OVERVIEW
To segregate more than 40,000 stickers a day with 30 variants in the running production line an AI-powered vision system was developed. In the process of segregation, with the help of image-based inspection, the defective stickers were routed to rejection bins. Deployment of this system helped in achieving desired inspection time, on the other hand, it freed human resources that can be utilized for other tasks.
PARAMETERS TO BE INSPECTED –
Color of the sticker
Design of the stickers
DEFECTS –
Cuts on the stickers
Smudged prints
Pinholes
Dull or fade colored stickers have to be rejected from the line)
Keywords: Sticker Classification, Computer Vision For Classification, Sticker Color Classification, Detect Smudged Prints, Computer Vision Image Classification
