A Maximum or Minimum Way of Forming 2-D Histogram for Image Optimal Multiple Value Segmentation

To resolve image multiple value segmentation difficulty, two novelty ways were putted forward in the light of the usually process to form 2-D entropy versus the traits of 2-D entropy in this text. In the first way, the 2-D histogram was defined as the central point's grey scale via maximum one of others 4 points in the same 4-neighborhood, In the second way, the 2-D histogram was defined as the central point's grey scale via minimum one of others 4 points in the same 4-neighborhood. Made a comparison with famous approach, test results of the first way on nearly whole of Matlab's images in image-processing toolbox that more details message about related targets in one image can be ideally kept, via test results of second way on introduced above images show that noise can be weakened better without losing any important outline message about related targets.