• 15:35-16:00
  • Zaal 1

Deep Learning Vision Solutions in the Age of Industry 4.0

  • Dohee Han
  • Sualab

1. Overview of Smart Factory Ecosystem 1) Device Level 2) Factory Level (SCADA, PLC) 3) Enterprise Level (ERP, MES)
2. Technology Disruption Trends in Machine Vision Field 1) Introduction of Automated Inspection 2) Components and Process of Conventional Machine Vision System 3) Drawbacks of Conventional Machine Vision System - Frequent Changes in Optical Conditions - Difficulties in Rule-setting under Short Product Cycle Trends - Complexity of Images
3. Introduction of Deep Learning and Remaining Problems 1) Concept of Machine Learning and Deep Learning 2) Deep Learning for Machine Vision and Remaining Problems 3) How SUALAB Solved those Remaining Problems
4. Deep Learning S/W Library for Machine Vision 'SuaKIT' and its Functions 1) Functions of SuaKIT - Segmentation - Classification - Detection 2) Deep Learning Architectures of SuaKIT - Single Image Analysis - Image Comparison - Multi Image Analysis - One Class Learning 3) SuaKIT Usage Scenario 4) Expected Effects Using SuaKIT - Detection of Defects Undetectable by Conventional Machine Vision - Filtering Error Detection - Classification of Various Detect Types - Replacement of Conventional Machine Vision Algorithm