Along with the times, the world of banking is also required to have growth in providing services to the public. The world of banking itself has a significant contribution in daily life. However, some deficiencies that arise due to the application of the procedures used are still often encountered. This situation can be seen from the number of customer self-registration data that is still done manually. Optical Character Recognition (OCR) technology on Citizenship Cards can be used to get the results of accuracy and speed and get the best reading results. The purpose of this study was to use Optical Character Recognition (OCR) technology in reading the Identity Card (KTP). The readings from OCR can be used to compare the size of the original, medium, and small images in color and grayscale images, so that the best results can be found in the processing of personal data on Identity Cards with Optical Character Recognition (OCR). This study resulted in the accuracy of reading grayscale Identity Card (KTP) data more accurately. Where the amount of 86.32% for the accuracy of the colored Identity Card (KTP) and 88.58% for the accuracy of the grayscale Identity Card (KTP).