Defect inspection system and method using artificil intelligence

A defect inspection method using artificial intelligence for a defect inspection system connected to an automatic visual inspection device, wherein the defect inspection system comprises a re-inspection server, a training terminal, a labeling re-inspection terminal, a classification terminal, and a re-inspection terminal, the defect inspection method comprising:

• receiving a defect image and a defect location from the automatic visual inspection device by the re-inspection server;
• receiving the defect image and the defect location from the re-inspection server by the classification terminal;
• reading a target trained module corresponding to the defect image, classifying the defect image according to the target trained module to generate a labeled defect image, and sending the labeled defect image to the re-inspection server by the classification terminal;
• receiving, by the re-inspection terminal, the labeled defect image sent from the re-inspection server and displaying the labeled defect image on a display of the re-inspection terminal;
• receiving a verified operation corresponding to the labeled defect image and sending the verified operation to the re-inspection server by the re-inspection terminal;
• receiving the verified operation and the labeled defect image from the re-inspection server and receiving a labeling result corresponding to the labeled defect image by the labeling re-inspection terminal;
• receiving the labeling result and the labeled defect image and sending the labeling result and the labeled defect image to the training terminal by the re-inspection server;
• and
• receiving the labeling result and the corresponding labeled defect image from the labeling re-inspection terminal and training a corresponding training module according to the labeling result and the corresponding labeled defect image by the training terminal.