Related Experiment Video
Updated: Jul 16, 2025

Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
Published on: January 5, 2024
GRP-YOLOv5: An Improved Bearing Defect Detection Algorithm Based on YOLOv5
Yue Zhao1, Bolun Chen1,2, Bushi Liu1
1Faculty of Computer and Software Engineering, Huaiyin Institute of Technology, Huaian 223003, China.
This study introduces GRP-YOLOv5, an enhanced algorithm for detecting bearing defects in chemical transmission equipment. The improved model achieves high accuracy but requires further optimization for real-time applications.
Area of Science:
- Mechanical Engineering
- Artificial Intelligence
- Materials Science
Background:
- Bearing defects in chemical transmission equipment cause significant production inefficiencies.
- Current detection methods may struggle with the visual similarity between defects and normal regions.
Purpose of the Study:
- To develop an improved bearing defect detection algorithm for enhanced accuracy and reliability.
- To address limitations in existing defect detection methodologies.
Main Methods:
- An improved bearing defect detection algorithm based on YOLOv5 (GRP-YOLOv5) was proposed.
- Image preprocessing involved gamma transformation to enhance contrast and grayscale.
- Feature extraction utilized ResC2Net for improved detail and semantic information capture.
- PConv convolution was integrated for feature fusion to increase network depth.
Main Results:
- The GRP-YOLOv5 model achieved a mean Average Precision (mAP@0.5) of 93.5% and mAP@0.5:0.95 of 52.7%.
- The model demonstrated excellent accuracy in bearing defect detection compared to other experimental models.
- The model size was optimized to 25 MB, but Frames Per Second (FPS) performance was noted as a limitation.
Conclusions:
- The GRP-YOLOv5 model offers a promising solution for accurate bearing defect detection.
- Further research is needed to optimize the model's processing speed for real-time applications.
Related Concept Videos
Bearings: Problem Solving
Improving Translational Accuracy
Detection of Gross Error: The Q Test
Bearing Stress
Due to the intricacy of these microforces, an average value, known as bearing stress, is often used by...
Detection of Black Holes
Their closest cousins are neutron stars, which are composed almost entirely of neutrons packed against each other, making them extremely dense. A neutron star has the same mass as the Sun but its diameter is only a few kilometers. Therefore, the escape velocity from their surface is close to the speed of light.
Not until the 1960s, when the first neutron...
Differential Leveling

