Identification of LED Lights with Different Colors Based on Computer Vision

In this study, a computer vision-based identification system is established for the identification of LED lights of different colors for industrial robots. By using a CCD camera, images of products on the industrial assembly line are collected. To reduce the effect of environmental brightness, the obtained images based on RGB model are converted into HSI color format. According to the obtained hue and saturation values, this system could accurately and effectively identify the color of the LED lights, which provides accurate identification and judgment for subsequent robot arm placement, correction, and other operations. The computer vision system could be applied to the industry to improve the automation level and product quality of industrial production.