基于预标定彩色散斑光场照明的三维轮廓测量技术研究

Color speckle is generated by random dammann grating，which is illuminated simultaneously by lasers with different wavelengths,and the pre-calibration of the color speckle is utilized in the research of three-dimensional profile measurement. Research a novel 3D profile measurement technique based on pre-calibrated color speckle illumination. Analysis the relationship among the regular Dammann array dots, camera calibration accuracy,and image matching accuracy,then optimize and fabricate random Dammann grating. Designed the multi-beam illumination system based on the random Dammann grating, and establish the experimental system to measure 3D profile by binocular vision and to obtain valuable results. Study the learning and pre-calibrating method of feature descriptor of color speckle based on standard sample. Study the growth matching method of neighborhood points based on the regional characteristic. Study the sub-pixel global refine method based on pre-calibrated color speckle feafure library. This is an innovative research by applying the diffractive optical element into computer vision, and this interdisciplinary research will bring a lot of valuable topics. The equipment based on this three-dimensional profile measurement technique has advantages of no mechanical limit, low-cost, simple structure, anti-interference, high resolution, high speed, etc. This equipment also has broad application prospects, such as reverse engineering, advanced manufacturing, cultural relics protection, and digital entertainment.

提出了利用多路激光照射随机达曼光栅产生彩色散斑光场，并将预标定彩色散斑光场应用到物体三维轮廓测量研究中。研究一种全新的基于预标定彩色散斑光场照明的三维轮廓测量技术。分析规则达曼点阵、彩色散斑特征、图像匹配精度的三者相互关系,以此优化设计并制作随机达曼光栅。设计随机达曼光栅多路激光照明系统，建立物体轮廓测量实验装置并获得有价值的实验成果。研究彩色散斑特征描述子标准样本学习与预标定方法。研究基于区域特征的规则点邻域随机点生长匹配方法。研究基于预标定散斑特征库的全局亚像素匹配求精方法。 该研究是将衍射光学元件应用于计算机视觉的一个创新性研究，此学科交叉研究会带来很多有价值的研究课题。基于该三维轮廓测量技术的设备具有无机械约束、成本低廉、结构简单、抗干扰、分辨率高、测量速度快等优点，在逆向工程、先进制造、文物保护、数字娱乐等领域也有着广泛的应用前景。

3D测量技术非常重要，在各个领域有广泛应用。尤其是面向工业机器人的3D测量设备对稳定性及精度有很高要求。本项目发明了基于散斑原理的3D测量技术。通过引入预标定散斑测量技术，改进了具有复杂反射特性的物体的3D测量精度，从而使所发明的技术可以测量各种材料的工业部件。通过附加二值宽带条纹，解决了散斑快速配准及单帧点云边界完整性问题。结合时域散斑序列的采集对复杂拓扑高度敏感的特点，使得所发明技术具备复杂拓扑结构的测量能力。通过锚点区域生长技术，改进了匹配速度和准确度。基于本项目资助发明的3D测量技术具备成本低廉、结构简单、抗干扰、分辨率高、测量速度快等优点，在逆向工程、先进制造等领域有着广泛的应用前景。

内容获取失败，请点击重试