Residential Areas Recognition by Fusing Region Data and Edge Features

A method for residential areas extraction based on region data and edge images is proposed.The region-based segmentation always provides the closed contour regions and makes use of relatively large neighborhoods for decision making.To obtain the correct segmentation results of residential areas,the algorithm based on the densities and linear distributions of edge point sets is proposed.The experimental results on real aerial images shows that the proposed method fusing region and edge data can improve the accuracy of extraction for residential areas.