Lake edge detection using Canny algorithm and Otsu thresholding

Abstract

Change that is happened to the lake can have implication to its surrounding. Monitoring the change can be done remotely by using remote sensing images. To do it automatically, we first need to detect the lake region in the images, then gain the edges between water and land area. This research aims to implement the Canny edge detection method, combining with Otsu thresholding to detect the edges. Otsu thresholding is used to gain threshold value for Canny Method. In result, some edges are well detected, but some others are missed. Some of the false detected edges are gained from thresholding process, where shadow or dark pixel area which have nearly same color as water, are also detected as water.