激光与光电子学进展, 2020, 57 (7): 071202, 网络出版: 2020-03-31
光亮表面双目立体视觉三维形貌测量方法 下载: 2084次
Measurement Method of Three-Dimensional Shape of Bright Surface with Binocular Stereo Vision
测量 双目立体视觉 三维形貌测量 匹配点搜索 相位偏折 位置标定 measurement binocular stereo vision three-dimensional shape measurement matching point search phase shift deflection position calibration
摘要
光亮表面因其反射特性,一般三维形貌测量方法对此难以测量,针对该问题,本文给出了基于双目视觉结合相位偏折法对光亮表面进行三维形貌测量的方案。双目系统布局选用相机横向摆放方式,完整的屏幕-相机-可调节载物台测量系统被集成在定制框架内。对相移法中存在的非线性相位误差进行校正,在主值相位图内进行反向相位误差补偿,提高解包裹精度,为减小标定误差,将系统标定得到的位置参数使用Levenberg-Marquardt算法优化。结合光亮表面法向量唯一性和相机的极线约束提高匹配点搜索效率,对传统三角法求空间点进行改进,提高待测物表面点求取准确性,实验结果验证了所提方案具有较高的测量精度和稳定性。
Abstract
Due to the reflection characteristics of the bright surface, the general three-dimensional shape measurement method are difficult to measure.To solve this problem,a scheme based on binocular vision and phase deflection method to measure the three-dimensional shape of bright surface is proposed. The binocular system layout uses the camera horizontal placement, and the complete screen-camera-adjustable platform measurement system is integrated into a custom frame.In order to reduce the calibration error, the Levenberg-Marquardt algorithm is used to optimize the position parameters obtained from the system calibration. The matching point search efficiency is improved by combining the uniqueness of the bright surface normal vector with the polar constraint of the camera.The traditional trigonometry method is improved to find the space points and the accuracy of the surface points of the object to be measured is improved. The experimental results verify that the proposed method has high measurement accuracy and stability.
程子怡, 卢荣胜, 毛翠丽. 光亮表面双目立体视觉三维形貌测量方法[J]. 激光与光电子学进展, 2020, 57(7): 071202. Ziyi Cheng, Rongsheng Lu, Cuili Mao. Measurement Method of Three-Dimensional Shape of Bright Surface with Binocular Stereo Vision[J]. Laser & Optoelectronics Progress, 2020, 57(7): 071202.