高分辨率遥感影像普通车辆自动检测方法探讨

Automatic Vehicle Detection Methods Based on High Resolution Remote Sensing Image

利用高分辨率遥感影像数据,研究停车场上普通车辆的自动提取方法。基于影像的灰度信息,分别采用边缘检测法、阈值分割法、区域生长法和点特征提取法等方法对遥感影像中的车辆目标进行检测。通过对比分析各种算法实验结果,探讨不同算法进行车辆检测的效果,并利用灰度线性变换的方法对阈值分割法进行了改进,提高了普通车辆目标检测效果。

Using high resolution remote sensing image, we researched the automatic detection methods of ordinary vehicle in the parking lot. We adopted edge detection method, threshold segmentation method, region growing method and point feature extraction method to detect the vehicles in the remote sensing image, all based on the gray information of the cars. By comparing and analyzing the results of the experiment, we probed the merits and demerits of various algorithms in this paper. What's more, the effects of the threshold segmentation method were improved observably by using gray linear transformation method.