Reverse Vision Measurement Simulation System Based On 3D Model

Traditional 3D visual inspection identifies and judges defects by obtaining 3D point clouds and analyzing point cloud features of products on the production line.The entire process is time-consuming and cannot adapt well to real-time detection of product defects on the produc-tion line.In addition,there is relatively little research on visual inspection of defects in curved objects such as steel pipes.Most enterprises still use inefficient traditional methods such as manual visual inspection to detect defects on the surface of steel pipes,and there is a great demand for steel pipe surface defect visual inspection systems that can cooperate with steel pipe production lines for real-time,high-precision,and high-precision detection.In response to these two issues,this article proposes a reverse vision measurement method based on the seamless steel pipe surface defect detection project for high-precision,real-time defect detec-tion and classification of products on the seamless steel pipe production line.By simulating the visual measurement scene and using a reverse vision measurement model,a standard state two-dimensional model of the tested steel pipe is obtained,which is compared with the real-time captured two-dimensional data of the tested steel pipe to achieve defect detection.The main work content and research results are as follows:(1)A reverse vision measurement method is proposed,which uses the multiline structured light measurement method to obtain the steel pipe point cloud.On this basis,the seamless steel pipe axis is extracted.Through the simulation of the multiline structured light measurement scene,the steel pipe standard state modeling is carried out.At the same time,the reverse vision model is established to map the standard state 3D steel pipe model to the two-dimensional plane,and finally the seamless steel pipe surface defect detection benchmark information is obtained.The experiment shows that there are a total of 138800 steel pipe point clouds obtained through the method in this article,with a point cloud density of 18.895/8)8)~2.The point cloud density is relatively high and can accurately reflect the detailed information of the steel pipe.The average deviation between the standard state two-dimensional model of seamless steel pipes and the center point of the standard non-destructive steel pipe light stripe is 0.1095 mm,which has high accuracy.(2)A steel pipe surface defect detection algorithm based on reverse vision measurement is proposed,which divides the specific types of defects into concave and convex types.The standard state information of seamless steel pipes and the center point of the light strip of the tested steel pipe are compared using a connected domain construction method,and a threshold is set according to the detection accuracy requirements to achieve real-time detection of the type and position of seamless steel pipe defects.The experiment shows that the accuracy of the surface defect detection method for steel pipes proposed in this article reaches 93.3%,and the average detection time of a single image does not exceed 66.7 ms,which can be used for practical industrial inspection.(3)A real-time detection system for seamless steel pipe surface defects based on reverse vision measurement has been established,and algorithms and system functional modules have been designed using technologies such as Qt and Open CV,and integrated into the seamless steel pipe surface defect detection system.The system mainly includes a reverse vision model building module,an image data processing module,a defect detection module,a data statistics module,and a data management module.This system can be used for automatic detection of surface defects on steel pipes on enterprise production lines,and staff can view system alarm information in real-time and handle defective steel pipes.