PROJECT TITLE :

Block-Based Major Color Method for Foreground Object Detection on Embedded SoC Platforms

ABSTRACT :

Background modeling and foreground object detection are crucial techniques for embedded image surveillance systems. The most well-liked and accurate methods are largely pixel based mostly, taking on additional memory and requiring longer execution times. Thus, these techniques don't seem to be suitable for embedded platforms. This paper presents a block-based mostly major color background modeling and a foreground detection algorithm that possesses economical processing and low memory demand during a complicated scene, making them possible for embedded platforms. Our proposed approach consumes 37p.c less memory and increases accuracy by a minimum of twop.c compared to existing ways. Last, implementing the object detection algorithm on the VIA VB8001 platform, we have a tendency to will achieve 22 frames per second for the benchmark video with image size 768$,times,$576.