面向复杂型面的成像前光学调制三维形貌测量方法研究

Complex surface is a basic component of high-end equipment. Study of three-dimensional shape measurement for complex surfaces is very demanding for the development of national high-end equipment manufacturing technology. Major drawbacks of current three-dimensional shape measurement technologies include large curvature surface due to the spatial frequency aliasing is uncertainty, high reflective surface due to strong light interfere is unavailable, and the dynamic measurement accuracy is far less than static measurement accuracy. To address these problems, this project presents a three-dimensional shape measurement method for complex surface based on optical modulation before image formation. The main research contents are listed as follows: Firstly, research on spatial frequency multiplexing measurement method for large curvature surface is conducted. Spatial sample frequency is quantitatively selected to enhance the phase sample accuracy of the grating projection image in the steep gradient change area. secondly, strong light removal method for high reflective surface is developed. Self-adaptive high dynamic range imaging system is designed to enhance the quantitative accuracy of the grating stripe, and programmable spectral imaging system is proposed to remove stray light noise. thirdly, we carry out research on dynamic three-dimensional shape measurement method based on snapshot coded multispectral imaging. By sampling multiple grating stripe images with phase shift deviation and resolving the modulated phase in single exposure time, high accuracy measurement of dynamic surfaces can be achieved. This project would focus on the current major difficulties in three-dimensional shape measurement for complex surfaces, and will aim at improving the development of national high-end equipment manufacturing technology.

复杂型面工件作为高端装备的基本构件，对其进行三维形貌测量方法研究已成为了我国装备制造技术发展的重大需求。目前面向复杂型面的三维形貌测量存在大曲率面形空间频率混叠测不准、高反射材料面形强光干扰测不出，以及动态测量精度远低于静态测量问题。为了解决以上三个问题，结合成像前光学调制技术优势，本项目提出了面向复杂型面的成像前光学调制三维形貌测量方法，主要内容包括：研究大曲率面形空间频率复用测量方法，参数化选择空间采样频率，提高光栅投影图像梯度突变处的相位测量精度；研究高反射面形强光去噪方法，构建自适应高动态范围成像系统提高对结构光栅条纹成像的量化精度，构建可调谐光谱成像系统滤除环境光噪声；研究多波段单拍编码动态三维形貌测量方法，在单次曝光中完成多幅移相条纹光栅图像的采集和相位解算，实现高精度动态三维形貌测量。本项目研究将解决当前复杂型面三维形貌测量中存在的主要难题，促进我国高端装备制造技术的发展。

复杂型面三维测量面临着高反射强光干扰测不出、大曲率面形测不准和动态测量精度低等问题。针对以上问题，本项目开展了以成像前光学调制为核心技术的新型成像系统的三维测量方法研究。首先，构建了基于DMD的多视点反射式成像模型，分析了成像特性和成像质量，并应用在核科学成像领域。在反射式成像机制的成像前光学调制基础上，明确了不同反射率、粗糙度的被测表面对结构光的镜面反射机理，采用低秩矩阵对高反射强光表面进行稀疏表达，实现被测目标在高光环境中成像，并应用在印制电路板加工工序质量检测中。成功设计了一套自适应高动态范围三维形貌测量装置，通过亮条纹饱和区光强调制达到对结构光的重建和相位补全，完成高反射强光表面三维测量。成功设计了一套PZT闭环微位移平台调控CCD与DMD的配准方式，根据目标特征的成像需求主动提高目标二维空间采样频率，测量矩阵由条纹相位图的方向梯度函数确定，压缩感知重建空间频率复用相位图像以实现大曲率面形三维测量，并应用在电机铜排加工检测领域。根据各图像块的信息熵的大小对各图像块进行分组处理，给各图像块自适应的分配采样率并设置测量矩阵进行压缩采样，利用最大熵准则光谱分段，选取合适的采样值局部压缩感知重构光谱反射率，完成了多光谱数据编码压缩高精度重建。最后，结构光投影子系统等间隔选取可见光谱光线N条波段作为投影光源，通过选取空间采样单元降低多波段数据的混叠，光谱波段稀疏采样消除了相邻通道数据的串扰，实现多波段通道动态三维测量。本项目是在负责人与多方专家学者交流下完成的，其相关研究成果正在电机制备、集成电路，核科学成像等工业检测领域进行转化和应用。截至2021年底，本项目资助发表期刊论文5篇，其中SCI检索3篇，全部EI检索，授权发明专利1项，实用新型2项，获得省技术发明二等奖1项，培养了4名硕士毕业。

内容获取失败，请点击重试