An Examination of Character Recognition on ID card using Template Matching Approach

Abstract

Identification card (ID cards) becomes the main reference in obtaining information of a citizen. Some business sectors require the information contained in the ID card to perform the registration process. In general, the registration process is still using a form to be filled in accordance with the data on the ID card, which will then be converted into digital data by means of retyping the information. The purpose of this research is to create a character recognition system on the ID card where the character recognition process included into four stages: pre-processing, text-area extraction, segmentation and recognition. The experiment includes some tests of greyscale, binarization and segmentation algorithm, as well as the combination of those algorithms. Text area extractor showed satisfactory results of identifying text-area on the ID card, which can scope all the entire area that consist of text. In the segmentation stage, approximately 93% of character can be cut off correctly. The actual character will be mapped to the template character using two algorithms where the division grid of each of them is different. Nevertheless, the recognition process of applying the template method matching still needs to be improved back.