Edge Detection and Linking Pattern Analysis Using Markov Chains

There have been many studies on developing more accurate edge detection algorithms and employing them for various applications - especially in geographic information system images and maps, say coastline images. In this paper the coastline images are implemented with an edge detection algorithm using the MATLAB programs, and the edge detected images are analyzed so as to reconnect broken links. In doing so Markov transition probability matrices are generated upon the edge images and processed in order to analyze the edge linking patterns of the coastline images. Furthermore, the image analysis using the Markov Chains shall be extended and acknowledged as another alternative that can be a replacement or supportive to traditional coastline mapping techniques, because the Markov Chain represents the pattern of the coastline images numerically.