Design and Implementation Stock Opname for Tire Using AI and API

Abstract

The stock opname application for tire sales must have features that help in recording, monitoring, and analyzing stock in real time. The traditional stock opname process is carried out by workers working in the warehouse using a scanner or by manually counting SKUs, but this requires careful workers so that the stock calculation is correct. The purpose of this study is to create a tire stock opname application using Computer Vision and AI (Artificial Intelligence) technology. By using tire photos, tire information can be obtained such as Brand, Model, DOT, and Size, so you only need to input the description and amount of stock. The stock opname application is built using Flutter, and is supported by Flask which acts as an API, utilizing YOLOv8 and TrOCR to extract text from the tire photos entered.