A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning

Regular aircraft engine inspections play a crucial role in aviation safety. However, traditional inspections are often performed manually, relying heavily on the judgment and experience of operators. This paper presents a data-driven deep learning framework capable of automatically detecting defects...

Full description

Saved in:
Bibliographic Details
Main Authors: Thibaud Schaller, Jun Li, Karl W. Jenkins
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Journal of Experimental and Theoretical Analyses
Subjects:
Online Access:https://www.mdpi.com/2813-4648/3/1/4
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850205623433560064
author Thibaud Schaller
Jun Li
Karl W. Jenkins
author_facet Thibaud Schaller
Jun Li
Karl W. Jenkins
author_sort Thibaud Schaller
collection DOAJ
description Regular aircraft engine inspections play a crucial role in aviation safety. However, traditional inspections are often performed manually, relying heavily on the judgment and experience of operators. This paper presents a data-driven deep learning framework capable of automatically detecting defects on reactor blades. Specifically, this study develops Deep Neural Network models to detect defects in borescope images using various datasets, based on Computer Vision and YOLOv8n object detection techniques. Firstly, reactor blade images are collected from public resources and then annotated and preprocessed into different groups based on Computer Vision techniques. In addition, synthetic images are generated using Deep Convolutional Generative Adversarial Networks and a manual data augmentation approach by randomly pasting defects onto reactor blade images. YOLOv8n-based deep learning models are subsequently fine-tuned and trained on these dataset groups. The results indicate that the model trained on wide-shot blade images performs better overall at detecting defects on blades compared to the model trained on zoomed-in images. The comparison of multiple models’ results reveals inherent uncertainties in model performance that while some models trained on data enhanced by Computer Vision techniques may appear more reliable in some types of defect detection, the relationship between these techniques and subsequent results cannot be generalized. The impact of epochs and optimizers on the model’s performance indicates that incorporating rotated images and selecting an appropriate optimizer are key factors for effective model training. Furthermore, models trained solely on artificially generated images from collages perform poorly at detecting defects in real images. A potential solution is to train the model on both synthetic and real images. Future work will focus on improving the framework’s performance and conducting a more comprehensive uncertainty analysis by utilizing larger and more diverse datasets, supported by enhanced computational power.
format Article
id doaj-art-36ab9d95f8bb4292a75b04ed4efc07fc
institution OA Journals
issn 2813-4648
language English
publishDate 2025-02-01
publisher MDPI AG
record_format Article
series Journal of Experimental and Theoretical Analyses
spelling doaj-art-36ab9d95f8bb4292a75b04ed4efc07fc2025-08-20T02:11:03ZengMDPI AGJournal of Experimental and Theoretical Analyses2813-46482025-02-0131410.3390/jeta3010004A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep LearningThibaud Schaller0Jun Li1Karl W. Jenkins2ISAE-Supméca—Institute of Mechanics of Paris, 3 Rue Fernand Hainaut, 93400 Paris, FranceSchool of Aerospace, Transport and Manufacturing (SATM), Cranfield University, Cranfield MK43 0AL, Bedfordshire, UKSchool of Aerospace, Transport and Manufacturing (SATM), Cranfield University, Cranfield MK43 0AL, Bedfordshire, UKRegular aircraft engine inspections play a crucial role in aviation safety. However, traditional inspections are often performed manually, relying heavily on the judgment and experience of operators. This paper presents a data-driven deep learning framework capable of automatically detecting defects on reactor blades. Specifically, this study develops Deep Neural Network models to detect defects in borescope images using various datasets, based on Computer Vision and YOLOv8n object detection techniques. Firstly, reactor blade images are collected from public resources and then annotated and preprocessed into different groups based on Computer Vision techniques. In addition, synthetic images are generated using Deep Convolutional Generative Adversarial Networks and a manual data augmentation approach by randomly pasting defects onto reactor blade images. YOLOv8n-based deep learning models are subsequently fine-tuned and trained on these dataset groups. The results indicate that the model trained on wide-shot blade images performs better overall at detecting defects on blades compared to the model trained on zoomed-in images. The comparison of multiple models’ results reveals inherent uncertainties in model performance that while some models trained on data enhanced by Computer Vision techniques may appear more reliable in some types of defect detection, the relationship between these techniques and subsequent results cannot be generalized. The impact of epochs and optimizers on the model’s performance indicates that incorporating rotated images and selecting an appropriate optimizer are key factors for effective model training. Furthermore, models trained solely on artificially generated images from collages perform poorly at detecting defects in real images. A potential solution is to train the model on both synthetic and real images. Future work will focus on improving the framework’s performance and conducting a more comprehensive uncertainty analysis by utilizing larger and more diverse datasets, supported by enhanced computational power.https://www.mdpi.com/2813-4648/3/1/4blade imagesdata augmentationfine-tuningmAP50uncertainty and sensitivity analysismodel and evaluation uncertainties
spellingShingle Thibaud Schaller
Jun Li
Karl W. Jenkins
A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning
Journal of Experimental and Theoretical Analyses
blade images
data augmentation
fine-tuning
mAP50
uncertainty and sensitivity analysis
model and evaluation uncertainties
title A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning
title_full A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning
title_fullStr A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning
title_full_unstemmed A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning
title_short A Data-Driven Approach for Automatic Aircraft Engine Borescope Inspection Defect Detection Using Computer Vision and Deep Learning
title_sort data driven approach for automatic aircraft engine borescope inspection defect detection using computer vision and deep learning
topic blade images
data augmentation
fine-tuning
mAP50
uncertainty and sensitivity analysis
model and evaluation uncertainties
url https://www.mdpi.com/2813-4648/3/1/4
work_keys_str_mv AT thibaudschaller adatadrivenapproachforautomaticaircraftengineborescopeinspectiondefectdetectionusingcomputervisionanddeeplearning
AT junli adatadrivenapproachforautomaticaircraftengineborescopeinspectiondefectdetectionusingcomputervisionanddeeplearning
AT karlwjenkins adatadrivenapproachforautomaticaircraftengineborescopeinspectiondefectdetectionusingcomputervisionanddeeplearning
AT thibaudschaller datadrivenapproachforautomaticaircraftengineborescopeinspectiondefectdetectionusingcomputervisionanddeeplearning
AT junli datadrivenapproachforautomaticaircraftengineborescopeinspectiondefectdetectionusingcomputervisionanddeeplearning
AT karlwjenkins datadrivenapproachforautomaticaircraftengineborescopeinspectiondefectdetectionusingcomputervisionanddeeplearning