Image Super-resolution with Multiple Regularized Terms
ZHU Qidan;SUN Lei;CAI Chengtao;College of Automation, Harbin Engineering University;
In order to suppress the ringing and jaggy artifacts during the super-resolution image reconstruction process, an image super-resolution algorithm with multiple regularized terms is proposed. Firstly, the image degradation model is given and the image reconstruction constraint item is analytically derived. The high-resolution image can be generated by using the reconstruction constraint item, which will have jaggy and ringing artifacts. In order to solve this problem, the autoregression model and filters prior are invented to regularize the reconstruction process. The autoregression model is used to restore the local image details and the adaptive parameters of the autoregression model can be generated through the natural cluster sets. Meanwhile, the filters prior are used to force the edges of high-resolution image to be sharp. Finally, the experimental results show that our algorithm outperforms other competing algorithms in terms of both quantity and quality.