DPCM edge detection using local histogram

An approach is proposed for selecting the thresholds of edge strength maps from their local histograms. This threshold selection technique is based on finding the threshold for small blocks of the edge map. For each block the threshold is chosen using an iterative procedure. Also, an approach to edge detection based on DPCM image processing is presented. A DPCM coder is used as a method of computing the differential image rather than the conventional edge detection operators. The merits of using the DPCM coder to compute the differential image is that the computations required by the iterative threshold selection technique is reduced considerably.<
>