Image Segmentation Based on D-S Evidence Theory and C-means Clustering

On the basis of the Dempster-Shafer evidence theory, this paper given the multi-source information fusion method based on Dempster-Shafer evidence theory, and applied information fusion technology of Dempster-Shafer evidence theory in the classification of bamboo image texture. The image data for the DS classification, the user need to train samples, and proposed an D-S method of automatically obtaining the training sample in accordance with image feature and C-means clustering algorithm. First, the image is divided into several regions, each region containing images using wavelet decomposition to remove the edge of the area, and then calculate the remaining energy of the mean smooth area as the feature value, use the C-means clustering algorithm to smooth the regional classification, the feature value and type of training samples labeled as DS, and finally training of the classifier to segment the image. Experimental results show that the proposed method has achieved good segmentation results.