Space SRCNN: lightweight spatial target super-resolution reconstruction algorithm

With the rapid development of space technology, the number of spacecraft in orbit has increased significantly, which puts higher requirements on the safe working environment of space facilities such as space stations. Computer vision-based space target observation research has become increasingly important. The signaling capability of space communications is severely limited due to the high cost of spacecraft manufacturing and maintenance. Therefore, the development of observation methods that occupy little storage space and have a large amount of information has become a hot research topic. In this study, a lightweight space target super-resolution reconstruction algorithm called Space SRCNN is proposed. The algorithm consists of only four layers of convolutional neural network, and the number of parameters is only 0.20 M. Its peak signal-to-noise ratio (PSNR) can reach 39.09 when reconstructing a spacecraft image. Space SRCNN can significantly reduce the cost of interstellar information transmission, provide a novel technological means for the accurate identification and monitoring of space targets, and provide a low-cost space target observation with a practical solution for low-cost space target observation.