Efficient Joint Source Channel Coding and Decoding Method for Remote Sensing Image Transmission

Remote sensing images play a crucial and indispensable role in fields such as environmental monitoring and precision agriculture. With the advancement of remote sensing acquisition technology, the generation rate of massive remote sensing data has experienced an explosive growth. However, the communication bandwidth rate of the current satellite-to-ground channel is far from meeting the requirements for the backhaul of remote sensing data. Inspired by semantic communication technology, this paper proposes an efficient joint source channel coding and decoding method for remote sensing image transmission. This method only requires the transmission of key semantic features in remote sensing images while discarding the redundant features, which significantly improves the transmission efficiency of remote sensing images. The proposed transmission method includes a feature extraction module, a key semantic feature selection module, a joint source - channel encode module, a joint source - channel decode module, and a task module. The experimental results demonstrate that the classification accuracy of our method is significantly better than that of other comparison methods on the NWPU-RESISC45 data set.