Procedural synthetic training data generation for AI-based defect detection in industrial surface inspection

Schmedemann, Ole; Baaß, Melvin; Schoepflin, Daniel; Schüppstuhl, Thorsten

doi:10.1016/j.procir.2022.05.115

Cited by 20 publications

(11 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Synthetic images are rendered from the scene. The pipeline for generating training data has been described in more detail in our previous work 8 . In this pipeline, image features that are not relevant for defect detection are randomised, such as the base color texture, the illumination intensity, or the viewpoint of the virtual camera.…”

Section: Approachmentioning

confidence: 99%

See 1 more Smart Citation

Adapting synthetic training data in deep learning-based visual surface inspection to improve transferability of simulations to real-world environments

Schmedemann,

Schlodinski,

Holst

et al. 2023

Automated Visual Inspection and Machine Vision V

View full text Add to dashboard Cite

Learning models from synthetic image data rendered from 3D models and applying them to real-world applications can reduce costs and improve performance when using deep learning for image processing in automated visual inspection tasks. However, sufficient generalisation from synthetic to real-world data is challenging, because synthetic samples only approximate the inherent structure of real-world images and lack image properties present in real-world data, a phenomenon called domain gap. In this work, we propose to combine synthetic generation approaches with CycleGAN, a style transfer method based on Generative Adversarial Networks (GANs). CycleGAN learns the inherent structure from real-world samples and adapts the synthetic data accordingly. We investigate how synthetic data can be adapted for a use case of visual inspection of automotive cast iron parts, and show that supervised deep object detectors trained on the adapted data can successfully generalise to realworld data and outperform object detectors trained on synthetic data alone. This demonstrates that generative domain adaptation helps to leverage synthetic data in deep learning-assisted inspection systems for automated visual inspection tasks.

show abstract

Section: Approachmentioning

confidence: 99%

“…1a shows example images from the use case. We created a dataset of rendered images with a procedural data generation pipeline 8 . Next we implemented the most common method for deep domain adaptation of unpaired image data, CycleGAN 7 .…”

Section: Introductionmentioning

confidence: 99%

Adapting synthetic training data in deep learning-based visual surface inspection to improve transferability of simulations to real-world environments

Schmedemann,

Schlodinski,

Holst

et al. 2023

Automated Visual Inspection and Machine Vision V

View full text Add to dashboard Cite

show abstract

“…In the first case, simulation software use 3D models of the object to represent generating random poses and simulating defects. [1][2][3] Defects are simulated through 3D modeling 2 or the application of textures. 1 Domain-adaptation technique are also applied to compensate the domain-gap between real and synthetic data.…”

Section: Generation Of Synthetic Data For Optical Quality Controlmentioning

confidence: 99%

“…1 Domain-adaptation technique are also applied to compensate the domain-gap between real and synthetic data. 2 The quality control neural network is trained with a combination of real and synthetic data or with synthetic data only and then tested with real data.…”

Section: Generation Of Synthetic Data For Optical Quality Controlmentioning

confidence: 99%

Synthetic data generation for improvement of machine learning-based optical quality control: a practical approach in the welding context

Mitri¹,

Frommknecht²,

Huber³

et al. 2023

Multimodal Sensing and Artificial Intelligence: Technologies and Applications III

View full text Add to dashboard Cite

Weld quality inspection allows the detection of defects that may compromise the quality and strength of the weld. Although visual optical inspection offers lower reliability than other non-destructive methods, it enables weld analysis at a significantly lower cost. In this context, developing machine learning-based algorithms for automatic optical weld quality recognition requires acquiring large amounts of data for training. This entails high costs in terms of time, material and energy required for test preparation. However, one possible approach to tackling the problem with limited datasets is to use synthetic data. Using such data increases the amount and variety of data available to the detection algorithm. With a focus on the context of welding, this paper presents an approach that uses synthetic data as a form of data augmentation to improve the performance of the optical detection of weld seams. Specifically, we propose a generative neural network for semantic image synthesis using a limited starting dataset. The network generates new data instances by receiving as input a semantic map of the image to be represented. Weld defects such as porosity or weld spatter are added to the semantic map so that the network synthesizes corresponding defect images. Analysing the performance on a segmentation network, experimental results show how adding synthetic data to the original data can ensure improvements in network performance.

show abstract

“…In [12] steel slab defects were generated procedurally in Blender to improve defect detection. In [13] blow holes on casting are detected using synthetic data. Existing approaches generate synthetic data procedurally while randomizing the process parameters.…”

Section: Introductionmentioning

confidence: 99%

Reinforcement learning optimized digital twin based synthetic data generation for defect detection of titanium spacer

Mohanty,

Su,

2023

Automated Visual Inspection and Machine Vision V

View full text Add to dashboard Cite

In the field of automatic defect detection, one of the major challenges for training accurate classifiers using supervised learning is the insufficient and limited diversity of datasets. Obtaining an adequate amount of image data depicting defective surfaces in an industrial setting is costly and time-consuming. Furthermore, the collected dataset may suffer from selection bias, resulting in underrepresentation of certain defect classes. This research aims to tackle the issue of surface defect detection in titanium metal spacer rings by introducing a novel approach that leverages a digital twin framework. The behavior of the digital representation is optimized using a reinforcement learning algorithm. Subsequently, the optimized digital twin is utilized to generate synthetic data, which is then employed to train a spacer defect detection classifier. The performance of this classifier is evaluated using real-world data. The results illustrate that the model trained with synthetic data outperforms the one trained on a limited amount of real data. This work emphasizes the potential of digital twin-based synthetic data generation and reinforcement learning optimization in enhancing spacer surface defect detection and addressing the data scarcity challenge in the field. When the generated synthetic data and real data combined is used to train inspection network, the inspection background accuracy reaches 93.07% and defect detection accuracy reaches 94.2% surpassing the defect detection performance of inspection network trained only using real data.

show abstract

Procedural synthetic training data generation for AI-based defect detection in industrial surface inspection

Cited by 20 publications

References 26 publications

Adapting synthetic training data in deep learning-based visual surface inspection to improve transferability of simulations to real-world environments

Adapting synthetic training data in deep learning-based visual surface inspection to improve transferability of simulations to real-world environments

Synthetic data generation for improvement of machine learning-based optical quality control: a practical approach in the welding context

Reinforcement learning optimized digital twin based synthetic data generation for defect detection of titanium spacer

Contact Info

Product

Resources

About