Perceiving, understanding and predicting motion is an important part of our daily lives. Motion Estimation and vehicles segmentation is an interesting yet difficult problem in intelligent transportation system. Segmentation of moving objects in a scene is often desired in applications such as video surveillance, video indexing, robotics where our interest is in monitoring for example, cars and people. One of the crucial elements of a traffic monitoring system is the motion analysis component, which segments vehicles from the scene and estimates the motion on the image plane. The objective of the project is to investigate the methods that can be used to extract vehicles from the background in captured video frame and to apply the techniques of motion estimation and segmentation. We have applied Lucas Kanade algorithm to a reference frame and determined the motion vectors of the vehicles. We then extended this by applying segmentation to the reference frame to segment the vehicles from the background and then applied Lucas Kanade algorithm. Lukas Kanade algorithm is selected due to its success in the past video coding standards and its simplicity for software implementations. This processed information will provide a brief information about the vehicle counting, vehicle tracking, vehicle recognition.