Segmentation of Text from Degraded Document Images by Local Threshold Method

Text segmentation from badly degraded document images is a very interesting task due to the high inter/intravariation of different document images. In this paper, a document image binarization technique that reports these problems by adaptive image contrast. An adaptive contrast map is first constructed for an input degraded document image by combination of the local image contrast and local image gradient. The contrast map is then binarized and combined with Cannys edge map to find the text stroke edge pixels. The text is further segmented by a local threshold method. Some post-processing is further applied to increase the document binarization quality. The proposed method is simple and involves minimum parameter tuning. To improve the quality of the text in the degraded document image using two thresholding techniques. One is OTSU with several edge detection (i. e. canny, sobel, and total variation) techniques applied to the degraded document image. Another is Adaptive threshold with several edge detection (i. e. canny, sobel, and total variation) techniques applied to the degraded document image. The qualities of these output images evaluated by PSNR and MSE. The best combination of threshold and edge detection techniques is selected by testing several degraded documents.