基于细化分割的遥感影像水体边缘轮廓提取

Extraction of edge profiles of remote sensing image based on detailed segmentation

为实现遥感影像的水资源特征识别,需要对遥感影像中的地表水体边缘信息进行有效检测识别,提出了一种基于细化分割的遥感影像水体边缘轮廓提取识别方法。采用卫星遥感技术进行高分辨率遥感水陆场景图像成像,通过灰度像素增强技术进行遥感影像空间分辨率增强处理,在不同场景纹理中进行遥感影像空间像素特征重构,以中心像素的灰度值为阈值建立遥感影像陆地地物的空间结构特征辨识模型,采用细化分割方法进行遥感影像的水体边界点分割处理,采用形态学滤波方法进行遥感影像水体边缘轮廓检测过程中的细化分割和滤波,对水陆粗分离结果进行形态学闭运算处理,根据细化分割结果进行水体边界平滑处理,实现对遥感影像水体边缘轮廓的提取。仿真结果表明:采用该方法进行遥感影像水体边缘轮廓提取的精度较高,水体边界平滑性较好,轮廓特征的辨识度较高。

In order to realize the feature recognition of water resources affected by remote sensing,it is necessary to effectively detect and identify the surface edge information of the remote sensing image,the method is proposed of extracting and recognizing the edge contour of the remote sensing image based on the fine segmentation of the remote sensing image.The satellite remote sensing technology is used to image the remote sensing image of the surface water body,the spatial resolution of the remote sensing image is enhanced by the gray pixel enhancement technology,the spatial pixel feature of the remote sensing image is reconstructed in the different scene texture,the spatial structure feature identification model of the remote sensing image land object is established based on the gray value of the center image,the method of thinning segmentation of the water body boundary point segmentation of the remote sensing image is used,and the method of filtering is used to perform the fine segmentation and filtering of the edge contour of the remote sensing image in the process of the remote sensing image,extracting the rough image of the surface image.The simulation results show that the method has high precision,good smoothness of water boundary and strong recognition of contour features.