Project Overview

PUPR - OCR (Optical Character Recognition) System is a web-based platform that automates data extraction from various document types, enabling admins to upload documents and quickly retrieve key details—such as name, NIP, email, and salary—without manual entry. The extracted data can be updated directly on the front end to account for OCR inaccuracies. Once approved, the verified document is sent via open API to the PUPR system, streamlining document processing and integrating seamlessly with existing workflows.

Key Features

• Enhanced Image Upload Quality:

  • Utilizes TextCleaner and ImageMagick to pre-process and enhance the quality of uploaded document images.
  • Automatically adjusts brightness, contrast, and sharpness to ensure that the OCR engine receives clear, high-resolution inputs for improved data extraction accuracy.

• Optical Character Recognition (OCR):

  • Employs a specialized OCR process that leverages regular expressions tailored for each document type.
  • Converts raw extracted text into structured data points—such as name, NIP, email, and salary—ensuring that only the relevant and specific information is captured efficiently.

• User-Friendly Data Update Interface:

  • Allows administrators to review and edit the extracted data directly on the front end.

• Document History Tracking:

  • Maintains a complete history of each document’s lifecycle, from initial upload through to extraction, edits, approval, or rejection.

Technologies and Stack

• Frontend:

  • The user interface is built with React.js, a well-known library for creating dynamic, responsive, and interactive web applications.

• Backend – Two Service Architecture:

  • Primary API Service:
    • Developed using Node.js with Express as the HTTP framework, providing the main RESTful APIs to support user interactions and document processing.
    • Utilizes Sequelize as the ORM to manage database operations efficiently with PostgreSQL.
  • OCR and Image Enhancement Service:
    • A dedicated Node.js service focused on scheduled tasks using cron jobs.
    • Responsible for running the OCR process and enhancing uploaded document images (applying tools like TextCleaner and ImageMagick as needed) to improve extraction accuracy.

• Database:

  • The project uses PostgreSQL as its sole database, chosen for its reliability, robust performance, and powerful support for complex relational data.

My Role and Responsibilities

• Update Image Processing Code:

  • Updated the textcleaner module by switching from the 'convert' command to the more efficient ImageMagick library.

• Server Environment Setup:

  • Configured utilities for deploying the application in both staging and production environments on the client’s cloud VPS secured by a VPN.

• Implementation of Regex Validation:

  • Developed and integrated a new regex checking system to support additional document types.

• Open API Configuration for Document Approval:

  • Set up an open API interface that enables seamless integration with the PUPR EHRM System for document approval processes.

Get In Touch

For business inquiries, collaborations, or further discussion about my projects, please feel free to reach out via email at [email protected]. You can also follow my work and stay updated on the latest developments by connecting with me on GitHub, LinkedIn, and Instagram.

---

Stay Curious and Happy Coding !!

---

← Back to projects