Improving Degraded Document Images Using Binarization Technique

Abstract Image segmentation is a set of segments that collectively cover the entire image or a set of contours extracted from the image. In the process of improving degraded document images segmentation is one of the difficult task due to background and foreground variation. This paper presents a new approach for enhancement of degraded documents. It consists of an adaptive image contrast based document image binarization technique that is tolerant to different type of document degradation such as uneven illumination document smear involving smudging of text seeping of ink to the other side of page degradation of paper ink due to aging etc. The images i.e. scanned copies of these degraded documents are provided as an input to the system. They are processed to get the finest improved document so that the contents are visible readable. Contrast image construction can be constructed using local image gradient and local image contrast. Further edge estimation algorithm is used to identify the text stroke edge pixels .The text within the document is further segmented by a thresholding technique which is based on the height and width of letter size present in degraded document image. It works for different format of degraded document images. The method has been tested on Document Image Binarization Contest DIBCO experiments on Bickley diary dataset consists of several challenging degraded document images.