Automated Defect Recognition of Castings defects using Neural Networks

TL;DR

This paper presents a CNN-based Automated Defect Recognition system for industrial X-ray images, achieving high accuracy and fast inference, which can improve reliability and efficiency in defect detection tasks.

Contribution

The paper introduces a CNN model that surpasses existing methods in accuracy for industrial defect detection and demonstrates its suitability for real-time deployment.

Findings

Achieved 94.2% accuracy (mAP@IoU=50%) on GDXray dataset.

Inference time is less than 400 ms per image, suitable for industrial use.

Hyper-parameter optimization improved accuracy from 75% to 94.2%.

Abstract

Industrial X-ray analysis is common in aerospace, automotive or nuclear industries where structural integrity of some parts needs to be guaranteed. However, the interpretation of radiographic images is sometimes difficult and may lead to two experts disagree on defect classification. The Automated Defect Recognition (ADR) system presented herein will reduce the analysis time and will also help reducing the subjective interpretation of the defects while increasing the reliability of the human inspector. Our Convolutional Neural Network (CNN) model achieves 94.2\% accuracy (mAP@IoU=50\%), which is considered as similar to expected human performance, when applied to an automotive aluminium castings dataset (GDXray), exceeding current state of the art for this dataset. On an industrial environment, its inference time is less than 400 ms per DICOM image, so it can be installed on production facilities with no impact on delivery time. In addition, an ablation study of the main hyper-parameters to optimise model accuracy from the initial baseline result of 75\% mAP up to 94.2\% mAP, was also conducted.