Super-resolution reconstruction of sequence image implemented on OMAP3530 platform

In remote sensing, surveillance, reconnaissance, medical diagnosis (CT, MRI) and other applications, high resolution (HR) images are required. However, due to the high cost and physical limitations of the acquisition hardware, the low resolution (LR) images are used frequently. Therefore, super resolution (SR) restoration is an emerged solution permitting to form one or a set of HR images from a sequence of LR images. In this paper, a novel sequence image SR restoration algorithm based on wavelet decomposition and frame difference was proposed and has been implemented on OMAP3530 platform, demonstrating the effectiveness of image reconstruction. Quantitative and qualitative experiments for the introduced method showed great superiority in obtaining high resolution images with high PSNR and much better visual quality.