Related Experiment Video
Updated: Jun 9, 2025
07:37
Full-field Strain Measurements for Microstructurally Small Fatigue Crack Propagation Using Digital Image Correlation Method
Published on: January 16, 2019
9.6K
The edge artifact extraction method for Si3N4 ceramic bearing rolling element microcracks characterization
Tao Chen1,2, Xin Xia1, Jianbin Zhou1
1School of Mechanical and Electronic Engineering, Jingdezhen Ceramic University, Jingdezhen, 333403, Jiangxi, China.
Scientific Reports
|October 29, 2024
Summary
This study introduces a novel method using 2D discrete wavelet transform and Otsu thresholding to enhance microcrack edge artifact detection in silicon nitride (Si3N4) ceramic bearings. The technique effectively improves image contrast and noise reduction for clearer feature extraction.
Area of Science:
- Materials Science
- Ceramic Engineering
- Tribology
Background:
- Microcrack detection in silicon nitride (Si3N4) ceramic bearings is challenging due to low contrast, noise, and background merging of edge artifacts.
- Accurate segmentation of these microcracks is crucial for assessing bearing integrity and performance.
Purpose of the Study:
- To develop and validate a robust method for enhancing the visibility and extractability of microcrack edge artifacts in Si3N4 ceramic bearing components.
- To address the limitations of existing image processing techniques in segmenting low-contrast and noisy microcrack features.
Main Methods:
- A novel approach combining 2D discrete wavelet transform (DWT) for noise reduction and detail restoration with Otsu threshold segmentation for artifact removal.
- Wavelet decomposition and reconstruction functional equations were created to improve image contrast and eliminate noise.
- Otsu thresholding was optimized by maximizing interclass variance to effectively remove difficult feature edge artifacts.
Main Results:
- The proposed method significantly enhanced the contrast of microcrack edge artifact features in Si3N4 ceramic bearing rolling bodies.
- Experimental results demonstrated an average Peak Signal-to-Noise Ratio (PSNR) close to 62.69 dB and an average Structural Similarity Index Measure (SSIM) of approximately 0.77.
- The technique achieved more complete feature extraction for point, line, and surface microcrack edge artifacts.
Conclusions:
- The developed 2D DWT and Otsu thresholding method provides an effective solution for segmenting challenging microcrack edge artifacts in Si3N4 ceramic bearings.
- This advancement offers improved image quality and more accurate feature extraction, contributing to better non-destructive evaluation of ceramic components.
