Detection and Classification of Aircraft Structural Defects for Database Creation and Findings Identification

Abstract

In aviation, maintaining structural integrity [1] It is crucial to maintain aviation safety and operational security. Surface wear, corrosion, and cracks are [2], [3], [4] typical structural defects that can seriously compromise components for aircraft. Employing innovative image processing techniques [5]This study provides a comprehensive approach to support the creation of systems that enable the automatic recognition and classification of these findings. The primary objective is to develop a verified image-based database that improves maintenance processes and inspection performance. The process basis is a structured finding catalogue that was created after an extensive examination of scientific and industrial sources. This catalogue standardises terminology and makes it easier to manually annotate and classify defects consistently. A rigorous pipeline that includes image collection from various sources, data augmentation to improve generalisation, manual annotation based on the catalogue, and expert validation to guarantee accuracy and consistency is used to build the dataset. A crucial component of this initiative is the Aircraft Inspection [6], [7], [8] Repository. By acting as a centralised platform that improves data accessibility, expedites maintenance workflows, and guarantees regulatory compliance, it is intended to address the challenges of gathering, monitoring, and analysing inspection data. The repository greatly improves maintenance planning and decision-making by arranging inspection records across various aircraft models, providing dynamic data analysis tools, and enabling collaborative access to findings.