Light Driven Visual Inspection System for Human Vision

This paper presents a study on optimizing light spectra for enhancing the contrast between two objects in visual inspection systems. The research focuses on using LED color mixing, a process that involves modulating the output of multiple narrowband LEDs, to generate optimized spectra that can best separate the color appearance of the objects. The algorithm used in the study maximizes contrast based on human perception. The study was conducted using two different light sources: a commercially available 4-channel luminaire and a custom-built 11-channel luminaire. The results showed that the 4-channel luminaire improved contrast by 31% and the 11-channel luminaire improved contrast by 72% for the 77 color pairs tested.