Industrial Consultancy & Sponsored Research (IC&SR) , IIT Madras
Method and Apparatus for Tracking of Object in Set of Video Frames
Categories for this Invention
Category – Computer Vision, Object Tracking
Applications – Video Surveillance, Visual Odometry, Optical Flow, Stereo Vision, Structure from Motion (SfM), Simultaneous Localization and Mapping (SLAM)
Industry – Security and Surveillance, Autonomous Vehicles, Robotics, Augmented Reality, Virtual Reality, Healthcare Imaging
Market -The Global Video Surveillance Systems Market size is estimated at USD 81.68 billion in 2024, and is expected to reach USD 145.38 billion by 2029, growing at a CAGR of 12.22% during the forecast period (2024-2029).
Image Gallery
Problem Statement
- Existing object tracking techniques, particularly corner point tracking, struggle with accuracy at object boundaries and depth discontinuities in video frames.
- There is a demand for a more robust and efficient object tracking method that can accurately track objects across frames, especially at boundaries, in computer vision and video surveillance applications.
Technology
Maximally Stable Level Line Segments (MSLLS):
- Key technology for robustly identifying stable features in video frames, enhancing object tracking accuracy.
Shape-Based Matching:
- Technique to compare shapes of object features across frames, aiding in reliable object identification and tracking.
Texture-Based Matching:
- Method to analyze and compare texture patterns of object regions, improving matching accuracy in varying conditions.
Part SSD Matching:
- Substantial Sum-of-Squared-Differences matching technique for comparing texture patches, enhancing object tracking performance in complex scenes.
Corner Point Detection:
- Algorithms for identifying distinctive corner points in images, facilitating precise object localization and tracking.
Key Features / Value Proposition
User Perspective:
- Enhanced object tracking for security.
- Improved navigation for robotics.
- Seamless virtual element integration in AR.
Technical Perspective:
- Novel maximally stable level line segment method.
- Flexible integration into computer vision systems.
- Advanced shape and texture-based matching.
Industry Perspective:
- Enhanced security surveillance.
- Improved autonomous vehicle navigation.
- Immersive AR and gaming experiences.
Questions about this Technology?
Contact For Licensing
sm-marketing@imail.iitm.ac.in
ipoffice2@iitm.ac.in
Research Lab
Prof. Anurag Mittal
Department of Computer Science and Engineering
Intellectual Property
- IITM IDF Ref. 1508
- IN 394760 (Patent Granted)
Technology Readiness Level
TRL- 4
Technology Validated in lab