This paper aims at developing robust techniques for evaluating image segmentation method based on a variation formulation using edge flow vectors. The key contributions of this work are as follows. First, we investigate the characteristics of existing edge detection method of image segmentation. Second, we develop a complete framework for evaluating a new edge flow vector based edge detection segmentation algorithms. Our system comprises evaluation measures, ground truth data. A number of measures are proposed to quantify edge performances that are illustrated for a number of edge detectors on a number of video sequences that have different properties or contexts. We validate our proposed method and compare with already exiting edge detection approaches. and demonstrate their performance. Experimental results of our proposed method is more accurate and efficient than the other approaches. © 2011 Binary Information Press.