A modern spatial enhancing method decreed on a robust MSRR and high-frequency synthesized SSRR for ultra-expansion ratio

Abstract

From the great requisitions of the modern digital image implementations, the refined spatial image is normally desired nevertheless the observed device for high resolution is ultimately overpriced thereby the SRR (Super Resolution Reconstruction) technique, which is founded on the arithmetical establishment for creating the better refined spatial image, is one of the modern scrutinized topics in DIP (Digital Image Processsing) and DSP (Digital Signal Processing) communities. In consideration of ultraexpansion ratio (at 16x), this research article aims to present a modern spatial enhancing method decreed on a robust MSRR (Multiframe Super Resolution), which is constituted on the ML (Maximum Likelihood) regularization with robust Andrew's Sine function, and a SSRR (Single Frame Super Resolution), which is constituted on high-band spectrum appraisement. Initially, for concealing outlier whereas preserving the region border, a sequence of captured images with poor quality is computed by a robust MSRR (Multi-Frame Super Resolution) for creating the better refined spatial image at 4x expansion ratio. Next, this enlarged image is computed by high-band spectrum appraisement SSRR (Single-Frame Super Resolution) for ultimately creating the refined spatial image at 16x expansion ratio. The comprehensive verification outcomes embellish that the modern spatial enhancing method can create the 16x spatial image with better refined information and less outlier.