基于稀疏表示的多幅图像快速超分辨率重建

摘要

Aiming at problems that image super-resolution reconstruction(SRR)based on sparse representation can improve image reconstruction quality,but generally there are large amount of computation and time-consuming,present a multiple images fast super-resolution reconstruction based on sparse representation. The sparse representation of the image is obtained by sparse representation algorithm of particle swarm optimization (PSO),and sparse coefficients of several images are fused.High resolution image is obtained by reconstruction according to sparse coefficient after fusion.Experimental results show that reconstruction speed of this method is faster and quality of reconstruction is higher.%针对基于稀疏表示的图像超分辨率重建(SRR)提高图像的重建质量,但一般存在计算量大、耗时长的问题,通过粒子群优化稀疏表示算法获得稀疏表示;对多幅图像的稀疏系数进行融合;根据融合后的稀疏系数重建得到高分辨率图像.实验结果表明:方法的重建速度更快,重建质量更高.