Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
May 31, 2026Proceedings of the Institution of Mechanical Engineers Part C Journal of Mechanical Engineering Science

Research on a large-scale non-planar metal surface defect detection method based on chocolate block array stitching approach and improved YOLOv8 algorithm

View Full Paper
Ask AI
Bookmark
Share

Authors

ZZZheng ZouLHLiang HeSMShoutao Ma

Discussion

Loading...

Member takes

Overview

Randomized trial demonstrates improved defect detection in non-planar metal surfaces, indicating efficiency benefits.

Key Points

  • This research aims to enhance defect detection on large-scale non-planar metal surfaces by improving image stitching and YOLOv8 algorithms.
  • Developed a Chocolate Block Array Stitching approach for optimal camera pose and efficient image synthesis.
  • Improved YOLOv8 algorithm with Dilation-Wise Residual Module, Efficient Channel Attention, and Wise Intersection over Union loss function.
  • Comparison of the proposed method's efficiency and precision against the original YOLOv8.
  • Achieved 1.8% increase in precision with proposed DEW-YOLOv8 algorithm.
  • 1.3% improvement in recall and mean average precision (mAP).
  • Combined approach reduced time cost by 65.5% compared to direct full image detection.

Cite This Study

Zou et al. (2026) studied this question.

synapsesocial.com/papers/6a1bd2ab5783ba022b6fe2a6https://doi.org/10.1177/09544062261453408
View Full Paper
Ask AI
Bookmark
Share