Image segmentation, as a fundamental issue of image processing, is the basis of image analysis and image comprehension. There are thousands of image segmentation methods, but none is appropriate for all images. This paper applies several basic image segmentation methods and sparse subspace clustering method to microscopy cell image respectively, and shows the final results. Using threshold segmentation and morphology methods acquires good results, but too much manual intervention is needed. Using the method of sparse subspace clustering can mainly realize image segmentation automati-cally, and the segmentation results depend on the generation of super pixels and how to choose the feature of super pixels. |