DEVELOPMENT OF AN IMAGE DATASET AND CLASSIFICATION METHODS FOR DETECTING AUTOMOTIVE BUMPER DEFECTS USING ARTIFICIAL INTELLIGENCE

Abstract

This article examines the methods for developing an image database and classifying defects for detecting automotive bumper defects using artificial intelligence. The study analyzes the key stages required for automating quality control systems, including image acquisition, database organization, annotation, and defect classification principles. In addition, a methodology for creating a dataset suitable for deep learning algorithms has been developed. The proposed approach contributes to the automation of quality control in the automotive industry, reduction of human factor influence, and improvement of defect detection accuracy.