Cascade convolutional neural network-based automatic detection method for welding seam of steel part

Abstract

The invention provides a cascaded convolutional neural network-based steel part welding seam automatic detection method, which comprises the following steps of S1, scanning a steel part by using an industrial camera and infrared laser rays to obtain welding seam data; s2, performing weld joint area labeling and center line labeling on the image, and completing manufacturing of a weld joint reference image set and a center line reference image set; s3, according to a training algorithm, using the weld data and the weld reference image set to train a cascade convolutional neural network, and using the weld data and the center line reference image set to train a center line extraction network; s4, scanning the to-be-detected steel part by using the industrial camera and the infrared laser rays to obtain to-be-detected weld data; and S5, inputting to-be-detected weld data into the trained detection model, and outputting a detection result. According to the method, the position of the welding seam can be accurately extracted, the anti-interference capability is greatly improved, the welding quality is ensured, and the self-adaptive capability of an automatic welding system is improved.

Description

technical field [0001] The invention belongs to the technical field of weld detection, and in particular relates to an automatic detection method for steel welds based on a cascaded convolutional neural network. Background technique [0002] Welding technology has become one of the widely used connection methods, mainly used in the fields of aerospace, electronics manufacturing, machinery manufacturing and shipbuilding. Yet the spot environment of welding is very harsh, and the harmful gas that welding torch produces during welding and the dazzling arc light that welding torch produces during welding are very easy to make welders' lives threatened. With the aging of the population, welding workers are decreasing year by year, but welding demand is increasing year by year. Traditional manual welding cannot meet the needs of the current society. [0003] As the development of artificial intelligence industry and electronic manufacturing technology provides new ideas and new t...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the deficiencies of the prior art, and to provide an automatic detection method for steel welds based on cascaded convolutional neural networks, so as to solve the problem that the weld detection method designed for a specific type of steel can achieve high precision. detection, and to improve the technical problems of low detection accuracy of steel welds of different types and strong arc interference. Accurately extract the welding seam position, center line position and feature point position under equal interference, which has the advantages of strong anti-interference ability and accurate recognition

Images