High-Resolution SAR Image Road Extraction Based on MRF Model

As an important ground object, remote sensing image road extraction is a very popular research subject and widely used. With the continuous development of SAR imaging technology, its spatial resolution has become higher and higher, and the road in the image is no longer displayed as a linear feature in low resolution, but a narrow and long planar feature, which leads to the previous low resolution road extraction method cannot be effectively used in high resolution, and a new high resolution road extraction method is needed. In this paper, a high-resolution SAR image road extraction method based on MRF model is proposed. The Contourlet threshold filter is used to reduce the impact of speckle noise, image segmentation is performed using MRF model, and road edge information is extracted based on Canny operator. Finally, geometric feature constraints and edge connections are performed. The results verify the accuracy and stability of the road extraction method proposed in this paper.