Cosmetic Defect Analysis
Detect subtle and unpredictable defects anywhere on the smartphone
Graphical programming environment for deep learning-based industrial image analysis
After a smartphone is fully assembled but before it proceeds to packaging, it must be inspected for scratches, cracks, chips, dents, misalignment, discoloration, and other defects which may be in multiple places anywhere on the housing and cover glass. These defects do not generally affect the function of the device but detract from the appearance of the product.
Traditional rule-based vision applications can be trained on a range of typical defects, such as a scratch in a predefined area, or a crack that tends to appear in a screen corner, but the range of possible defects is extremely large and can appear anywhere on the phone. Even a relatively infrequent defect, such as discoloration of the housing, or dent caused by the impact of a robot arm, needs to be caught before packaging. Given the rate at which phones are produced, human inspection is inconsistent with low efficiency.
Cognex Deep Learning’s defect detection tool can learn to find a wide range of unacceptable product defects throughout the manufacturing process. The tool examines the screen, the band, and the back to detect any combination of dents, scratches, and discolorations anywhere on the smartphone. To spot all defects, the inspection leverages special lighting and proper part presentation to ensure that only products without cosmetic defects move on to the packaging step.