Remote Sensing Image Segmentation Using Mean Shift Method

• Published in: Advanced Research on Computer Education, Simulation and Modeling pp. 86 - 90
• Main Authors: Wan, Fang, Deng, Fei
• Format: Book Chapter
• Language: English
• Published: Berlin, Heidelberg Springer Berlin Heidelberg 2011
• Series: Communications in Computer and Information Science
• Subjects: adaptive adjustment; Mean shift; remote sensing; segmentation
• ISBN: 9783642218019; 3642218016
• ISSN: 1865-0929; 1865-0937
• DOI: 10.1007/978-3-642-21802-6_14

Mean shift is a Feature space analysis algorithm widely used in natural scene images and medical image segmentation. It is also used in the high-resolution remote sensing image segmentation process. But one bottleneck of the mean shift procedure is the cost per iteration, especially in the huge data processing. We present an improved mean shift based image segmentation algorithm for the remote sensing images. Given initial parameter of windows, in each iteration step, the algorithm can adaptively adjust window size, which makes the iteration times reduce and speed up the segmentation process. Experiments proved the method can bring good result and satisfying performance.