Adopting opto-mechanical integration solutions, machine vision replaces the human eye, and deep learning is used to automatically detect product appearance defects, which facilitates timely adjustment of the front-end process and improves the yield rate.
Escape(Functional defects)
%
0
Escape(Non functional defects)
%
<0.3
Overkill
%
<3
Capacity UPH
pcs/h
1200
Retrofit time
min
<10
Items
Scratched, crushed, oiled, shiny, sticky, missing label content, crooked labels, broken differentials, artwork seam width.
Advantages
Compatible with a wide range of models, materials and sizes of adapters. If the equipment needs to expand the compatible models in the future, it can achieve other sizes of product inspection by replacing the simple workpiece. Convenient to adjust the front-end process in time to improve the yield rate.
Effects
With appearance inspection, material sorting, MES upload, PQM upload function. Detect the surface appearance defects, label content and structure dimensions appearing on six sides and twelve prongs of the power adapter.