基于显微视觉的微纳米尺度三维动态形貌观测

Research on visual, high resolution, and three dimensional static/dynamic observation on micro/nano scale is a great challenge in micro/nano science and technology. Since micro vision imaging has many advantages, such as real time imaging,visuality, none destruction and so on., it is widely used in two dimensional observation on micro scale. However, due to diffraction and low resolution, micro vision technology is hardly used in three dimentional observation on micro/nano scale.Therefore,in order to improve micro/nano observation technology and its relevant science and technology, it has great significance to develop a real time, visual, dynamic and three dimensional observation technology based on micro vision. In this proposal, we plan to research on blurred measurment models of defocus and diffraction under the depth of focus of micro scale; research on the relationship between the burred measurment and the depth of samples, and propose a depth optimization inverse solution and three dimensional reconstruction method based on the burred measurment; research on precision improvement method based on the enhanced scattering model, and reveal the principle of three dimensional dynamic shape reconstruction based on blurred measurment on micro/nano scale. Finally, we plan to validate our algorithms with standard samples ( for example nano grids) and high resolution microscopies, and reconstruct their three dimensional shapes with a resolution of 50-100 nm. On the base of it, we will propose an three dimensional dynamic reconstruction method with an image sequence. In a word, with the research results in this proposal, we will provide a new effective three dimensional observation method for micro/nano science and technology, and it is has great significance to improve micro/nano observation technology and its relevant science and technology.

微纳米尺度的直观、高分辨率三维静/动态观测是当前微纳米科技面临的挑战性问题。由于具有直观、实时、无损伤等优点，显微光学成像已广泛用于微观尺度二维观测，但衍射和分辨率等问题的存在使得其尚无法应用于微纳米尺度的三维观测应用中。针对此问题，本申请将基于显微光学成像机理，研究微景深条件下的光学成像离焦与衍射模糊测度建模方法；研究图像模糊测度与景物深度的关系，提出基于模糊测度的深度信息反解与三维重构方法；研究基于扩射模糊增强模型的深度灵敏性提升方法，并开展反解算法优化研究，以达到微纳米尺度下基于模糊测度的景物三维形貌重构。此外，本研究将以标准样品（纳米栅格等）为观测对象，开展相关实验研究，实现50-100纳米精度的景物三维重构，并在此基础上，实现基于序列图像的形貌变化三维动态恢复。本研究将为微纳米科技研究提供一种新的、有效的三维观测方法，对促进微纳米观测技术和相关领域科技的发展具有重要意义。

微纳米尺度下直观、高分辨率三维静/动态观测是当前微纳米科技面临的挑战性问题。由于显微光学成像具有直观、实时、无损伤等优点，光学显微镜一直都是实现精确二维观测的主要手段。但是，传统的光学显微镜以及改进的混合光学显微镜都遵循几何光学成像原理，由于光学衍射的影响，当前的光学显微镜均存在分辨率的“阿贝极限”：小于200 纳米的物体不能被清晰成像。正因如此，微纳米尺度下的光学成像及观测技术一直没有取得突破性进展。针对此问题，本项目通过深入探索微纳米尺度下基于光学显微系统的高精度动态三维视觉观测问题，提出了微观光学显微系统中光学衍射模糊测度与建模方法、基于模糊测度的深度信息反解与三维重构方法、基于扩散模糊增强模型的灵敏度提升机理和方法。在此基础上，开展了相应的深度反解优化算法研究和实验研究。项目实现了基于光学菲涅尔衍射模糊测度的景物三维形貌动态重构、基于离焦/衍射模糊的动态精准定位，为微纳米科学技术研究提供一种有效的高精度三维视觉动态观测方法。最后，本项目的研究成果均已经得到了充分的实验验证：纳米栅格形貌恢复实验和原子力显微镜的悬臂梁形变恢复实验验证了本项目提出的静态固定成像波长时形貌恢复方法，三维恢复精度最高可以达到100纳米；微透镜的三维形貌重建实验验证了基于光学菲涅尔衍射的微纳米尺度模糊测度模型，所获得的横切面误差最小为70纳米。同时，为了提高深度重建算法的灵敏性和精度，本项目提出了一种全局散射模糊深度获取方法，以达到小透镜情况下提高深度获取方法精度和灵敏度的目的。本研究的研究成果为微纳米科技研究提供了一种新的、有效的三维观测方法，对促进微纳米观测技术和相关生物、医学等领域科技的发展具有重要意义。