Foreground Segmentation in a Video by using a Novel Dynamic Codebook

Publisher:
IEEE
Publication Type:
Conference Proceeding
Citation:
Proceedings of the 2016 13th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON), 2016, pp. 1 - 6 (6)
Issue Date:
2016-01-01
Full metadata record
Files in This Item:
Filename Description Size
07561253.pdfPublished version317.55 kB
Adobe PDF
Foreground segmentation in a video is a key process in many applications such as object detection, an object tracking, and a behavior analysis. Since the extracted foreground objects are often used in the analytical process, the quality of the foreground is a significant factor to the success of these applications. However, there are many key challenges in the foreground segmentation, including dynamic backgrounds, gradual illumination changes, sudden illumination changes, shadows, and long-term scene changes. This paper proposes a novel dynamic codebook method to address such challenges. The dynamic codebook aims to significantly improve the conventional well-known codebook technique by introducing a technique to make a dynamic boundary of each codeword. In this technique, the lab color space is used in order to make the model more resilient to the illumination change. The experimental results and comprehensive comparisons demonstrate that the proposed method can achieve very promising performance.
Please use this identifier to cite or link to this item: