论文部分内容阅读
光学遥感图像海陆边界分割是海洋近岸目标检测和识别的重要技术基础。由于光学遥感图像分辨率的提高,海陆边界分割存在背景复杂、干扰多等问题。为解决复杂背景下高清遥感图像海陆边界分割问题,提出了一种新的海陆边界分割算法。该方法包含三个部分:使用均值漂移算法将遥感图像处理成若干同质区域;采用一种新的基于扫描线的方法选择海水区域种子像素点,利用区域增长算法填充海水区域;通过连通区域分析的方法分离陆地部分,得到海陆分割结果。实验证明,该方法能够对于复杂背景下的光学遥感图像实现准确、稳定的海陆分割,算法具有较强的鲁棒性和准确性。
Optical remote sensing image segmentation of land and sea is an important technical basis for the detection and identification of marine nearshore targets. Due to the increase of the resolution of optical remote sensing images, there are many problems such as complex background and interference in the land-sea boundary segmentation. In order to solve the problem of sea-land boundary segmentation in complex remote sensing images, a new algorithm for sea-land boundary segmentation is proposed. The method consists of three parts: the mean shift algorithm is used to process the remote sensing image into several homogeneous regions; a new method based on the scan line is used to select the seed pixels in the seawater region, the sea area is filled by the region growth algorithm; The method of land separation, the result of land-sea segmentation. Experimental results show that this method can achieve accurate and stable sea-land segmentation for optical remote sensing images with complex background, and the algorithm has strong robustness and accuracy.