DEVELOPMENT OF A SYSTEM FOR UNDERSTANDING BPKB FORMAL DOCUMENTS AND HANDWRITTEN TEXT USING OPTICAL CHARACTER RECOGNITION (OCR) TECHNIQUES

• Main Author: Ananda Pratama Resyaly, Daffa
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/75327
• Institution: Institut Teknologi Bandung

This final project aims to develop a system for understanding formal documents, namely BPKB, and handwritten text using optical character recognition (OCR) techniques. This system requires an internet connection in the document recognition phase. The process in this system involves three main stages, namely text detection, text recognition, and named entity recognition. In the text detection phase, the system uses a text detection algorithm to identify areas in the formal document that potentially contain text. Then, in the text recognition phase, an OCR algorithm is applied to recognize text in each area detected previously. Subsequently, in the named entity recognition phase, the system applies a natural language processing (NLP) algorithm to recognize certain entities, such as the owner's name in the formal document. Testing was carried out using various formal documents with image quality variations, and the test results showed a reasonable level of accuracy and inference time. With the existence of this system, it is hoped that the process of understanding and extracting information from formal documents will be more efficient and accurate. This system can be used in the automation and digitization process of formal documents, improving information accessibility, and reducing human errors in data processing.