基于灰度迭代阈值的高分辨率影像分割研究

图像分割技术为遥感图像解译和分类的一种重要方法，目前主要应用在中分辨率影像中，由于高分辨率影像的信噪比低，直接进行图像分割，效果不佳．对2010年3月湖北省襄阳市高分辨率的快鸟影像，进行降噪处理及改进阈值策略来改善灰度迭代阈值法的图像分割效果．首先运用数学形态学的闭运算对影像进行降噪处理，然后采用灰度迭代阈值法和最大类问方差法的图像分割对影像进行分割对比研究．结果表明，灰度迭代阈值法在时间效率及影像分割效果方面优于最大类间方差法．

Segmentation based on high resolution image of the gray-scale iterative threshold

As a kind of important method in remote sensing image interpretation and classification, image segmentation technology has been mainly used in the moderate resolution image recently. Due to the low signal to noise ratio（SNR） of high resolution image, directly image segmentation is not that good. In this paper, we applied the noise suppression and improved threshold strategy to improve image segmentation effect of the gray iterative threshold method using the Quickbird high resolution image of Xiangyang City in Hubei province in march 2010. At first we used mathematical morphology closed operation for image noise suppression, then carried on a segmentation comparative study of image segmentation effect of the gray iterative threshold method and the maximum variance method. The results showed that the time efficiency and image segmentation effect of the gray iterative threshold method was better than that of the maximum variance method.