A Novel OCR Approach Based on Document Layout Analysis and Text Block Classification
Weiheng Zhu, Yuanfeng Liu, Hao Liang · 2016
Document layout helps users to focus on important content of the documents while neglecting the rest whenever possible. This paper presents a novel Optical Character Recognition (OCR) algorithm whose performance is enhanced by post-processing based on information collected from document layout analysis. Initial OCR results are used for text block classification, whose results are then used to fine-tune the final results. Experimental results show that it outperforms the start-of-the-art OCR algorithms.