Road Extraction from Satellite Images Based on K-Mean Clustering
Satellite image processing plays a major role in the areas of urban planning, climate modeling, crop, forest, and
disaster relief management. For example in urban planning, a high-resolution satellite image helps in identifying rivers,
vegetation areas, buildings, roads, various man-made structures, etc. In road detection, a high-resolution satellite images
consists of multiple layers that identify other high-density objects on the road. Road identification is a challenging task for
planning urban areas. The road extraction process consists of two primary steps: detecting roads that may contain non-road
parts such as buildings etc. In this paper, an approach is presented to extract the roads from satellite images. For extraction,
we use K -means clustering, histogram equalization, thresholding to extract roads from the satellite images.The results
presented for the extracted road from the images of urban area.
Keywords - Segmentation, K-means Clustering, Filtering, Histogram Equalization, Contours Detection.