This paper deals with a robust video based system for the traffic surveillance system on the highway for vehicle detection, vehicle classification for effective traffic analysis using watershed. Image segmentation plays a major role in the extraction of foreground from the video input. It plays an important role in transportation for monitoring vehicles. This paper proposes the technique called a watershed segmentation algorithm for a video in image processing. The overall process based on image segmentation for segmenting an image ROI (region of interest) is the foreground objects. This is used in the application for video supervision, traffic monitoring and SAR (synthetic aperture radar). It plays a vital role in highways to avoid traffic and is mainly used for monitoring the traffic in signals.