Dependent Random Access Point Pictures in HEVCEricsson
This paper describes the concept and a number of system aspects for the Dependent Random Access Point (DRAP) picture, a new feature that was recently introduced in the High-Efficiency Video Coding (HEVC) standard.
Filmic Tone Mapping, a presentation at Electronic Arts on a technique from film that became very applicable to games with the addition of support for HDR lighting and rendering in graphics cards.
High-quality point clouds have recently gained interest as an emerg- ing form of representing immersive 3D graphics. Unfortunately, these 3D media are bulky and severely bandwidth intensive, which makes it difficult for streaming to resource-limited and mobile de- vices. This has called researchers to propose efficient and adaptive approaches for streaming of high-quality point clouds.
In this paper, we run a pilot study towards dynamic adaptive point cloud streaming, and extend the concept of dynamic adaptive streaming over HTTP (DASH) towards DASH-PC, a dynamic adaptive bandwidth-efficient and view-aware point cloud streaming system. DASH-PC can tackle the huge bandwidth demands of dense point cloud streaming while at the same time can semantically link to human visual acuity to maintain high visual quality when needed. In order to describe the various quality representations, we pro- pose multiple thinning approaches to spatially sub-sample point clouds in the 3D space, and design a DASH Media Presentation Description manifest specific for point cloud streaming. Our initial evaluations show that we can achieve significant bandwidth and performance improvement on dense point cloud streaming with minor negative quality impacts compared to the baseline scenario when no adaptations is applied.
Slides from "hands on video" course
explains the color model as well as YUV decimation and packing in theory and practice using FFMPEG, FFPROBE and YUV player
Dependent Random Access Point Pictures in HEVCEricsson
This paper describes the concept and a number of system aspects for the Dependent Random Access Point (DRAP) picture, a new feature that was recently introduced in the High-Efficiency Video Coding (HEVC) standard.
Filmic Tone Mapping, a presentation at Electronic Arts on a technique from film that became very applicable to games with the addition of support for HDR lighting and rendering in graphics cards.
High-quality point clouds have recently gained interest as an emerg- ing form of representing immersive 3D graphics. Unfortunately, these 3D media are bulky and severely bandwidth intensive, which makes it difficult for streaming to resource-limited and mobile de- vices. This has called researchers to propose efficient and adaptive approaches for streaming of high-quality point clouds.
In this paper, we run a pilot study towards dynamic adaptive point cloud streaming, and extend the concept of dynamic adaptive streaming over HTTP (DASH) towards DASH-PC, a dynamic adaptive bandwidth-efficient and view-aware point cloud streaming system. DASH-PC can tackle the huge bandwidth demands of dense point cloud streaming while at the same time can semantically link to human visual acuity to maintain high visual quality when needed. In order to describe the various quality representations, we pro- pose multiple thinning approaches to spatially sub-sample point clouds in the 3D space, and design a DASH Media Presentation Description manifest specific for point cloud streaming. Our initial evaluations show that we can achieve significant bandwidth and performance improvement on dense point cloud streaming with minor negative quality impacts compared to the baseline scenario when no adaptations is applied.
Slides from "hands on video" course
explains the color model as well as YUV decimation and packing in theory and practice using FFMPEG, FFPROBE and YUV player
Advanced Mechanisms for Delivering High-Quality Digital ContentMikołaj Leszczuk
In this paper, a practical solution for optimal coding parameters using different bandwidth requirements is presented. The obtained specification is based on the analysis of a large database of more than 10,000 sequences compressed with different parameters. The obtained parameters can be used both for adaptive streaming or storage optimization.
Fixation Prediction for 360° Video Streaming in Head-Mounted Virtual RealityWen-Chih Lo
Published in NOSSDAV'17 on June 2017.
We study the problem of predicting the Field-of-Views (FoVs) of viewers watching 360° videos using commodity Head-Mounted Displays (HMDs). Existing solutions either use the viewer's current orientation to approximate the FoVs in the future, or extrapolate future FoVs using the historical orientations and dead-reckoning algorithms. In this paper, we develop fixation prediction networks that concurrently leverage sensor- and content-related features to predict the viewer fixation in the future, which is quite different from the solutions in the literature. The sensor-related features include HMD orientations, while the content-related features include image saliency maps and motion maps. We build a 360° video streaming testbed to HMDs, and recruit twenty-five viewers to watch ten 360° videos. We then train and validate two design alternatives of our proposed networks, which allows us to identify the better-performing design with the optimal parameter settings.
