Binarization of degraded document images based on hierarchical deep supervised network. (February 2018)

• Record Type: Journal Article
• Authors: Vo, Quang Nhat; Kim, Soo Hyung; Yang, Hyung Jeong; Lee, Gueesang
• Is Part Of: Pattern recognition. Volume 74(2018:Feb.)
• Page Start: 568
• Page End: 586
• Publication Date: 2018-02
• Subjects: Document image binarization -- Convolutional neural network -- Document analysis; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2017.08.025
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We propose a supervised binarization method based on the deep supervised networks. The multi-scale deep supervised network for binarization has not been reported yet. A hierarchical architecture is designed to distinguish text from background noises. Different feature levels are dealt by the multi-scale architecture. The performance results are considerably better than state-of-the-art methods. Abstract: The binarization of degraded document images is a challenging problem in terms of document analysis. Binarization is a classification process in which intra-image pixels are assigned to either of the two following classes: foreground text and background. Most of the algorithms are constructed on low-level features in an unsupervised manner, and the consequent disenabling of full utilization of input-domain knowledge considerably limits distinguishing of background noises from the foreground. In this paper, a novel supervised-binarization method is proposed, in which a hierarchical deep supervised network (DSN) architecture is learned for the prediction of the text pixels at different feature levels. With higher-level features, the network can differentiate text pixels from background noises, whereby severe degradations that occur in document images can be managed. Alternatively, foreground maps that are predicted at lower-level features present a higher visual quality at the boundary area. Compared with those of traditional algorithms, binary images generated by our architecture have cleaner background and better-preserved strokes. The proposed approach achieves state-of-the-art results over widely used DIBCO datasets, revealing the robustness of the presented method.