摘要
Abstract
To ensure that the detection process can adapt to the characteristics and requirements of various complex remote sensing images,improving edge detection accuracy,this paper proposed a water body edge detection method for remote sensing images based on the high-low threshold Canny operator.In the traditional Canny operator,after applying mathematical morphology filtering for image smoothing and denoising,the gradient computation of the remote sensing image was enhanced by adding statistical information on the gradient in the tilted direction.Based on the gradient computation results,the improved Otsu method was used to adaptively select high and low thresholds for filtering water body edge points.The method then performed local edge refinement in the water body,traversed the entire remote sensing image water body,removed false edges,and connected real edge lines to achieve high-quality water body edge detection.Experimental results show that this method can detect edges in different water body remote sensing images,and its noise removal performance is superior to other methods,with minimal distortion.关键词
高低阈值/坎尼(Canny)算子/遥感影像/水域边缘检测/数学形态学滤波/梯度计算/边缘检测/自适应Key words
high-low threshold/Canny operator/remote sensing image/water body edge detection/mathematical morphology filtering/gradient computation/edge detection/adaptivity分类
计算机与自动化
