Home > About Fuji Electric > Research & Development > R&D Activities > Representative R&D Projects > Automatic Appearance Inspection Technology Using Deep Learning

Automatic Appearance Inspection Technology Using Deep Learning

  • Digital innovation Laboratory
  • Energy Solution
  • Industry Solution
  • Fundamental and cutting-edge technologies
  • Production Technologies
  • Safety and Security
  • IoT
  • Automatic feature extraction using deep learning
  • Multiple classifier combination technology
  • Abnormal location heat map display

Algorithms for industrial product appearance inspection were previously developed by specialists. Moreover, depending on the content being evaluated, it was often necessary to rely on manual sensory evaluation.

In response to this, Fuji been engaged in the research and development of automatic visual inspection technology using deep learning to facilitate the automation of sensory evaluation without having to rely on specialists to develop algorithms. This technology contributes to improvements in both manpower saving and quality.

Flow example for automatic visual inspection using deep learning