Indonesian ID Card Recognition using Convolutional Neural Networks

Abstract

Indonesian ID Card can be used to recognize citizen of Indonesia identity in several requirements like for sales and purchasing recording, admission and other transaction processing systems (TPS). Current TPS system used citizen ID Card by entering the data manually that means time consuming, prone to error and not efficient. In this research, we propose a model of citizen id card detection using state-of-theart Deep Learning models: Convolutional Neural Networks (CNN). The result, we can obtain possitive accuracy citizen id card recognition using deep learning. We also compare the result of CNN with traditional computer vision techniques.

Implementation of Verification and Matching E-KTP with Faster R-CNN and ORB

“…In research conducted by Firhan Maulana, extracting the data fields inside the EKTP card reached an F-Score of 0.78 with the timely processing of 4510 milliseconds per ID card [10]. Furthermore, Pratama et al [11], improving the OCR technique with CNN and increased the average F-Score of 0.84. However, various images with good and bad quality become a challenge for the OCR engine in EKTP card data extraction [10][11].…”