Projection onto convex sets super-resolution image reconstruction based on wavelet bi-cubic interpolation

The traditional projections onto convex sets super-resolution image reconstruction algorithm leads to the halo effect in reconstructed high resolution image, so we present a super-resolution reconstruction method based on wavelet bicubic interpolation algorithm. The new algorithm utilizes wavelet bi-cubic interpolation algorithm to obtain the initial image of POCS algorithm and introduces bilateral filter to estimate point spread function. Bilateral filtering makes non-edge pixel having minimum influence on the edge region. Algorithm also uses adaptive relaxation parameters to reduce the impact of error motion information. Compared with traditional algorithm, experiment results show that the proposed reconstruction algorithm eliminates the halo effect obviously, thus the reconstructed image can achieve a good visual effect.