Automatic road-side extraction from large scale imagemaps

In recent years, many approaches have been exploited for automatic road extraction. Most of these approaches are based on edge detection algorithms. In this paper, a new object-based approach for automatic extraction of main roads in large scale imagemaps is proposed. The gray-scale imagemap is converted to a simplified imagemap using Gray scale Morphological Algorithms (GMA). At this point, the proposed algorithm consists of two parallel stages. The first stage deals with straight lines extraction and the second stage deals with roads skeleton extraction. In the first stage, the simplified imagemap is segmented and converted to a binary image. Next, the binary imagemap objects are labeled and then, the straight line segments are extracted. In the second stage, the resolution of the simplified imagemap is reduced so that the width of roads are reduced to 2–3 pixels. Next, the reduced resolution image is converted to a binary image. Then, the skeleton of roads in the binary reduced resolution image is extracted. By combining the results from the two stages, the roadsides are extracted. The skeleton of roads and straight line segments are combined using searching road-side algorithm. The test area was an imagemap with a scale of 1:8000 over the Kish region of Iran. The program for this study is developed in Visual C++ language under Windows 98 operating system. © 2002 Elsevier Science B.V. All rights reserved.