Research on Regional Water Resources Classification Algorithm Based on Improved Deep Residual Network

In order to realize the rapid and accurate classification of regional water resources, this paper designs the regional water resources classification algorithm based on the improved deep residual network. This paper analyzes the overall situation of regional water resources distribution in China, and takes the remote sensing image of Gaf-2 satellite as the basis of algorithm design, and preprocesses the remote sensing image of regional water resources, such as enhancement, marking and segmentation. On this basis, the pyramid network is used to improve the depth residual network, and the improved depth residual network is combined with the pre-processing results of remote sensing image to obtain the regional water resources classification results. The experimental results show that the proposed algorithm has short preprocessing time, low classification error rate, short classification time and better practical application effect.