Image segmentation based on improved SLIC and spectral clustering

In order to solve the problem of single spectral clustering algorithm, which has the disadvantages of low efficiency, low accuracy of boundary and target segmentation, an improved algorithm of SLIC with spectral clustering is proposed. Firstly, the improved SLIC is used to pre-segment the image, and a suitable number of superpixel blocks are generated as the input nodes of spectral clustering. The similarity matrix is constructed by using the distance and color mean information in the superpixel. Finally, the NJW spectral clustering algorithm is used to merge the superpixel blocks to segment the target area. The experimental results show that SLIC algorithm is used to preprocess the segmentation, and spectral clustering is used as the follow-up clustering segmentation method. The experimental results show that compared with the single segmentation algorithm, the algorithm has a significant improvement in time efficiency and the degree of preserving the extracted object boundary.