Application of Machine Vision Technology in Intelligent Recognition Devices for Miscellaneous Objects

Le résumé fourni par la source

This paper proposes a machine vision technology solution based on RCIR camera to address the difficulty of intelligent identification of impurities in tobacco sorting. Firstly, choose RCIR camera as the main image acquisition device, and enhance the recognition ability of complex backgrounds and substances with similar colors by combining the three spectral channels of red, cyan, and infrared. Subsequently, a series of image preprocessing methods were employed, including Gaussian filtering, histogram equalization, and image normalization, to enhance image quality and resolution. Then, by extracting shape, color, and texture features, a support vector machine (SVM) classifier is used to accurately distinguish between tobacco and miscellaneous items. The experimental results show that the average recognition accuracy of RCIR cameras is 92%, which is 7 percentage points higher than the 85% of traditional RGB cameras; The recall rate has increased from 82% for RGB to 90%; Meanwhile, the processing speed has been reduced from 1.6 seconds for RGB cameras to 1.1 seconds. The experimental results verified the significant advantages of the machine vision system based on RCIR camera in recognition accuracy, efficiency, and production line automation level, highlighting its potential application in tobacco sorting.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.