论文部分内容阅读
基于对机场的图像特征、目标特征的全面分析,提出了一种从大幅面高分辨率光学遥感图像(10000×10000 pixel)中快速识别机场目标的算法。首先通过预处理去除图像中大部分与目标特性不相关的区域,然后在剩余区域中作精确的检测,其中用到的目标特征有空间频率、均值、方差和能量等,使用canny边缘检测、线段分割与四邻域跟踪技术对目标进行精确识别。研究结果表明,本文算法能够实现对机场的快速识别,识别算法时间小于2 s。
Based on the comprehensive analysis of the image features and the target features of the airport, an algorithm to quickly identify the airport targets from large-format and high-resolution optical remote sensing images (10000 × 10000 pixel) is proposed. Firstly, most of the regions of the image that are not related to the target characteristics are removed by preprocessing, and then the remaining regions are used for accurate detection. The target features used are spatial frequency, mean value, variance and energy. Using canny edge detection, Segmentation and four neighborhood tracking technology to accurately identify the target. The results show that the proposed algorithm can quickly identify the airport and the recognition time is less than 2 s.