A Novel Framework for Super-Resolution Reconstruction of Equestrian Videos Considering Compressed Sensing Algorithm

An ideal video super-resolution algorithm not only up-converts the resolution of the original image, but also performs corresponding processing on the color gamut, dynamic range, bit depth, etc. Therefore, keeping the prior information available will be essential, and this paper proposes the novel framework for super-resolution reconstruction of the equestrian videos considering compressed sensing algorithm. The wide application of video super-resolution algorithms and the substantial increase in the number of training videos will help to improve and improve the training model, hence, the system firstly conduct the modelling of down-sampling to understand the factors that influence the image quality. Then, the designed algorithm is discussed in detail, the proposed method combines the optimized wavelet analysis and optical flow method to achieve the better reconstruction performance. Through the comparison testing and select the factors of PSNR and SSIM, the accuracy is tested.