Sistem Penimbangan Cerdas Berbasis Vision

Abstract

The application of automation technology in weighing systems to improve efficiency and accuracy is becoming increasingly popular. Problems with conventional weighing systems, such as the inability to automatically recognize object types, inefficient weight measurements, and manual payment processes, have led to the development of the Vision-Based Smart Weighing System. This system is designed to automatically recognize fruit types, measure weight, and print payment receipts by combining Machine Learning, YOLO (You Only Look Once), and Convolutional Neural Network (CNN) technologies to enhance visualbased identification accuracy and weight estimation. In addition to using Load Cell sensors, the systemalso employs visual weight detection methods as an additional support for accuracy. Integrated with a payment gateway, the system enablesfast and secure digital payments. When fruit is placed on the scale, the camera captures an image, identifiesthe type and weight of the fruit, calculates the total price based on a combination of visual and Load Cell methods, then displays complete information on the LCD and automatically prints the transaction receipt. The locking door only opens after payment is successfully completed to ensure process security. This system makes the weighing process faster, more accurate, and more practical, and is highly suitable for implementation in supermarkets, grocery stores, and fruit shops to reduce human error and enhance customer convenience.