基于二值编码条纹虚拟相位解调的微器件3D形貌检测方法研究

On-line nondestructive topography testing an important guarantee for the quality control of large -scale micro-devices processing.In recent years, structural light micro-structure measurement technology is an important research direction of micro-device topography testing. The latest research results improve the measurement accuracy by improving the measurement model and reconstruction method, but it is still difficult to meet the requirements of online testing in terms of measurement efficiency and reliability.Our previous studies have found that binary encoding and phase modulation and demodulation of the optical field can greatly improve the measurement efficiency and reliability while guaranteeing the measurement accuracy.In this project, a method for on-line testing of 3D topography of micro-devices based on virtual phase demodulation of binary coding fringes is proposed by using the double distal optical microscopy measurement model. The main research contents include: 1) design and optimization of optical field binary encoding;2) aiming at the limited depth of field, the calibration model and method based on the phase feature points of the orthogonal fringes are studied;3) study the construction of virtual phase plane and the method of fast phase analysis to realize the high-precision testing of 3D topography of micro-device.The research will provide a support for the industrial application of micro-device online topography measurement method based on micro-structure light field projection, and lay a foundation for the next independent research and development of digital intelligent 3D topography testing equipment for micro-device in China.

在线无损形貌检测是大规模微器件加工品控的重要保证。结构光显微形貌测量技术是近年来微器件形貌检测的重要研究方向，最新的研究成果不断改进测量模型和重建方法，提高测量精度，但在测量效率和可靠性等方面仍然难以满足在线检测的要求。我们的前期研究发现，通过对光场的二值编码和相位调制解调，可以在保证测量精度的同时，大幅提高测量效率和可靠性。本项目利用双远心光学显微形貌测量模型，提出基于二值编码条纹虚拟相位解调的微器件3D形貌在线检测的方法，主要研究内容包括：1）光场二值编码设计及优化；2）针对有限测量景深，研究正交条纹相位特征点的标定模型及方法；3）研究虚拟相位平面的构建略及引导相位快速解析的方法，进而实现微器件3D形貌高精度检测。本项目的研究将为显微结构光场投影的微器件在线形貌测量方法走向产业应用提供支撑，为我国下一步微器件数字化智能化3D形貌检测设备的自主研发奠定基础。

微器件是指外观尺寸在毫米级、特征尺寸在微米量级的功能性器件，在多个领域有着巨大的应用价值。在线无损形貌检测是大规模微器件加工品控的重要保证。结构光显微形貌测量是利用器件表面形貌对结构光场的相位调制，通过相位解析实现三维形貌的高精度检测，非常适合复杂环境微器件三维形貌的高精度在线检测，但在在测量效率和可靠性等方面仍然难以满足在线检测的要求。在本项目的资助下，开展了基于二值编码条纹虚拟相位解调的微器件3D形貌在线检测的方法的研究。主要研究工作包括：1）充分考虑3D形貌重建过程相位解析的特点，提出了基于遗传算法的光场二值编码设计及优化策略，显著提升了二值编码光场的正弦性；2）针对标定过程中，没有适当约束的情况下会导致优化结果与实际情况不相符，严重影响测量精度的问题，提出结构光显微形貌测量的双远心摄像机标定优化方法，重投影误差降至0.03像素；3）搭建了3D形貌实验系统，测量对象尺寸在几毫米-几十毫米范围，精度可达0.005mm。本项目的研究成果为航空工业和医疗机构等单位提供了微器件3D形貌检测服务，取得了较好的应用反馈。.培养研究生7名（含博士生2名）、论文8篇（SCI论文6篇）和国家发明专利9项。. 培养研究生7名（含博士生2名）、论文8篇（SCI论文6篇）和国家发明专利9项。