Algorithmic research based on image segmentation

Abstract

Image segmentation is means we divide the image into several areas based on the characteristics of gray level, color, space texture, and geometry pattern. These areas are mutually exclusive. In one area, some image characteristics might consistent or similar. However, some characteristics in the different areas expresses the obviously difference. Image segmentation is one of the most the important research for the computer visual, management and evaluation of image. It is the basic primary for the model identification and image understanding. The segmentation quality will directly influence the image management effect with important research value. There have various segmentation methods with different thinking. However, there still has no segmentation algorithm that suit for the entire image. The most algorithms are provided aiming at the detailed questions. Therefore, people are searching for the new segmentation algorithm with more potential for the better effect. This article evaluates the usual segmentation algorithm and concludes the applicative area for the commonly used algorithms © 2013 Springer Science+Business Media.