A vehicle segmentation approach by fast mean computation using integral image in intelligent transportation system

In ITS, it is important to segment vehicle for vehicle detection system based on computer vision and the segmentation speed is very important in real time application. A simple vehicle segmentation algorithm is proposed in this paper, firstly, an effective preprocessing method to eliminate the affection of different scale gray pixels in different region, which can make the image be easily binarized; Secondly, a small threshold can be set to binarize the preprocessed image; After that, in order to avoid the small regions enclosed in big region, a inner region filling algorithm is presented; Thirdly, the image processed by first and second step can be easily segmented; Lastly, several transportation scene images are used to test the proposed algorithm, the result indicate the validity of the proposed segmentation algorithm.