Remote Sensing Images Of The K-means Clustering And Watershed Segmentation Algorithm

This article'content is image segmentation' study and realization, which are part of method that update GIS vector data automatically by remote sensing images. It mainly study edge detection segmentation algorithm of the remote sensing images, K-Means cluster image segmentation algorithm and watershed segmentation algorithm based on mathematical morphology.The paper also introduce the knowledge of remote sensing images preprocess, mathematical morphology theory and image segmentation. The main content is below:PartⅠIt describes the method and study content of using remote sensing images to update GIS vector data.PartⅡIntroduce base knowledge of remote sensing image's preprocess and primarily study the method of image enhancement and summarize some effective enhancement methods based on remote sensing images' features.PartⅢIt describes the theoretical knowledge about the remote sensing images segmentation and mathematical morphology. Especially use this knowledge in segmentation of remote sensing images.PartⅣThe article's study emphasize on Canny edge detection algorithms, algorithms based on K-Means cluster image segmentation and watershed segmentation algorithms based on mathematical morphology. Then improve these algorithms and gain better effect.PartⅣUse mathematical morphology processed the result of segmentation, after used K-Means cluster segmentation. This method can get single target informations,which could be better to target recognition.PartⅥIt is proved that single segmentation can not perform better, except that mixed these segmentation algorithms. The experiment indicates that Canny edge detection algorithms is efficient used by watershed segmentation algorithms.