Abstract:

A novel image segmentation method based on improved sparse subspace clustering is presented. The image to be segmented is overpartitioned into some uniform subregions called superpixels, and color histogram of each superpixel is computed as its feature data. Then by employing an improved sparse subspace representation model, the sparse representation coefficient matrix is computed and used to construct the affinity matrix of a graph. Finally, the spectral clustering algorithm is used to obtain the image segmentation result. Experiments show that the proposed improved sparse subspace clustering method performs well in clustering and is robust to noise. It can obtain good segmentation results for natural color images.