Vehicle detection based on color analysis

Detecting moving objects in videos is an important task in many computer vision applications, including human interaction, traffic monitoring and Structural Health Monitoring. When having a stationary camera, a basic method to detect the objects of interest is background subtraction. However, precise moving object detection using such a method is an extremely difficult task in a varying environment. This paper introduces a new technique for detecting vehicles in a traffic scene based on frame differencing and color analysis for foreground region (moving region). The foreground region consists of pixels belonging to moving objects, shadows and possibly part of the background in the current frame. The image will be scanned in both horizontal and vertical directions to create threshold lines to eliminate pixels belonging to shadows and the background. The technique is fast and reliable for real-time applications and experimental results show the accuracy of the method.