结合相位一致和分水岭变换的高分辨率遥感影像分割方法

发布时间：2018-02-12 12:28

本文关键词： 遥感高分辨率遥感影像相位一致性分水岭变换影像分割　出处：《激光与光电子学进展》2017年09期 　论文类型：期刊论文

【摘要】：针对噪声导致高分辨遥感影像分割存在过分割或者欠分割的问题,提出结合相位一致和分水岭变换的高分辨率影像分割方法。该方法首先采用基于光谱相似性的相位一致的模型方法来获得边缘响应幅度,再采用自动标记分水岭算法对影像进行分割;基于相邻分割对象的空间位置、形状、面积等特征多重约束,提出相邻分割对象合并代价函数模型,对分割结果进行优化并获取最终分割结果。选择典型地区实验影像进行分割实验,通过目视评价和监督评价,并与典型分割方法进行比较,验证所提分割方法的有效性。
[Abstract]:In view of the problem of over-segmentation or under-segmentation of high-resolution remote sensing image caused by noise, A high resolution image segmentation method based on phase agreement and watershed transform is proposed. Firstly, a phase consistent model based on spectral similarity is used to obtain the amplitude of edge response. Then the image is segmented by automatic marking watershed algorithm, and based on the spatial position, shape and area of adjacent segmentation object, a cost function model of merging adjacent segmentation object is proposed. The segmentation results are optimized and the final segmentation results are obtained. The experimental images of typical regions are selected for the segmentation experiment. The effectiveness of the proposed segmentation method is verified by visual evaluation and supervised evaluation, and compared with the typical segmentation methods.
【作者单位】：辽宁工程技术大学测绘与地理科学学院;中国测绘科学研究院;
【基金】：国家自然科学基金(41101452) 高等学校博士学科点专项科研基金(20112121120003) 辽宁省教育厅科研项目(LJYL010)
【分类号】：TP751

【相似文献】