采用遗传算法在相位域对抖动算法产生的二值条纹图案进行优化, 同时, 为了解决遗传算法的耗时问题, 用对二值块的优化来代替整个条纹图案的优化.实验结果表明: 本文方法对不同离焦量具有鲁棒性, 能极大改善抖动算法的测量质量; 在轻度离焦、中度离焦和重度离焦的情况下, 相位均方根误差分别由1.493, 1.209, 0.989 rad, 减小到0.972, 0.749, 0.603 rad.

A genetic framework to optimize the dithering binary pattern in the phase domain was proposed, which searches a best binary patch instead of the whole pattern to handle the time-consuming difficulty of genetic algorithm. Experimental results show that the proposed technique can significantly improve the dithering fringe quality and it is robust to different amounts of defocusing. The phase root mean square error reduces from 1.493, 1.209, 0.989 rad to 0.972, 0.749, 0.603 rad respectively, when the projector is mildly defocused, moderately defoucsed, and severely defocused.