No abstract

No abstract

No abstract

In order to solve the problems such as unreasonable plan, opaque process, insufficient flexibility and excessive manual participation in the process of screening, identification and quality consistency inspection of electronic components in China, a software and hardware construction scheme of intelligent detection system for electronic components is proposed, which integrates the cutting-edge technologies such as machine vision, deep learning, IOT(internet of things), data mining. Relying on the National Semiconductor Device Quality Supervision and Inspection Center, part of the construction work has been completed, and the information, automation and intelligence of the testing business have been realized.

Aiming at the problems of low efficiency and low accuracy caused by manual defect detection of electronic components, FCN, SegNet/DeconvNet and DeepLab and other deep learning technologies are studied. Using Caffe, Keras, PyTorch and other frameworks, an automatic defect identification system for electronic components is developed to distinguish qualified products from unqualified ones, so as to realize intelligent defect detection of electronic components. At the same time, the detection accuracy rate is greatly improved, and the quality of electronic components is guaranteed.