作者: Alessandro Foi , Vladimir Katkovnik , Sakari Alenius
DOI:
关键词: Artificial intelligence 、 Quality (physics) 、 Mathematics 、 Block (data storage) 、 Deconvolution 、 Deblurring 、 Image (mathematics) 、 Computer vision 、 Noise (signal processing) 、 Pointwise
摘要: A novel deconvolution technique for blurred observations corrupted by signal-dependent noise is presented. Deblurring performed with a transform-domain inverse-Þltering applied locally, on sliding block of adaptively-selected pointwise varying size. Simulation results demonstrate good quality the proposed method, which versatile and can be easily combined other processing. 1. INTRODUCTION AND MOTIVATION