A novel approach proposed to deal with challenging task for developing an efficient real time moving object detection by using an improved background subtraction method towards traffic surveillance system. In this proposed mechanism vehicles are detected by using improved background subtraction. This attempt to overcome the limitations in the existing background subtraction method and also improve the detection accuracy by avoiding interrupt from the shadow and occlusion occurrence due to some changes in the weather and intensity of light this problem will be handled accordingly by using the combination of gamma correction with the existing background subtraction method to reduce this impact of illumination. This methodology will be carried out by reading frame by frame RGB video data and convert into gray scale data and then vehicle labeling are carried out with the technique of 2D model. The performance evaluation is carried out regarding, detecting accuracy and processing time together to determine the efficiency.
Volume 11 | 01-Special Issue
Pages: 580-586