Mobile Device Speaker Mesh Inspection
Confidently flag defects on complex and visually confusing meshes
Related Products
Powered by In-Sight ViDi Deep Learning-Based Vision Software
Graphical programming environment for deep learning-based industrial image analysis
The speaker mesh is a piece of perforated metal that protects a cellphone’s acoustic module from dust and damage while minimally affecting the sound.
Meshes must be visually inspected for cracks, scratches, missed perforations, deformation, broken edges, foreign inclusions such as dirt or hair, and damage from the perforation process. This mesh is visible to the end user, and so cosmetic damage must also be detected.
Because meshes are three-dimensional textured parts, they create complex lighting patterns of reflections and highlights. Even a small tilt in the mesh can completely change the pattern. Meanwhile, defects such as contamination or irregular scratches can appear anywhere on the part. This can be on the outer surface, or in the gaps between the wires. These variations make it difficult to program traditional machine vision to handle every possible case.
To identify speaker mesh defects, Cognex Deep Learning’s defect detection tool is trained on a wide selection of speaker meshes to learn the full variation of normal parts, including common lighting variations and the acceptable level of cosmetic defects. As it scans through meshes it analyzes and flags any that are outside of the acceptable range, while minimizing false positives.
Cognex Deep Learning can work in concert with traditional machine vision, which is best for precision alignment tasks and gauging the speaker mesh dimensions. The combination of deep learning plus traditional vision tools far outstrips that of manual inspection, at significantly higher speed. And it does so while maintaining consistency across shifts and production lines.
