Moving Vehicle Detection

Abstract

This paper describe to detect an moving vehicle from an input sequence of video where an input video is taken as input and it is divided into a fix size frame where number of frames are  fixed  and  initial  works  on  single  frame  where  it outputs the number  of  frame  and  captures  that  particular frame and vehicle detected using Gaussian mixture model is capture and further processing as grey scale image is saved whereas then preprocessing is further done to remove any noise and unwanted disturbance as we focused on foreground image it captures only foreground image then on single frame only we apply the bbox method to provide the boundary to the detected vehicle which ensures the area of interest. We then applying preprocessing we get a clean foreground image with rectangular or square boundary as specified which allows to be focused on detected image and on single frame only we applied a counting method of detected image that results out a  counting of vehicle on a frame which then can be calculated for number of frames and also can specify the vehicle classification.