Optimisation of a coastline extraction algorithm for object-oriented matching of multisensor satellite imagery

The authors examine the optimized extraction of coastlines in multisensor satellite imagery as a first step in the object-oriented matching of these images, using whole coastlines as the matching features instead of a point based method. Typically, they want to extract these coastlines with the best positional accuracy possible, together with suppression of unnecessary detail. For this purpose, they add two additional masking steps, an edge focusing and a closing step to the edge detection process.