Flying Metrology and Defect Identification for Aircraft Surface Inspection

Abstract

A commercial aircraft is struck by lightning on average once a year. To rule out possible damage to the aircraft, a time‐consuming visual inspection of the aircraft is carried out by maintenance staff. In order to reduce overall maintenance costs and aircraft on ground time, an autonomous unpiloted aerial vehicle is utilized. The mobile unit allows an easy inspection of the compromised area, by carrying industrial camera technology to digitalize the aircrafts’ surface. So this approach applies and extends classic machine vision as well as machine learning algorithms to automatically detect maintenance‐relevant surface defects in location‐indexed images recorded by the system.

Flexible High-Resolution Continuous 3-D Scanning for Large-Scale Industrial Components

“…Surfacestructured light is more commonly used, and Fourier transform profilometry (FTP) [15] or phase-shifting profilometry (PSP) [16] is often used to generate a high-density point cloud. Point cloud registration methods include mechanical positioning 2 [17], iterative closest point (ICP) [18], registration based on marks distributed on the measured surface [19,20], and registration based on pose measurement [4,21]. Methods for point cloud generation regarding line-scan camera systems include FTP [22], the fringe skeleton method [23], pseudorandom fringe [24], and the time correlation method [25].…”

Damage simulation experiment of aircraft skin coating based on hyperspectrum

“…At present, aircraft surface damage detection technology mainly includes manual vision inspection and machine vision automatic detection, among which, manual visual detection accounts for about 90% of the total detection [3][4] . However, the manual inspection method based on visual inspection has three shortcomings: one is the large number of fasteners on the aircraft, as many as hundreds of thousands of fasteners on the fuselage and wings [5] .…”

“…Third, due to the heavy testing tasks of aircraft, in order to timely complete the testing and maintenance of aircraft in service, the airport needs to hire a large number of inspectors under the condition of protecting the rights and interests of employees, resulting in the increase of airport operating costs. Machine vision detection mainly identifies damage by comparing pictures, and the recognition efficiency is low when the color and shape are the same but the material is different [3] . Therefore, it is urgent to develop a new technology for aircraft surface nondestructive testing.…”