摘要
针对结构光三维测量系统中由于被测物表面反射率不均匀和曲率变化较大引起的条纹灰度退化现象,从而导致条纹中心提取精度下降的问题,提出了一种灰度调整方法。首先通过投影单色灰度图像到被测物体上,计算出摄像机采集图像中各个像素点相对于物体表面的调制信息,然后逐点将其代入条纹图像中进行灰度运算,以消除灰度退化项,实现条纹的灰度调整。分别对黑白相间的平板和人脸石膏模型进行实验,结果表明:经灰度调整后,各条纹在横截面上基本满足高斯分布,且条纹灰度分布均匀,从而有效修正了因反射率突变引起的重心偏移;同时解决了复杂物体图像中因亮度不均匀引起的条纹中心局部断线问题,从而保证了测量数据的完整性。上述结果验证了该方法的有效性和可靠性。
In 3-D measurement of structured light system,the fringe grayscale degradation caused by the uneven reflectivity and the large curvature change of measured object surface is frequently encountered.A grayscale adjustment method is thus proposed to improve the accuracy of stripe center extraction.The modulation of each image pixel related to the object surface is obtained by projecting pure gray images to the measured object.And the modulation is applied to the fringe pattern point by point to eliminate the grayscale degeneration term by grayscale operation to adjust the fringe grayscale.The tests with a plate in black and white check and a human face model are conducted.It is shown that each stripe basically follows Gauss distribution on its cross section after the grayscale adjustment,and the fringe grayscale is well distributed,so the deviation caused by the mutation of reflectivity is corrected effectively.The local disconnection of the center line caused by the unevenness of fringe brightness in complex object image is solved meanwhile to ensure the integrity of measured data.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2018年第2期30-36,共7页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(51375377)
陕西省自然科学基础研究计划资助项目(2017ZDJC-21)
关键词
结构光测量
中心提取
灰度退化
调整
structured light measurement
center extraction
grayscale degradation
adjustment