// AI news · 2026-11-07 · Nature
Machine Vision System for Steel Slab Dimension Measurement
A machine vision system using deep neural networks measures steel slab dimensions in real production conditions.
Published 2026-11-07 08:00 (Beijing time)Source NatureAuto-summarised, fact-checked against source
01Summary
It combines dual-camera imaging, DeepLabV3+ with EfficientNet-B5 backbone, RANSAC boundary fitting, and height-dependent calibration. Evaluated on 65 unseen slabs, mean absolute errors are 2.56 mm (length), 1.84 mm (width), and 1.04 mm (thickness). Within-tolerance agreement rates reach 98.5%, 96.9%, and 95.4% under plant-specific tolerances. The full measurement cycle takes ~4.2 seconds per slab and is deployed on a production line.
Summarised from the original article by an open-source model, then checked automatically: every number and English name must appear in the source. The original article is authoritative.
02Key points
- Uses dual-camera imaging and DeepLabV3+ with EfficientNet-B5 for dimension measurement.
- Mean absolute errors are 2.56 mm, 1.84 mm, and 1.04 mm for length, width, and thickness.
- Achieves 98.5% to 95.4% within-tolerance agreement rates under factory tolerances.
- Deployed on a production line with 4.2-second per slab measurement cycle.
03Source
Nature · published 2026-11-07 08:00 Beijing time
05Related news
- Same source · Multi-omics reveals SALM macrophage phenotype in CRC2026-11-07 · Nature
- Same source · ChGBO_QRNN for Intrusion Detection in IoT-SDN2026-11-07 · Nature
- Same topic · Comparative analysis of long-term electricity load forecasting2026-11-07 · Nature
- Same topic · MultiSHAP explains cross-modal interactions in multimodal AI2026-11-07 · Nature
- Same day · Network toxicology and ML reveal DOTP's role in SLE pathogenesis2026-11-07 · Nature
From AIMap AI news: collected from 1,100+ official and media sources. Summaries only restate the original article; follow the link for the full story. All stories of 2026-11-07 · Archive by date · All AI news · Today