卫星遥感图像中类圆形油库的自动识别方法

油库是典型军事目标之一，对其识别是卫星图像判读的一项重要内容，传统的方法是通过判读员进行人工判读，工作量非常大是其缺点之一。为了克服这一缺点，本文提出了一种类圆形油库的自动识别方法。首先利用Kapur熵法对图像进行阈值分割，得到二值图像：然后对二值图像中的白像素进行最近邻聚类形成团块，并计算其面积以及体态比和矩形度等形状参数；最后利用油库近似圆形和成群分布的特点对油库群进行识别和定位。实验结果表明该方法对于高分辨率卫星遥感图像中的类圆形油库的识别是很有效的。

Automatic recognition method for quasi-circular oil depots in satellite remote sensing images

Oil depot is one type of typical martial targets and recognizing it is an important work for satellite image interpretation, which is typically implemented by manual operation of interpreters. To avoid hard work, an automatic recognition method for quasi-circular oil depots is presented. Firstly, the entropy method introduced by Kapur is used to get the optimum threshold, according to which, a binary image is obtained; then the white pixels in the binary image are clustered with the rule of nearest neighbor to generate some blobs whose area, posture ratio and rectangle tolerance are calculated at the same time. In the end, oil depots are recognized and their clusters are located.by the knowledge that the shape of an oil depot is similar to a circle and some nearer oil depots are located. Results on the satellite images with high resolution are given to demonstrate the efficiency of our method.