Optical Character Recognition of Lens Holder Part Association Codes
Accurately trace smartphone camera lenses through the manufacturing process by reading alphanumeric lens holder codes even when worn or damaged
Related Products
Powered by In-Sight ViDi Deep Learning-Based Vision Software
Graphical programming environment for deep learning-based industrial image analysis
Camera lenses are placed in holders for transportation from one station to another. These holders are designed to hold more than one size of lens. To track these, each holder has alphanumeric codes and orientation marks on it which associate the holder with the appropriate lens and enable traceability through the manufacturing process.
These holders are reused many times, and the alphanumeric codes and orientation marks become worn and difficult to read. Traditional machine vision can lose the ability to distinguish the closely spaced codes long before the lens holder itself is worn enough to require replacement. If these codes are read incorrectly, it can lead to incorrect lenses being delivered, or lens assemblies being misplaced.
Cognex Deep Learning’s OCR tool reads embossed and printed codes on challenging backgrounds and adjusts for wear, tight spacing, or ambiguous symbols. The OCR tool comes with a pretrained font library, making it easy to set up and deploy. After being trained on a set of images indicating the types of symbols used and their locations, including examples of worn, damaged, or ambiguous codes, this tool can then locate and read the camera lens holder codes. This enables the use of camera lens holders for their maximum lifetime while preventing misread issues from solutions that aren’t as well versed at varied OCR solutions.
