Extraction of unstructured roads from satellite images using binary image segmentation

Abstract

In this paper, a new methodology is proposed for extraction of unstructured road networks from satellite images using binary image segmentation. The segmentation method is implemented by determining a global threshold which is obtained from a statistical evaluation using a number of sample satellite images. The thresholding method is based on the observation that the roads in unstructured environments are brighter than non-road areas. In addition, it is observed that in an aerial image, the area covered by roads is only a small percentage. Hence, road pixels are generally populated at the brighter end of the image histogram. After thresholding, morphological operations are applied to the resultant binary image. The results showed that proposed method can be used for automatic extraction of unstructured road network from a variety of satellite images.