A binarization method for low-quality document images based on background estimation and energy minimization

Abstract

The invention discloses an inferior-quality document image binaryzation method based on background estimation and energy minimization, and the method comprises the steps: firstly carrying out the gray scale preprocessing of a color document image; secondly carrying out the noise reduction, background image estimation, background deduction and image enhancement of the image through bilateral filtering; thirdly constructing an energy function, and constructing a network map; finally achieving the minimization of the energy function through employing an augmenting path graph cut algorithm. The method remarkably improves the document image binaryzation effect under the complex background, can be suitable for writing in various colors, stroke gradual changes, ink mark wetting, page stain or texture, illumination non-uniformity, and low contrast.

Description

technical field [0001] The invention belongs to the technical fields of digital image processing, pattern recognition and machine learning, and in particular relates to a low-quality document image binarization method based on background estimation and energy minimization. Background technique [0002] Document Analysis and Recognition (DAR) technology has been widely used in the digitization of ancient books, layout analysis and text recognition, video subtitle extraction, text information retrieval and other fields, mainly including image acquisition, binarization, skew correction, character segmentation and recognition and other processes . Image binarization is one of the key preprocessing links, which converts grayscale images into binary images, so as to realize the separation of character foreground and document background. The effect of the binarization algorithm directly affects the performance of the entire DAR system, so many scholars have studied it in recent ye...

AI Technical Summary

Problems solved by technology

The effect of the binarization algorithm directly affects the performance of the entire DAR system, so many scholars have studied it in recent years and proposed many algorithms; however, it is affected by factors such as poor image contrast, ink infiltration, page stains, or uneven illumination. , binarizing low-quality document images remains a challenge

However, none of the above methods can well solve image binarization under complex document backgrounds such as low contrast, ink soaking, gradient lighting, smudges and textures.

Images