Proposed Algorithm to Determine the Skew Angle Using Hough / Radon Transforms

Abstract

    Optical Character Recognition (OCR) is the most important areas seeking researchers to make progress to provide easy transfer of handwritten documents or printed documents by computer, which have only a paper document about it and work to save it in digital form to use it in the transmissions of different text script. In order to increase the efficiency of recognition of these characters go through the processing of symbols scanned many of the steps, and the first step in this area is to find the skew angle, which may cause the development of the page in your scanner are skewed a certain angle which make the process of recognition very difficult. .

Through this research we have proposed the use of a special algorithm to reduce data input page which will be passed on to the Hough / Radon transforms in order to know the capacity of each of these transforms at the expense of skew angle depending on the algorithm proposed. And then we will rotate the original text page  by the skew angle measured to get later on the same document clearly in the correct order to facilitate used later in any dealing interest with this area. We adopted in our present set of handwritten texts to be used as a database for the deviation of the proposed modified algorithm.