Conversion from gray scale or color document image into binary image is the main and important step in most of optical character recognition (OCR) systems and document analysis. Most of the document images after digitization often suffer from poor contrast, noise, uniform lighting, and shadow. Clean up and binarization is an active subject in image processing, which addresses these problems. Most of the previous binarization methods that depend on local thresholding consume more time. The other methods that depend on global threshold are fast but don't work well when the document image is degraded. In this paper we present fast and efficient document image clean up and binarization method based on retinex theory and global threshold. The proposed method is fast and produces high quality results compared to the previous works. © 2013 IEEE.