, T. V.Recognition, vol. 1,pp. 37-44, 2006. and Yianilos, P. N. The Bayesian image retrieval system, P. Alvarado, A. Berner and S. Akyol, “Combination of pichunter: Theory, implementation and psychophysicalHigh-Level Cues in Unsupervised Single Image Segmentation experiments. IEEE Tras. Image processing 9, 1, 20-37, 2000.Using Bayesian Belief Networks”, proc. Int’l Conf Imaging  Hoi, C.-H. and Lyu, M. R. 2004a. “Group-based relevanceScience, Systems and Techonolgy, vol. 2, pp. 675-681,2002. feeedbacks with support vector machine ensembles. In Proc. X. Feng, C. Williams, and S. Felderhof, “Combining Belief IEEE ICPR,2004.Networks and Neural Networks for Scene Segmentation”, IEEE  Hoi, C.-H. and Lyu, M. R. 2004b. A novel log basedTrans., Pattern Analysis and Machine Intelligence, vol. 24, no. relevance feedback technique in content based image retrieval.4, pp. 467-483,Apr. 2002. In Proc. ACM Multimedia,2004. 283 All Rights Reserved © 2012 IJARCET
ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology Volume 1, Issue 5, July 2012Authors –Mr. Santanu Bhowmik, M.Tech, MCA,Ph.D Scholar, NIT AgartalaShan.firstname.lastname@example.org.Mr. Viki Datta, MCATechnical Asstt., NIT Agartalavdatta49@gmail.com 284 All Rights Reserved © 2012 IJARCET
Email sent successfully..