Segmentation and morphological analysis of wear track/particles images using machine learning

Article Environmental Sciences

Automated identification and quantification of tire wear particles (TWP) in airborne dust: SEM/EDX single particle analysis coupled to a machine learning classifier

Juanita Rausch et al.

Summary: The proportion of non-exhaust particles, including tire wear particles, has increased in recent years, but quantifying TWP remains challenging due to their low concentrations and heterogeneous characteristics.

SCIENCE OF THE TOTAL ENVIRONMENT (2022)

Add to Collection

Article Acoustics

A New Transfer Learning Ensemble Model with New Training Methods for Gear Wear Particle Recognition

Chunhua Zhao et al.

Summary: In this paper, an integrated model of LCNNE based on transfer learning is proposed to solve the acquisition problems of wear particle data of large-modulus gear teeth and few training datasets. Experimental results verify the superiority of the LCNNE model in wear particle identification and classification.

SHOCK AND VIBRATION (2022)

Add to Collection

Article Engineering, Mechanical

Lessons learned using machine learning to link third body particles morphology to interface rheology

Rabii Jaza et al.

Summary: This study investigates the correlation between the morphology of third body particles and the rheology of the contact interface using Machine Learning algorithms. The results suggest that Machine Learning has potential in quantitative tribological analysis when the morphological and rheological databases are properly enriched.

TRIBOLOGY INTERNATIONAL (2021)

Add to Collection

Article Engineering, Mechanical

Image processing applied to tribological dry contact analysis

Alizee Bouchot et al.

Summary: The study enriches the quantification of the rheology of the tribological interface through advanced image processing, including the study of traditional descriptors and new relevant descriptors. This approach shows promising results in understanding the mechanisms involved in dry contacts.

WEAR (2021)

Add to Collection

Article Engineering, Mechanical

FFWR-Net: A feature fusion wear particle recognition network for wear particle classification

Suli Fan et al.

Summary: A novel wear particle recognition network, FFWR-Net, is proposed for wear particle images classification by fusing features extracted from both traditional image processing method and deep learning convolutional neural network method. The proposed FFWR-Net classifier shows better accuracy and effectiveness compared to previous convolutional neural network models when tested on the same wear particle dataset.

JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

Towards a quantitative characterization of wear particles using image analysis and machine learning

Alizee Bouchot et al.

Summary: This study aims to achieve quantitative characterization through four steps: conducting pin-on-disk experiments, acquiring images of third body particles, processing the images, and extracting quantitative characteristics of third body particles.

FIFTEENTH INTERNATIONAL CONFERENCE ON QUALITY CONTROL BY ARTIFICIAL VISION (2021)

Add to Collection

Article Engineering, Mechanical