Develoment of an automated pill medicine defect system using camera and web application

Anang Syafrizal Anwar, Muhammad Denial (2024) Develoment of an automated pill medicine defect system using camera and web application. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Preview

Text (Full Text) Develoment of an automated pill medicine defect system using camera and web application.pdf - Submitted Version Download (2MB) | Preview

Abstract

Automated defect detection systems have revolutionized industrial quality control by enhancing efficiency and consistency. This project introduces an automated pill inspection system for pharmaceutical manufacturing, aiming to improve product quality and safety. The system integrates a camera-based inspection mechanism, a web application interface, a Raspberry Pi controller, and a DC motor-driven conveyor. By utilizing OpenCV, the system processes high-resolution images captured by a connected camera in real-time to detect defects such as size deviations, colour inconsistencies, and shape irregularities. Specifically, the system categorizes pills into three groups Pill A (Paracetamol) in the yellow-green category, Pill B (Aspirin) in the white-blue category, and Pill C (Adezio) in the red-black category. Initial tests demonstrated a defect detection accuracy of 86%, with the system inspecting up to 30 pills per test by continuously refilling the dispenser, which has a single-fill capacity of 15 pills. The web application enables real-time monitoring, data visualization, and defect categorization, fostering informed decision-making. A servo-controlled dispenser ensures defective pills are promptly removed from the production line. Preliminary results highlight the system's potential for practical implementation in pharmaceutical production, addressing industry demands for automation while contributing to sustainable manufacturing practices through reduced product waste.

Actions (login required)

View Item