A Review of Document Image Enhancement Based on Document Degradation Problem

Document image enhancement methods are often used to improve the accuracy and efficiency of automated document analysis recognition tasks such as character recognition. These images could be degraded or damaged for various reasons including aging, fading handwriting, poor lighting conditions, watermarks, etc. In recent years, with improvement computer performance continuous development deep learning, many have been proposed enhance quality these images. this paper, we review six degradation, namely, background texture, page smudging, fading, watermarking, blurring. We summarize main models each degradation problem well work, binarization model that can deal textures, lettering smudges. When facing a stroke connectivity used, while other three problems mostly learning models. discuss current limitations challenges task introduce common public datasets metrics. identify several promising research directions opportunities future research.