Border noise removal and clean up based on retinex theory

Conversion from gray scale or color document image into binary image is the main step in most of Optical Character Recognition (OCR) systems and document analysis. After digitization, document images often suffer from poor contrast, noise, uniform lighting, and shadow. Also when a page of book is digitized using a scanner or a camera, a border noise, which is an unwanted text coming from the adjacent page, may appear. In this paper we present a simple and efficient document image clean up by border noise removal and enhancement based on retinex theory and global threshold. The proposed method produces high quality results compared to the previous works. © Springer Science+Business Media Singapore 2014.