Aerospace composite material defect positioning and identification method and system

Abstract

The invention provides an aerospace composite material defect positioning and identification method and system. The method comprises the following steps: S1, constructing a detection network for positioning and identifying defects of aerospace composite materials; S2, pre-training the detection network on an ImageNet data set, and obtaining pre-training weight values of a backbone network and a feature extraction network; S3, acquiring aerospace composite material defect image samples, and marking labels of the image samples; S4, inputting an image sample set and the labels into the whole detection network for training to obtain a weight, a defect position and category information of the detection network; and S5, obtaining a to-be-detected aerospace composite material defect image, inputting the image into the detection network, and calculating to obtain the defect position and defect category information in the to-be-detected image. According to the invention, defect positioning and identification are carried out on images containing aerospace composite material defects, the problem that the real-time performance is not high when X-ray images containing the aerospace composite material defects are manually identified is solved, and the defect identification precision is guaranteed.

Description

technical field [0001] The invention relates to the technical field of non-destructive testing and evaluation, in particular to a method and system for locating and identifying defects in aerospace composite materials. Background technique [0002] Compared with traditional metal materials and their alloys, carbon fiber reinforced plastics (CFRP) have excellent thermal and physical properties, including high strength and high stiffness-to-weight ratio, low thermal expansion coefficient, high fatigue resistance, inherent Corrosion resistance and low electromagnetic reflectivity. Therefore, it is more and more used in high-tech fields such as aerospace. However, there will be some defects in the production process and use of these composite materials, such as inclusion defects, porosity defects and loose defects caused by careless operation in the production process, and the aircraft parts made of this composite material are in Impact damage caused by accidents during aircra...

AI Technical Summary

Problems solved by technology

However, these algorithms often require a lot of calculation and memory consumption, and are difficult to run on embedded hardware platforms with limited computing power. At the same time, these algorithms cannot complete defect location, and manual visual inspection is still required for defect location

Images