Trace-driven simulation results show the merits of our proposed fixation prediction networks compared to the existing solutions, including: (i) lower consumed bandwidth, (ii) shorter initial buffering time, and (iii) short running time.
We develop custom Image Recognition systems for Aerospace and defence applications. Using algorithms like Deep Convolutional Neural Networks and Regional Convolutional Neural Networks.
Our algorithms for Target Recognition and Tracking are designed from the beginning to be run on embedded systems. We target both GPU and FPGA devices.
To Train and Validate our algorithms we developed a process to generate photorealistic 3D environments.
Those 3D Environments are used to produce realistic video streams of the targets in different environmental conditions (lighting, adverse meteorological conditions, camouflage, point-of-view).
The same technology can be used to Train and Test Automotive Vision Systems.
Real-time Bangla License Plate Recognition System for Low Resource Video-base...MD Abdullah Al Nasim
Automatic License Plate Recognition systems aim to provide a solution for detecting, localizing, and recognizing license plate characters from vehicles appearing in video frames. However, deploying such systems in the real world requires real-time performance in low-resource environments. In our paper, we propose a two-stage detection pipeline paired with Vision API that provides real-time inference speed along with consistently accurate detection and recognition performance. We used a haar-cascade classifier as a filter on top of our backbone MobileNet SSDv2 detection model. This reduces inference time by only focusing on high confidence detections and using them for recognition. We also impose a temporal frame separation strategy to distinguish between multiple vehicle license plates in the same clip. Furthermore, there are no publicly available Bangla license plate datasets, for which we created an image dataset and a video dataset containing license plates in the wild. We trained our models on the image dataset and achieved an AP (0.5) score of 86% and tested our pipeline on the video dataset and observed reasonable detection and recognition performance (82.7% detection rate, and 60.8% OCR F1 score) with real-time processing speed (27.2 frames per second).
Multispectral imaging in Forensics with VideometerLab 3Adrian Waltho
Multispectral imaging provides valuable information on the quality and safety of a vast array of materials from pharmaceuticals to raw meat and burned biscuits. The same techniques can be used to measure skin sensitivity to sticking plaster, detect counterfeit drugs and packaging and gain insight into historical artefacts such as medieval manuscripts and weapons.
The VideometerLab 3 takes 19 images at 19 specific wavelengths in ultraviolet, visible and infrared light. With multivariate statistical analysis we can reveal hidden patterns in our 19-dimension colour space and find information that would be otherwise 'invisible' to our human eyes.
Advanced Mechanisms for Delivering High-Quality Digital ContentMikołaj Leszczuk
In this paper, a practical solution for optimal coding parameters using different bandwidth requirements is presented. The obtained specification is based on the analysis of a large database of more than 10,000 sequences compressed with different parameters. The obtained parameters can be used both for adaptive streaming or storage optimization.
Fixation Prediction for 360° Video Streaming in Head-Mounted Virtual RealityWen-Chih Lo
Published in NOSSDAV'17 on June 2017.
We study the problem of predicting the Field-of-Views (FoVs) of viewers watching 360° videos using commodity Head-Mounted Displays (HMDs). Existing solutions either use the viewer's current orientation to approximate the FoVs in the future, or extrapolate future FoVs using the historical orientations and dead-reckoning algorithms. In this paper, we develop fixation prediction networks that concurrently leverage sensor- and content-related features to predict the viewer fixation in the future, which is quite different from the solutions in the literature. The sensor-related features include HMD orientations, while the content-related features include image saliency maps and motion maps. We build a 360° video streaming testbed to HMDs, and recruit twenty-five viewers to watch ten 360° videos. We then train and validate two design alternatives of our proposed networks, which allows us to identify the better-performing design with the optimal parameter settings.
Trace-driven simulation results show the merits of our proposed fixation prediction networks compared to the existing solutions, including: (i) lower consumed bandwidth, (ii) shorter initial buffering time, and (iii) short running time.
We develop custom Image Recognition systems for Aerospace and defence applications. Using algorithms like Deep Convolutional Neural Networks and Regional Convolutional Neural Networks.
Our algorithms for Target Recognition and Tracking are designed from the beginning to be run on embedded systems. We target both GPU and FPGA devices.
To Train and Validate our algorithms we developed a process to generate photorealistic 3D environments.
Those 3D Environments are used to produce realistic video streams of the targets in different environmental conditions (lighting, adverse meteorological conditions, camouflage, point-of-view).
The same technology can be used to Train and Test Automotive Vision Systems.
