Translator Disclaimer
8 September 2011 Improved sum-of-squared-differences tracking algorithm for thermal vision systems
Author Affiliations +
A modification of Sum-of-Squared-Differences algorithm is proposed to improve tracking efficiency of small objects in infra-red image sequences. The reason to use SSD algorithm is its better performance in tracking small objects, than in model based tracking algorithms. However traditional Sum-of-Squared-Differences (SSD) algorithm is sensitive to partial or full occlusions, background clutter and changes in object appearance. To increase immunity to this kind of noises the modification in model update procedure was developed. The experimental results illustrate that the proposed modification to SSD algorithm can improve overall algorithm performance in infrared operation. The paper describes the Sum-of-Squared Differences algorithm and its principal features in tracking objects on thermal image sequences. Next modification to SSD algorithm is described. Finally the experimental results are presented with comparison between traditional and modified SSD algorithm.
© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Grzegorz Bieszczad, Tomasz Sosnowski, and Henryk Madura "Improved sum-of-squared-differences tracking algorithm for thermal vision systems", Proc. SPIE 8193, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Infrared Imaging and Applications, 81932R (8 September 2011);

Back to Top