Geometric image change detection in urban environment

A proficient technique to distinguish changes in the geometry of an urban environment is proposed in this paper. Generally, many of the change detection techniques involve a pixel to pixel comparison by using an algebraic or a transform method of change detection. This technique profoundly depends on the optimum choice of threshold value to separate the real altered pixels. Moreover all these techniques are capable of calculating only the two dimensionality change in the environment, whereas in the proposed technique a differential geometry approach is used to detect changes from images which are done by involving the geometric property of the pixels with respect to its environment. Finally the quality of image is measured using various performance parameters like PSNR and MSE.