Real-time Bangla License Plate Recognition System for Low Resource Video-base...MD Abdullah Al Nasim
Automatic License Plate Recognition systems aim to provide a solution for detecting, localizing, and recognizing license plate characters from vehicles appearing in video frames. However, deploying such systems in the real world requires real-time performance in low-resource environments. In our paper, we propose a two-stage detection pipeline paired with Vision API that provides real-time inference speed along with consistently accurate detection and recognition performance. We used a haar-cascade classifier as a filter on top of our backbone MobileNet SSDv2 detection model. This reduces inference time by only focusing on high confidence detections and using them for recognition. We also impose a temporal frame separation strategy to distinguish between multiple vehicle license plates in the same clip. Furthermore, there are no publicly available Bangla license plate datasets, for which we created an image dataset and a video dataset containing license plates in the wild. We trained our models on the image dataset and achieved an AP (0.5) score of 86% and tested our pipeline on the video dataset and observed reasonable detection and recognition performance (82.7% detection rate, and 60.8% OCR F1 score) with real-time processing speed (27.2 frames per second).
Multispectral imaging in Forensics with VideometerLab 3Adrian Waltho
Multispectral imaging provides valuable information on the quality and safety of a vast array of materials from pharmaceuticals to raw meat and burned biscuits. The same techniques can be used to measure skin sensitivity to sticking plaster, detect counterfeit drugs and packaging and gain insight into historical artefacts such as medieval manuscripts and weapons.
The VideometerLab 3 takes 19 images at 19 specific wavelengths in ultraviolet, visible and infrared light. With multivariate statistical analysis we can reveal hidden patterns in our 19-dimension colour space and find information that would be otherwise 'invisible' to our human eyes.
Target Detection and Classification Performance Enhancement using Super-Resol...sipij
Long range infrared videos such as the Defense Systems Information Analysis Center (DSIAC) videos usually
do not have high resolution. In recent years, there are significant advancement in video super-resolution
algorithms. Here, we summarize our study on the use of super-resolution videos for target detection and
classification. We observed that super-resolution videos can significantly improve the detection and
classification performance. For example, for 3000 m range videos, we were able to improve the average
precision of target detection from 11% (without super-resolution) to 44% (with 4x super-resolution) and the
overall accuracy of target classification from 10% (without super-resolution) to 44% (with 2x superresolution).
TARGET DETECTION AND CLASSIFICATION PERFORMANCE ENHANCEMENT USING SUPERRESOLU...sipij
Long range infrared videos such as the Defense Systems Information Analysis Center (DSIAC) videos usually
do not have high resolution. In recent years, there are significant advancement in video super-resolution
algorithms. Here, we summarize our study on the use of super-resolution videos for target detection and
classification. We observed that super-resolution videos can significantly improve the detection and
classification performance. For example, for 3000 m range videos, we were able to improve the average
precision of target detection from 11% (without super-resolution) to 44% (with 4x super-resolution) and the
overall accuracy of target classification from 10% (without super-resolution) to 44% (with 2x superresolution).
TARGET DETECTION AND CLASSIFICATION PERFORMANCE ENHANCEMENT USING SUPERRESOLU...sipij
Long range infrared videos such as the Defense Systems Information Analysis Center (DSIAC) videos usually
do not have high resolution. In recent years, there are significant advancement in video super-resolution
algorithms. Here, we summarize our study on the use of super-resolution videos for target detection and
classification. We observed that super-resolution videos can significantly improve the detection and
classification performance. For example, for 3000 m range videos, we were able to improve the average
precision of target detection from 11% (without super-resolution) to 44% (with 4x super-resolution) and the
overall accuracy of target classification from 10% (without super-resolution) to 44% (with 2x superresolution).
Towards diagnosis of rotator cuff tears in 3-D MRI using 3-D convolutional ne...Wesley De Neve
Towards diagnosis of rotator cuff tears in 3-D MRI using 3-D convolutional neural networks. Paper presented at the Workshop on Computational Biology at the International Conference on Machine Learning, Long Beach, USA, 2019.
Investigating the biological relevance in trained embedding representations o...Wesley De Neve
Investigating the biological relevance in trained embedding representations of protein sequences. Paper presented at the Workshop on Computational Biology at the International Conference on Machine Learning, Long Beach, USA, 2019.
Towards reading genomic data using deep learning-driven NLP techniquesWesley De Neve
Towards reading genomic data using deep learning-driven NLP techniques. Slides presented at BIOINFO 2016 – Precision Bioinformatics & Machine Learning.
