摘要
提出了一种在卫星图像中采用域值分割的方法来识别港口目标。运用基于灰度级的权值矩阵来代替通常所用的基于图像像素的权值矩阵来描述图像各像素的关系,算法所需的存储空间及实现的复杂性比其他基于图论的图像分割方法大大减少,从而有利于应用在自动目标识别和其他方法中的应用。实验结果表明,其优越的分割性能对港口目标和海域以及陆地的区分起到非常关键的作用,从而能更好地识别港口目标。
Present a novel segmentation shareholding algorithm method to recognize port targets in satellite image. The weight matrices used in evaluating the graph cuts are based on the gray levels of an image, instead of the commonly used based on image pixels. The number of gray levels is much smaller than the number of pixels. Therefore, the proposed algorithm occupies much smaller storage space and requires much lower computational costs and implementation complexity than other graph based image segmentation algorithms. This fact makes the proposed algorithm attractive in automatic target recognition and various means. The experiment result indicated that the ascendant segmentation performance appear a very pivotal function to distinguish port target with sea area and the land, thereby we can recognize port target effectively.
出处
《装备制造技术》
2009年第11期22-23,共2页
Equipment Manufacturing Technology
关键词
图谱理论
域值分割
港口识别
卫星拍摄
graph special theory
threshold division
port recognition
