Robotic enhanced visual inspection of aircraft skin

摘要

Visual inspection is the most widely used method of commercial aircraft surface inspection. We have developed a remote visual inspection system, designed to facilitate demonstrating the feasibility and advantages of remote visual inspection of aircraft surfaces. We describe experiments testing image understanding algorithms to aid remote visual inspection by enhancing and recognizing surface cracks and corrosion from live imagery, and key features of the mobile robot platform and infrastructure that delivers the image. We discuss performance of the image understanding algorithms and speculate on their future use in aircraft surface inspection for crack and corrosion detection. A neural net classifier worked best for corrosion, whereas a fuzzy logic algorithm worked best for cracks.