Deep Machine Learning for Making Sense of Biotech Data - From Clean Energy to...Wesley De Neve
Deep Machine Learning for Making Sense of Biotech Data - From Clean Energy to Smart Farming. Presentation given at the Korea-Europe International Conference on the 4th Industry Revolution.
Biotech Data Science @ GUGC in Korea: Deep Learning for Prediction of Drug-Ta...Wesley De Neve
Biotech Data Science @ GUGC in Korea: Deep Learning for Prediction of Drug-Target Interaction and DNA Analysis.
Poster presented at the BIG N2N Symposium 2016.
Towards using multimedia technology for biological data processingWesley De Neve
Towards using multimedia technology for biological data processing.
Presentation given during the Ghent University Global Campus (GUGC) Research Seminar on 19/1/2014.
In his public lecture, Christian Timmerer provides insights into the fascinating history of video streaming, starting from its humble beginnings before YouTube to the groundbreaking technologies that now dominate platforms like Netflix and ORF ON. Timmerer also presents provocative contributions of his own that have significantly influenced the industry. He concludes by looking at future challenges and invites the audience to join in a discussion.
Unlocking Productivity: Leveraging the Potential of Copilot in Microsoft 365, a presentation by Christoforos Vlachos, Senior Solutions Manager – Modern Workplace, Uni Systems
GraphRAG is All You need? LLM & Knowledge GraphGuy Korland
Guy Korland, CEO and Co-founder of FalkorDB, will review two articles on the integration of language models with knowledge graphs.
1. Unifying Large Language Models and Knowledge Graphs: A Roadmap.
https://arxiv.org/abs/2306.08302
2. Microsoft Research's GraphRAG paper and a review paper on various uses of knowledge graphs:
https://www.microsoft.com/en-us/research/blog/graphrag-unlocking-llm-discovery-on-narrative-private-data/
GridMate - End to end testing is a critical piece to ensure quality and avoid...ThomasParaiso2
End to end testing is a critical piece to ensure quality and avoid regressions. In this session, we share our journey building an E2E testing pipeline for GridMate components (LWC and Aura) using Cypress, JSForce, FakerJS…
State of ICS and IoT Cyber Threat Landscape Report 2024 previewPrayukth K V
The IoT and OT threat landscape report has been prepared by the Threat Research Team at Sectrio using data from Sectrio, cyber threat intelligence farming facilities spread across over 85 cities around the world. In addition, Sectrio also runs AI-based advanced threat and payload engagement facilities that serve as sinks to attract and engage sophisticated threat actors, and newer malware including new variants and latent threats that are at an earlier stage of development.
The latest edition of the OT/ICS and IoT security Threat Landscape Report 2024 also covers:
State of global ICS asset and network exposure
Sectoral targets and attacks as well as the cost of ransom
Global APT activity, AI usage, actor and tactic profiles, and implications
Rise in volumes of AI-powered cyberattacks
Major cyber events in 2024
Malware and malicious payload trends
Cyberattack types and targets
Vulnerability exploit attempts on CVEs
Attacks on counties – USA
Expansion of bot farms – how, where, and why
In-depth analysis of the cyber threat landscape across North America, South America, Europe, APAC, and the Middle East
Why are attacks on smart factories rising?
Cyber risk predictions
Axis of attacks – Europe
Systemic attacks in the Middle East
Download the full report from here:
https://sectrio.com/resources/ot-threat-landscape-reports/sectrio-releases-ot-ics-and-iot-security-threat-landscape-report-2024/
Communications Mining Series - Zero to Hero - Session 1DianaGray10
This session provides introduction to UiPath Communication Mining, importance and platform overview. You will acquire a good understand of the phases in Communication Mining as we go over the platform with you. Topics covered:
• Communication Mining Overview
• Why is it important?
• How can it help today’s business and the benefits
• Phases in Communication Mining
• Demo on Platform overview
• Q/A
Generative AI Deep Dive: Advancing from Proof of Concept to ProductionAggregage
Join Maher Hanafi, VP of Engineering at Betterworks, in this new session where he'll share a practical framework to transform Gen AI prototypes into impactful products! He'll delve into the complexities of data collection and management, model selection and optimization, and ensuring security, scalability, and responsible use.
The Art of the Pitch: WordPress Relationships and SalesLaura Byrne
Clients don’t know what they don’t know. What web solutions are right for them? How does WordPress come into the picture? How do you make sure you understand scope and timeline? What do you do if sometime changes?
