Binarization of Texts with Varying Lighting Conditions Using Fuzzy Inclusion and Entropy Measures.

There are many applications in which we need to recognize or improve the content of a scanned document by converting it into its binary form. Since global image thresholding techniques fail when we have an input image of poor quality or of non - uniform illumination, we have to turn to some local thresholding technique. In this presentation, we introduce a method of binarizing texts with varying lighting conditions based on specific fuzzy inclusion and entropy measures. These indicators measure specific attributes of the neighborhood of a pixel and then - based on these values - an appropriate threshold is calculated. We don't use any statistical measures or sensitivity parameters depending on the input image. The whole process isn't the fastest one but it's an open, automated and adaptable procedure. [ABSTRACT FROM AUTHOR]