Research on Optimization of Assembly Quality Inspection System of Industrial Robot Based on Machine Vision and Improvement of Enterprise Benefit

Abstract

In the process of Industry 4.0 and industrial intelligence, industrial machine vision plays a crucial role in the quality inspection of workpiece assembly. It can effectively address production and manufacturing issues caused by workpiece quality problems during assembly, thereby significantly improving production quality and efficiency. This paper optimizes traditional industrial robot assembly quality inspection methods by adopting machine vision technology, combining hardware and algorithms to enhance the machine vision system. The TMS320C6711 digital signal processor serves as the core for image information processing, integrated with the EMIF interface, DMA, and interrupt mechanisms to achieve high-speed data transmission and real-time image processing capabilities. A multi-module collaborative solution based on SAA7111, CPLD, and MCU is adopted to optimize the image acquisition module, further enhancing system stability and flexibility. Additionally, an intelligent lighting optimization method combining PWM control algorithms with neural network scene recognition is employed to effectively address the issue of low assembly quality detection accuracy under varying lighting conditions, ensuring reliable performance in complex lighting environments. Results show that compared to traditional machine vision systems, the newly optimized system offers higher accuracy and classification accuracy, with detection accuracy improving from 92.77% to 94.15%, representing an 11.38% increase. Additionally, compared to traditional detection methods, the system significantly reduces detection time, enhancing the operational efficiency of industrial robots. It provides a reliable basis for assembly quality inspection in industrial robots and offers valuable insights for the application of machine vision in industrial fields.