作者: Claus Henrik Pedersen , Stephen Bernard Pollard
DOI:
关键词: Supersampling 、 Mathematics 、 960H Technology 、 Artificial intelligence 、 Color histogram 、 Channel (digital image) 、 Pixel connectivity 、 Pixel 、 Computer vision 、 Non-local means 、 Color image
摘要: An image processor receives color signals representing a or black-and-white image, typically containing text and non-text areas. A sliding window swath of the is processed which progressively moves over virtual image. spatial filter applied to sharpen it then classified into regions. The data from regions subjected black enhancement process in signal single channel (here green channel) thresholded against two thresholds, T1, T2. lower (darker) threshold T2 identifies pixels for being set black, whereas T1 “support” used component connectivity analysis. Having defined connected using both pixels, statistics making up are analyzed determine whether should be rendered black. If so, enhanced by snapping halo around white.