Image Segmentation Algorithm of Variable Precision Based on Granular Matrix Model

Abstract

In order to deal with space correlation of image information, a new image segmentation algorithm based on granular matrix model is proposed. Firstly, we define granular matrix model, which is used to build knowledge granular of an image. Secondly, we introduce classifying error precision to the model, and construct granular layers of an image by it. Thirdly, for the need of segmentation precision, we choose unit granular layer and realize reduction by granular matrix. Finally, implementing the combination of the similar regions and the image segmentation is accomplished. In order to certify the new algorithm, it is applied to image segmentation tests. The results indicate that it is more suitable for actual need, which not only reduce complexity of space and time, but also provide new thoughts in image process. Keywords-image segmentation; granular computing; variable precision