Automatic Optimal Lighting Adjustment and Control for Vision Recognition

Nelson Yen-Chung Cang ； Chin-Chia Wu

《Proceedings of the 14th IFToMM World Congress》 14th-3 (2015/10) Pp. 484-487

https://doi.org/10.6567/IFToMM.14TH.WC.OS13.091

並列摘要

Lighting configuration has always affected how a vision recognition system performs. In industry application, the adjustment of lighting configuration relied on experienced engineer’s manual tuning. However, this is laborious, time consuming, and sometimes not effective. To make lighting configuration easier, less time consuming, and optimal, this work proposed a fully automatic adjustment approach to tune lighting configuration, which includes lighting height, angle, and strength, based on photo-realistic rendering technique. Experiments compared semi-automatic adjustment and fully automatic adjustment approaches on an inspection task. Experiment results showed that fully automatic adjustment approach was 3.12 times faster than semi-automatic adjustment approach. Experiment also suggested lighting configurations that were more likely to give better inspection performance.

並列關鍵字

Vision ； lighting ； optimization