Aircraft skin surface damage detection method and system based on deep learning

Abstract

The invention discloses an aircraft skin surface damage detection method and system based on deep learning, and relates to the technical field of image processing, and the method mainly comprises the steps: obtaining a currently collected aircraft skin surface image; inputting a currently collected aircraft skin surface image into the trained aircraft skin surface damage detection model to perform damage category detection and damage area segmentation; the trained aircraft skin surface damage detection model is determined based on a deep learning neural network and a training data set; the deep learning neural network comprises a feature extraction network, an attention module, a multi-path multi-scale feature fusion module, a multi-path multi-scale feature fusion module, a target detection network and a full convolution segmentation network; and the attention module is used for recoding the original feature map output by the feature extraction network and determining a multi-scale feature map. According to the invention, low-cost, high-efficiency, accurate and damage-free detection can be carried out on the surface damage of the aircraft skin.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method and system for detecting damage to an aircraft skin surface based on deep learning. Background technique [0002] Due to its fast and convenient features, airplanes have become one of the common means of transportation for long-distance travel. Therefore, it is extremely important to ensure flight safety, and it is necessary to check the airplane and troubleshoot it in time. Aircraft skin damage is one of the main factors affecting flight safety. Typical skin damage in civil aviation aircraft includes paint peeling, cracks, loose rivets, missing or pits, etc. These damages are not only appearance, in some cases Structural damage, such as cracks at joints, missing rivets, etc., will result in serious skin damage, which will pose a safety hazard and affect the normal use of the aircraft. Therefore, aircraft skin damage detection and maintenance is an impo...

AI Technical Summary

Problems solved by technology

[0003] However, at present, visual inspection methods are mostly used for the detection of aircraft skin damage. Due to the large size of the aircraft, it will take a lot of time to rely solely on visual inspection by inspectors, with high work intensity and low efficiency. Subjective factors, prone to missed inspections, if the damage cannot be discovered and maintained in time, it may lead to serious safety hazards

Although traditional aircraft skin non-destructive testing methods mainly include ultrasonic testing, penetrant testing, and eddy current testing, these methods have high reliability without compromising the performance of the skin, but the testing costs are expensive and the testing process is cumbersome. Subjective judgment by professionals

Images