【摘要】 根据像素比计算芨芨草面积占区域总面积的百分比,由此整体估算呼伦贝尔草原芨芨草的盖度。图像分割采用了最大类间方差法,得到类间最大方差值,即最佳阈值。利用阈值对灰度化图像进行分割得到芨芨草和非芨芨草两类物的二值化图像。采用中值滤波和数学形态学相结合的边缘检测法,提取芨芨草图像,将其与非芨芨草图像分割。基于无人机影像,采用图像分割方法,能够有效提取芨芨草图像,与实测芨芨草盖度相比,估算结果准确度高达97.3%。更多还原

【Abstract】 The proposed image segmentation method mainly includes three stages: the first stage is using Otsu method to get the maximum variance of gray image between clusters,namely the optimal threshold. The second stage is gray image is segmented by the obtained threshold to generate achnatherum splendens and non achnatherum splendens binary image. The third stage is utilizing edge detection method combines mean filtering with mathematical morphological to extract achnatherum splendens image. and divide it with non achnatherum image.Based on unmanned aerial vehicle(UAV) image,use image segmentation method,achnatherum splendens can be extracted effectively.Results of estimation can obtain a higher accuracy of 97.3 % compared with measured achnatherum splendens coverage.更多还原