Real-time single frame superresolution

Tarhan, Cem
A demand in real-time applications for superresolution increases as the surveillance and low resolution camera usage is spread for cost optimization. In this thesis two real-time superresolution algorithms have been proposed. The first algorithm (NDUID) is composed of non-dyadic upsampling cascade for smooth interpolation and an edge enhancement block that uses a non-blind deconvolution. Second algorithm (EDAT) starts with Total Variation decomposition. The structure component is interpolated with a fast edge adaptive interpolation and the edges are enhanced using a shock filter. Texture channel only upsampled by bicubic interpolation. The EDAT is also implemented on a Spartan 3A DSP FPGA to demonstrate the results on a real application. Experimental results show that implemented algorithm is capable of upsampling 320x240 resolution images by a factor of two at 52 fps with better quality than other real-time algorithms.
Citation Formats
C. Tarhan, “Real-time single frame superresolution,” M.S. - Master of Science, Middle East Technical University, 2014.