All these questions and more will be explored as we talk about matching clients’ needs with what your agency offers without pulling teeth or pulling your hair out. Practical tips, and strategies for successful relationship building that leads to closing the deal.
Threats to mobile devices are more prevalent and increasing in scope and complexity. Users of mobile devices desire to take full advantage of the features
available on those devices, but many of the features provide convenience and capability but sacrifice security. This best practices guide outlines steps the users can take to better protect personal devices and information.
PHP Frameworks: I want to break free (IPC Berlin 2024)Ralf Eggert
In this presentation, we examine the challenges and limitations of relying too heavily on PHP frameworks in web development. We discuss the history of PHP and its frameworks to understand how this dependence has evolved. The focus will be on providing concrete tips and strategies to reduce reliance on these frameworks, based on real-world examples and practical considerations. The goal is to equip developers with the skills and knowledge to create more flexible and future-proof web applications. We'll explore the importance of maintaining autonomy in a rapidly changing tech landscape and how to make informed decisions in PHP development.
This talk is aimed at encouraging a more independent approach to using PHP frameworks, moving towards a more flexible and future-proof approach to PHP development.
GraphSummit Singapore | The Future of Agility: Supercharging Digital Transfor...Neo4j
Leonard Jayamohan, Partner & Generative AI Lead, Deloitte
This keynote will reveal how Deloitte leverages Neo4j’s graph power for groundbreaking digital twin solutions, achieving a staggering 100x performance boost. Discover the essential role knowledge graphs play in successful generative AI implementations. Plus, get an exclusive look at an innovative Neo4j + Generative AI solution Deloitte is developing in-house.
UiPath Test Automation using UiPath Test Suite series, part 4DianaGray10
Welcome to UiPath Test Automation using UiPath Test Suite series part 4. In this session, we will cover Test Manager overview along with SAP heatmap.
The UiPath Test Manager overview with SAP heatmap webinar offers a concise yet comprehensive exploration of the role of a Test Manager within SAP environments, coupled with the utilization of heatmaps for effective testing strategies.
Participants will gain insights into the responsibilities, challenges, and best practices associated with test management in SAP projects. Additionally, the webinar delves into the significance of heatmaps as a visual aid for identifying testing priorities, areas of risk, and resource allocation within SAP landscapes. Through this session, attendees can expect to enhance their understanding of test management principles while learning practical approaches to optimize testing processes in SAP environments using heatmap visualization techniques
What will you get from this session?
1. Insights into SAP testing best practices
2. Heatmap utilization for testing
3. Optimization of testing processes
4. Demo
Topics covered:
Execution from the test manager
Orchestrator execution result
Defect reporting
SAP heatmap example with demo
Speaker:
Deepak Rai, Automation Practice Lead, Boundaryless Group and UiPath MVP
UiPath Test Automation using UiPath Test Suite series, part 4
Improved license plate recognition for low resolution cctv forensics by integrating sparse representation-based super-resolution
1. Improved License Plate Recognition for
Low-Resolution CCTV Forensics by Integrating
Sparse Representation-based Super-Resolution
12th International Workshop on
Digital Forensics and Watermarking
Auckland, New Zealand
October 4, 2013
Hyun-seok Min, Seung Ho Lee,
Wesley De Neve, Yong Man Ro
Image and Video Systems Lab
Department of Electrical Engineering
Korea Advanced Institute of Science and Technology (KAIST)
Daejeon, Republic of Korea
2. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Overview
• Introduction
• Proposed method for license plate recognition
• Experimental results
• Conclusions
2/20
3. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
License Plate Recognition
• Automatic license plate recognition (LPR)
– important for CCTV forensics
– important for privacy protection
3/20
Large-scale Privacy Protection
in Google Street View
LPR in a CCTV system
4. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Challenges
• Low effectiveness of LPR for real-world imagery
– due to difficult capturing conditions
• E.g., frequent use of long-distance CCTV cameras
– output noisy LP images with a low resolution
4/20
Example input images captured by long-distance CCTV cameras in Korea
5. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Super-Resolution (1/2)
• Super-resolution can be used to improve the
effectiveness of current LPR systems
– enhances images by producing high-resolution
images from one or more low-resolution images
5/20
6. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Super-Resolution (2/2)
• Shortcomings of current approaches
– focus on maximizing the SNR between the original
high-resolution image and the reconstructed image
• do not focus on maximizing the effectiveness of LPR
– focus on the use of general image information
• do not take advantage of specific image priors, such as the
font, location, and color of LP characters
6/20
Super-resolution
Training images
Similar?
7. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Proposed LPR Method (1/2)
• Uses high-resolution LP images for
1) creating a dictionary for sparse representation-based
(SR-based) super-resolution of the test LP images
2) the training of LP character classifiers (SVM)
7/20
Proposed LPR system
License
plate
detection
Sparse
representation-
based super-
resolution
Image patchesInput image
High-quality
training images
...
...
...
...
...
...
...
...
Classifier for 1
Classifier for 2
...
Classifier for 0
Recognition of
license plate
characters
...
Output
CCTV
8. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Proposed LPR Method (2/2)
• Advantages
– takes into account specific image priors
• e.g., the font, location, and color of LP characters
– shares specific image priors among different steps
• SR-based resolution enhancement of the test LP images
• SVM-based LP character recognition
8/20
9. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Sparse Representation-based
Super-Resolution for LPR (1/2)
• We enhance the quality of noisy and low-resolution
LP images through SR-based super-resolution
– to that end, we make use of a dictionary that contains a
large number of LP image patches that have been
extracted from high-quality LP images
9/20
Kd
Ki
×
= R],...,,...,[ 1 ∈xxxD
xi : the feature vector of the ith LP image patch extracted
from the set of high-resolution training images
d : the dimensionality of the feature vectors
K: the number of LP image patches in the dictionary
10. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Sparse Representation-based
Super-Resolution for LPR (2/2)
10/20
License
plate
detection
Sparse
representation-
based super-
resolution
Image patches
High-quality
training images
...
...
...
...
...
...
...
...
11. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Setup:
Test Set (1/2)
• Test set characteristics
– 600 input images captured by CCTV cameras in Korea
– resolution of input images: 1280×720
– we extracted 772 LP images from the 600 input images
by making use of sliding concentric windows (SCW)
• each input image contains one or more LP images
11/20
Example input images captured by long-distance CCTV cameras in Korea
12. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Setup:
Test Dataset (2/2)
• We divided the LP images in the test set into four
groups, taking into account their spatial resolution
1) Resolution Group 1: LP image width ≤ 40 pixels
2) Resolution Group 2: 40 pixels < LP image width ≤ 55 pixels
3) Resolution Group 3: 55 pixels < LP image width ≤ 70 pixels
4) Resolution Group 4: 70 pixels < LP image width
12/20
Resolution Group 1 Resolution Group 2 Resolution Group 3 Resolution Group 4
13. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Setup:
Dictionary Construction
• Dictionary construction (& classifier training)
– we made use of 100 high-resolution LP images that
have been downloaded via Google Image Search
– we randomly extracted 10,000 image patches from the
training images
• image patches have a size of 10×10 pixels
– we made use of a dictionary with 10,000 image patches
13/20
Example high-resolution LP images
14. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Results:
Influence of Resolution Enhancement (1/2)
14/20
15. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Results:
Influence of Resolution Enhancement (2/2)
• Observation
– the effectiveness of the proposed LPR method is
better than the effectiveness of
• LPR without resolution enhancement
• LPR using bicubic interpolation
• Visual examples
15/20
(a) original low-resolution LP images (b) LP images that are the
output of bicubic
interpolation
(c) LP images that are the
output of SR-based super-
resolution
45×22 56×25
16. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Results:
Dictionary Influence (1/2)
• We enhanced the resolution of the test LP
images by means of two types of dictionaries
1) a dictionary consisting of high-resolution LP images
2) a dictionary consisting of general images
16/20
Examples of general imagesExample high-resolution LP images
17. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Results:
Dictionary Influence (2/2)
• Observation
– SR-based super-resolution by means of high-quality
LP training images is more effective than SR-based
super-resolution by means of general images
17/20
18. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Experimental Results:
Influence of Low-Level Visual Features
• Observation
– color Gabor wavelet features are most effective
18/20
19. Improved License Plate Recognition for Low-Resolution CCTV Forensics by
Integrating Sparse Representation-based Super-Resolution
Conclusions
• We introduced a novel LPR method
– applies SR-based super-resolution to test LP images
– uses a common set of high-quality LP images to
facilitate
• SR-based super-resolution of LP test images
• SVM-based LP character recognition
• Our results indicate that the proposed method
allows improving the effectiveness of LPR in
real-world imagery
19/20
20. Any questions or comments?
Contact information
Hyun-seok Min: hsmin@kaist.ac.kr
Web: http://ivylab.kaist.ac.kr
20/20