A noise attribute thresholding method for document image binarization

Abstract

A new thresholding method, called Noise Attribute Thresholding method (NAT), for document processing is presented. This method has overcome a difficulty encountered by many conventional global thresholding methods which utilize the image gray level histogram. The difficulty is that some objects of interest in the image do not form prominent peaks in the histogram. The NAT method utilizes the noise attribute features extracted from the image. These features are based on a simple noise model and are independent of the strength of the signals (objects) in the image. Experimental results show that this method is very effective.