一种遥感图像海上船舶多级自适应聚类分割方法

Multilevel adaptive cluttering segmentation method for marine ship in remote sensing image

采用一种改进型图像信息熵来表征目标区域的复杂程度,并以改进型信息熵和Otsu聚类算法为基础,提出一种遥感图像海上目标多级自适应分割方法。实验结果表明,该方法对于中分辨率遥感图像中相对复杂海域背景下的小目标图像分割有着较好的效果。

Modified image information entropy which could express the complex degree of the target region was used in this paper. Based on the modified image information entropy and Otsu cluttering algorithm, a multilevel adaptive segmentation algorithm for marine target under complex background images was put forward. Experiment results testify the validity of the proposed methods.