Automated Optical Inspection

Optical Drift

Illumination intensity degrades over time as light sources age, which shifts colour temperature and luminance values across the image field. Lens distortion and mechanical vibration alter spatial mapping between physical coordinates and pixel locations. Regular calibration against certified glass targets re-establishes geometric accuracy and brightness uniformity across the field of view.

False Rejection

Algorithmic classification depends on defined tolerance bands surrounding nominal component dimensions and solder fillet geometry. When component packaging variances or board warp occur within acceptable fabrication limits, automated optical inspection systems frequently flag benign variations as critical assembly defects. Production environments manage this over-rejection rate by adjusting specular reflection thresholds and refining component footprint libraries.

Operators verify flagged locations on dedicated review stations to prevent unnecessary rework of conforming assemblies. Misclassified components require manual re-inspection, adding operational cost and cycle time to high-volume manufacturing lines.

Threshold Calibration

Verification protocols establish golden board standards by capturing statistical distributions across known good production samples. Quality engineers define upper and lower limits for contrast, edge detection and pad coverage to balance fault coverage against false alarm rates. Final qualification requires verification against standard defect catalogs prior to line release.