Adaptive Contrast Binarization Using Standard Deviation for Degraded Document Images
Jeena Joy, Jincy Kuriakose · 2014
Document image binarization is of great significance in the document image analysis and recognition process because it affects further steps of the recognition process. The variation between the foreground text and the background of different document images is a challenging task. This paper presents a new document image binarization technique that focus on these issues using adaptive image contrast. The adaptive image contrast is a combination of the local image contrast and the local image gradient so as to tolerate the text and background variation caused by different types of document degradations. In the proposed image binarization technique, an adaptive contrast map is first constructed for an input degraded document image which is then adaptively binarized and combined with Sobel’s edge detector to identify the text stroke edge pixels. The document text is further segmented using a local threshold. Keywords:- Binarization, thresholding, adaptive image contrast, stroke width. I.