关键词:标记分水岭分割; 遥感影像分割; Butterworth低通滤波
Improved marked-based watershed algorithm for segmentation of high spatial resolution remote sensing imagery
ZHANG Gui-feng, WU Zhao-cong, YI Li-na
(School of Remote Sensing & Information Engineering, Wuhan University, Wuhan 430079, China)
Abstract:This paper proposed an improved marked-based watershed image segmentation method based on Meyer’s algorithm. The method extracted the markers automatically according to the gradient distribution. In the flooding process,processed the unmarked pixels sequentially with an ascending order according to the gradient value, and used the forward order priority queue and reverse order priority queue to record the current processed pixel. Experiment results show the proposed algorithm performs well with high efficiency and low computer memory cost.
Key words: marked-based watershed; remote sensing imagery segmentation; Butterworth low-pass filtering ......