An Efficient OCR System for Extracting Information from Indian Identity Documents using CTP

Pages:15-24

Vridhi Sachdev, Chirag Sandil, Abhaysinh Landge, Ashish Kolhe, Bhagyashree Dhakulkar

Abstract

This paper presents an innovative Optical Character Recognition system designed to extract information from three prominent Indian identity documents like Aadhar Card, PAN Card, and GST Certificate. The system's methodology leverages Connectionist Text Proposal Network to efficiently process and extract relevant data, thereby facilitating the Know Your Customer process in Logistics industry. The proposed OCR model aims to streamline and automate the extraction of customer data from these identity documents, enhancing operational efficiency and accuracy in identity verification processes. Through a detailed description of our methodology and experimental results, this research elucidates the effectiveness and reliability of our approach in handling diverse Indian identity documents. The findings underscore the potential of the proposed system to contribute significantly to the improvement of identity verification procedures, particularly in contexts requiring seamless integration of digital technologies for compliance and regulatory purposes.