Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
August 27, 2024Open Access

Automating Multi-Analytical Tasks in Machine-Vision Enabled Rail Surface Inspections: A Three-Stage Deep Learning Based Method

View Full Paper
Ask AI
Bookmark
Share

Authors

TWTianxin WangUniversity of Oxford

Discussion

Loading...

Member takes

Implication

Computational evaluation demonstrates accurate rail surface defect identification in railway images, indicating feasibility for self-powered inspection systems.

Key Points

  • A three-stage deep learning framework automates detection, pixel-level localization, and classification of rail surface defects across real and synthetic datasets.
  • Evaluation of machine vision on rail image datasets confirms reliable defect classification, balancing computational load and inference time for real-world setups.
  • An autoencoder based generative model identifies anomalies to trigger a segmentation model, which may enable autonomous monitoring on self-powered railway vehicles.

Cite This Study

Tianxin Wang (2024) studied this question.

synapsesocial.com/papers/68e5ac88b6db643587545e66https://doi.org/10.4203/ccc.7.8.4
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 2 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Automatic Defect Detection on Hot-Rolled Flat Steel Products2012 · 318 citations
  2. 2Augmentation for small object detection2019 · 744 citations