Document digitalization is done in a large scale for hand written and printed documents, where the documents are scanned and stored in a digital form. Maximum nu mber of these documents are hand written, and hence, they comprise of various errors like angular skewness of words, or sentence as a whole. Before the document can be digitalised using OCR software, it is pre-processed for angular skew detection and their removal from the scanned document. Detection and correction of angular skewness of the words is the most difficult task for the process, which must be carried out as this skewness is present in almost all of the hand written words. In this paper, we propose and test algorith ms independent of the script to detect and correct angular skewness.
展开▼