论文部分内容阅读
随着遥感技术的不断发展,高效、快速的遥感影像处理成为人们不断研究的热点。作为内业处理的基础,人们对影像直线特征提取的重视不断加强。本次试验采用辽宁科技大学1∶2000正射影像图,针对教学区内建筑物、绿化区和道路混合的特点,提出了一种自动识别并高效提取建筑物直线特征的算法。该算法将直线拟合和Hough变换进行结合,加入角度先验方法。试验结果表明:该算法的直线提取性能优于传统Hough变换,能很好地适应教学区域,对建筑物的识别能力、直线提取效率及精度有很大的提高。
With the continuous development of remote sensing technology, efficient and rapid remote sensing image processing has become a hot spot that people continue to study. As the foundation of internal processing, people pay more and more attention to the feature extraction of images. In this experiment, 1: 2000 orthophoto map of Liaoning University of Science and Technology was adopted. According to the characteristics of buildings, green areas and roads mixed in the teaching area, an algorithm of automatically identifying and efficiently extracting the linear features of buildings was proposed. The algorithm combines linear fitting with Hough transform and adds angle prior method. The experimental results show that the proposed algorithm has better linear extraction performance than the traditional Hough transform, which can adapt to the teaching area well and greatly improve the recognition ability, linear extraction efficiency and accuracy of the building.