Text recognition on PAN card using template matching method

Abstract

Object recognition in computer vision refers to detecting a particular object in a video or an image. Computer vision being a branch of Machine Learning helps us to achieve a goal by making the machine automate the entire process through multiple algorithms and paradigms. PAN (Permanent Account Number) is a 10 character alpha-numeric code which is issued by Government of India to each of its citizen, especially those who are tax payee. The purpose of this research is to identify the PAN code and extract them so as to automate the process of retyping it manually. We achieve the above through text recognition process of template matching under Normalized Cross Correlation (NCC) method.