Aircraft surface defect detection system based on cloud edge cooperation and deep learning

Abstract

The invention discloses an aircraft surface defect detection system based on cloud edge cooperation and deep learning, which relates to the field of image detection and comprises a cloud end, an edgeside and a terminal, the system uses two neural networks for defect detection, one of the neural networks is a lightweight small neural network deployed on the edge side, and the other of the neural networks is a large neural network deployed on the cloud; and the edge side detects the picture acquired by the terminal to obtain a preliminary detection result, filters out a defect picture, uploadsthe defect picture to the cloud, accurately detects the defect position and the defect type by using the large neural network, and finally returns the result to the edge side. According to the invention, through cloud edge cooperation, the overall time delay of the system is reduced, and the utilization rate of network resources is also improved.

Description

technical field [0001] The invention relates to the field of image detection, in particular to an aircraft surface defect detection system based on cloud-edge collaboration and deep learning. Background technique [0002] Aircraft surface defect detection is a very critical part of aircraft manufacturing and daily maintenance. If the corrosion, potholes, and cracks on the fuselage surface are not remedied in time, it is likely to cause irreparable losses during the flight. At present, the detection of defects on the surface of the aircraft is generally carried out by human eye observation, and the possible defects are found by arranging workers with technical experience to inspect a certain area on the surface of the aircraft. The manual inspection method is relatively stable but there are many defects. First of all, human subjective will may lead to misjudgment of defects. Tired working conditions and different lighting environments may also cause workers to miss and misjud...

AI Technical Summary

Problems solved by technology

The manual inspection method is relatively stable but there are many defects. First of all, human subjective will may lead to misjudgment of defects. Fatigue working conditions and different lighting environments may also cause workers to miss and misjudge surface defects.

Secondly, the size of the aircraft is extremely large compared to humans, and maintenance workers need to rely on an external platform when performing surface inspections, which is extremely inconvenient and likely to lead to safety accidents

Third, the efficiency of manual inspection is extremely low, and it may be difficult to meet the quality inspection needs of its enterprises during frequent flights and mass production of aircraft

For the image detection method based on deep learning, first of all, the network resources in the industrial environment are limited, uploading pictures to the cloud server will consume a lot of network resources, and may crowd out the network resources of other industrial applications

Secondly, the results of surface defect detection need to be displayed in real time, but in the cloud system, due to the limitation of the on-site network, a large amount of delay will be generated, which will affect the real-time performance of the detection system

Images