Multi-view information attention interaction network for light field super-resolution

The invention discloses a multi-view information attention interaction network for light field super-resolution, which is characterized in that a group of low-resolution light field images are used as the input of the network, and a top branch comprises a remodeling process, a feature extraction module and four double-attention modules; the bottom branch comprises a feature extraction module and four multi-view attention modules; the feature information extracted by the two branches is converged to the fusion reconstruction module together, firstly, the global image feature is utilized to update each single-view-angle image feature, so that each view angle feature can be fused with the global view angle feature, then the global view angle feature passes through the up-sampling module, and finally, the global view angle feature is added with the bicubic up-sampling image of the input image to obtain the final output image. According to the invention, the multi-view attention module is utilized to make the weig