基于形态学梯度重构的高分辨率遥感影像分割方法

基于形态学梯度重构提出一种用于高分辨率遥感影像的分割方法.针对遥感图像的特点构建多形状结构元素,然后使用该结构元素对图像提取形态学梯度并进行开闭重构;根据人的视觉特征,对梯度的高对比度区域进行还原,用于保证较高的局部对比度;最后使用浸没式分水岭变换获得分割结果.对IKONOS影像进行分割实验,结果表明了该方法的有效性.

Segmentation of High-resolution Remote Sensing Image Based on Morphological Gradient Reconstruction

A method of segmentation of high-resolution remote sensing image based on morphological gradient reconstruction is proposed.In this algorithm the multi-shape structure elements are constructed according to the characteristics of remote sensing image.Then morphological gradient image is obtained by utilizing multi-shape structure elements.Opening and closing operations of reconstruction are employed to gradient image.According to the characteristic of human vision,regions with high contrast are recovered,for saving the high local contrast information of the gradient image.At last,we get the final results by immersion simulation watershed transform.Experiments have been implemented to IKONOS images and the results show good performance of this segmentation method in quality.