Road satellite image recognition with improved edge detection

The research project shows that road networks play an important role in a number of geographic applications, such as mapping, infrastructure planning, and traffic routing software. Automatic and semi-automatic road network extraction techniques have significantly increased the extraction rate of road networks. The performance of road detection is severely affected by noise or low frequency. Road edge detection images play an important role in detecting the direction of the road and the specific location of obstacles, size, and speed of obstacles on the road. In this paper, several road detection techniques are theoretically analyzed, and a new road detection method using median filter is proposed. By comparing it with some well-known techniques, it is found that there are many gaps in the techniques presented so far. By comparing the proposed algorithm with the proposed one, it is proved that the proposed algorithm does not introduce artifacts in the old techniques.

کلیدواژه ها:

Road images ، road detection ، satellite remote sensing ، edge detect