Artificial Intelligence-Based Super-Resolution Reconstruction Algorithm for Pulsed Multi-Frame Images

Pulsed SRR is a research hotspot that has emerged in recent years, and it has certain practicality, but there are still some limitations in practical applications. Pulsed SRR is a method that can reconstruct images according to different image requirements. The image SRR can be achieved by using the principle of multi-frame imaging. The main objective of this paper is to investigate the pulse multiframe image SRR algorithm based on artificial intelligence. In this paper, a MFI dataset is used to evaluate the effectiveness of the system reconstruction model. Firstly, a method with strong similarity is applied to generate matching images to improve the reconstruction effect and quality. The method is optimized on Data Market Alpha version (ML) and is able to retain image files with higher frame rate under the same conditions. The results obtained after several experimental trainings are a significant improvement compared to the classical algorithm. There are also significant differences in the predicted resolution and contrast obtained by the two different versions of the algorithm on the test dataset. In addition, the algorithm is optimized to improve efficiency and avoid local network performance degradation due to inter-frame interference, since a larger bandwidth needs to be used for transmission. This paper focuses on using this approach to reduce the memory footprint during compression, effectively improving the processing speed performance of the system. In practical applications, this method can be used for image SRR and other applications to improve the resolution.