GVF algorithm has been studied actively and it will be able to apply to quite many applications. This paper presents GVF algorithm for the segmentation of a sequence of images. Traditional GVF algorithm can extract contour of object in an image. However, in the video sequence images or CT images traditional GVF algorithm had some problems. We can treat these problems under motion tracking using Motion Estimation using Mean Square Error (MSE). Initial point problem is very important in GVF algorithm. We applied extracted initial point of first image to the second image. In other word, previous initial contour can be used to the next initial contour.