论文部分内容阅读
本研究采用IKONOS遥感图像,选取波谱特征区,通过区域像元统计并计算各波段权重系数,将影像进行波谱归一化处理,使归一化的影像岛陆与海水特征更明显。在此基础上,采用最大类间方差法确定最佳分割阈值,对归一化的影像进行二值化,从而提取海岛岸线。该方法采用特征曲线法进行水陆分离,得到的二值图像保持了原图的有效边缘,采用二值形态学提取的海岸线连续可靠、信噪比高。结果表明,该方法简单、快速,能有效提取海岸线,具有实用价值。
In this study, the IKONOS remote sensing image was used to select the spectral characteristic area. By using the pixel statistics of the area and calculating the weight coefficients of each band, the image was normalized by the spectrum to make the normalized image island land and seawater characteristics more obvious. Based on this, the best segmentation threshold was determined by the method of maximum variance between classes, and the normalized image was binarized to extract the island coastline. The method uses the characteristic curve method for water and land separation, and the obtained binary image maintains the valid edge of the original image. The coastline extracted by the binary morphology is continuous and reliable, and the signal-to-noise ratio is high. The results show that the method is simple and fast, which can effectively extract the coastline and has practical value.