Application of thresholding in image processing into finding counter of phone card

Abstract

This article presents results of a study on methods of detecting digit areas, going into the technical analysis to select thresholds used in separating digits in the number of digits in the phone scratch card. Based on the theory of image processing, the authors propose the method of choosing the non-fixed threshold using the loop and putting into practice through the Python language, using the OpenCV library. The results are for the purpose of comparing and evaluating the effectiveness of 03 methods: selecting the fixed threshold, selecting the non-fixed threshold, and selecting the non-fixed threshold according to the loop. From there, the author drew conclusions about the optimization of the non-fixed threshold (by loop).