论文部分内容阅读
传统遥感图像云检测方法在处理山地、雪地、暗云等场景时极易发生错判,准确度较低。通过对遥感图像中云与地物的不同特点进行分析,提出一种新的遥感图像边缘特征描述方法,结合图像的边缘特征和灰度特征使用AdaBoost分类器进行云图分类,并利用图像的空间相关性对分类结果进行修正。经10万余幅图像测试结果表明:该算法与传统算法相比准确度极大提高,正确率达到96%以上,且运算速度快,满足实时性要求。
The traditional remote sensing image cloud detection method is very easy to make wrong judgment when dealing with scenes of mountain, snow and dark clouds, and the accuracy is low. By analyzing the different characteristics of clouds and ground objects in remote sensing images, a new method for describing the edge features of remote sensing images is proposed. The edge features and gray features of the images are used to classify the clouds using the AdaBoost classifier, and the spatial correlation of the images Sexually correct classification results. The test results of more than 100,000 images show that the accuracy of the algorithm is greatly improved compared with the traditional algorithm, the correct rate is over 96%, and the computing speed is fast, meeting the real-time requirements.