Deep Residual Network Based Road Detection Algorithm for Remote Sensing Images

As an important recognition target in remote sensing images, roads can be widely used in many fields and has great significance. However, there are still some difficulties in the accurate identification of roads. Firstly, remote sensing images have high-resolution features, which provide more detailed features but also generate more complex background interference; secondly, different spatial resolution leads to different morphology of roads in images, such as different sizes of rural and urban roads, different road materials. These cause great difficulties for the detection of road targets. At present, most of the methods are based on the classification of spectral features, while ignoring other high-dimensional features. In recent years, with the development of artificial intelligence, deep learning has played a unique and outstanding role in many fields, image target detection has become a hotspot field, and many outstanding technologies have emerged. In this paper, a deep residual network-based remote sensing image road detection model is adopted to address the above problems, and the model is finally trained on Massachusetts Roads Dataset to perform road detection. The experimental results show that the model is trained to effectively and accurately identify road targets in the remote sensing images.