Multi-Scale Fusion And Splice For Moving Images During Space Interaction And Docking

In this thesis we studied the image fusion and splice algorithms in the process of two spacecraft closing to each other during interaction and docking. Then we analyzed and studied the existing popular image fusion algorithms, and proposed an improved fusion algorithm based on SCDPT as well as a hierarchical algorithm for far and near distance fusion. Meanwhile, a tool package for image fusion and splice in network environment was implemented. Research and innovation are as follows:(1) We proposed a multi-scale fusion rule. Then we used the rule in the SCDPT algorithm and it was validated by experiments. We proposed a multi-scale image fusion algorithm based on SCDPT and gave the implementation steps. Then we used standard image sets and space images to make some experiments by using subjective and objective evaluation criteria. The experiments showed that this method was good for image fusion both in spatial pictures and standard images.(2) We proposed a hierarchical fusion strategy. We detected and extracted the moving region by using moving target detection and extraction method, then took different image fusion methods for the moving region and the non-moving region. In the moving target region we considered fusion effect first, for the non-moving target region we focused on fusion rate. We made some experiments on both standard image set and spatial images with proposed algorithm and wavelet, UDCT etc. The experiments results showed that the improved algorithm was verified.(3) We proposed an improved SIFT space moving image stitching algorithm. We introduced the process of SIFT image stitching algorithm in detail, and then we improved the algorithm. By taking the moving target segmentation into consideration, this method reduced the complexity of the feature point calculation. Finally we used ordinary image and space image to make some experiments, and we proved that the algorithm can get effective results.(4) We made a space moving image fusion and splicing tool. The tool acquired image information from the space simulation network and remote server. After selecting a different image fusion algorithm or stitches algorithm for image fusion or splicing operation, the results will be displayed at the screen and stored at the same time. We used objective indicators to assess the image fusion result.