RESEARCH ON DESIGNING COLOR-BASED PRODUCT CLASSIFICATION SYSTEM APPLYING DIGITAL IMAGE PROCESSING TECHNOLOGY

Abstract

This paper presents the process of designing a product classification system according to the color characteristics of the product using high precision image processing technology. The system can recognize up to 58 different colors with many advantages over the classification systems using the number of color sensors that are limited in practice. Product classification system uses LabView-based image processing tools to classify products into different categories and integrates additional monitoring and remote control functions to bring high economic efficiency to the system. In the scope of the article, the authors applied an algorithm to detect the image colors obtained from the camera and then processed on 3 popular colors: red, yellow, and blue. Experimental results were conducted in four specific cases. The data shows that the system operates stably, with high accuracy: with the red product 100% accuracy; yellow products with 98% accuracy and blue products with 96% accuracy.