Visual inspection of automotive assembly parts

As a well-known domestic and international research and development enterprise of packaging intelligent automation equipmentThe technical services of Shanghai Lujia Automation Technology Co., Ltd. provide intelligent visual inspection equipment technology solutions for industrial components that are synchronized with international standards for China's manufacturing industry. Visual inspection of automotive assembly partsapply toMajor industries such as pharmaceuticals, food, beverages, daily chemicals, health products, electronics, electrical appliances, chemicals, automotive industry, and plastics and hardware!

Intelligent visual inspection equipment for automotive assembly partsexistDigital image processing technology is an emerging technology industryIt has been widely applied in fields such as automation systems, automotive parts testing, and intelligent recognition. It has become one of the important solutions for traditional manual detection with slow speed and low detection efficiency. Due to the numerous defects in the details of industrial parts in actual production, it is necessary to select appropriate algorithms for accurate identification and detection. This article focuses on the design of an image detection system for the back panel components of automotive energy absorbing boxes. An experimental hardware platform is built, and various components used in the vision system and the composition of the lighting system are described in detail. The camera system is then calibrated to correct distortion effects. After obtaining the corrected image, key technologies such as image preprocessing, edge detection, and measurement of geometric parameters of parts were focused on. In preprocessing, the noise category of the image was first analyzed, and various filtering algorithms were compared to find the suitable filtering algorithm for the image in this article. Furthermore, in image edge detection, classic edge detection algorithms were compared, providing a foundation for subsequent feature extraction. When detecting the basic features of an image, circles and lines in the image were separately detected, and the parameters of the detection results were optimized to improve the detection performance of circles and lines. When detecting grooves in the image, a template matching algorithm was used to accurately identify the position of the grooves. After the inspection of part dimensions, the article also studied the classification and recognition methods for intact parts, welded parts, and scratched parts. Firstly, through edge detection, while ensuring clear and complete image edges, the gradient direction histogram algorithm is used for feature extraction, and probabilistic neural networks and SVM are used for classification and recognition, achieving good classification results. However, due to the high dimensionality of feature vectors and the overlapping of feature extraction information, it is difficult to fully utilize the key information in the image. The gradient direction histogram algorithm was improved in the article by bilinear interpolation of the gradient direction histogram feature extraction algorithm to obtain feature vectors that better reflect detailed features. Then, neural networks and support vector machines were used for recognition, which not only improved the anti aliasing effect of feature values, but also increased the accuracy of image classification and recognition. The implementation of this module is based on Visual C++and MATLAB, including the development of visual system interfaces and algorithm writing. This article realizes the detection of part features and the classification and recognition of different types of parts. The research results in the article reflect certain engineering value, and provide certain reference significance for the application of image measurement technology and the classification and recognition of parts.