Wang Yuezong, Zhao Zhizhong
College of Mechanical Engineering and Applied Electronics Technology, Beijing University of Technology, Beijing 100124, China.
College of Mechanical Engineering and Applied Electronics Technology, Beijing University of Technology, Beijing 100124, China.
Micron. 2019 Jan;116:46-53. doi: 10.1016/j.micron.2018.09.011. Epub 2018 Sep 27.
Microscopic stereovision systems with a stereo light microscope have been used for quantized observations and measurements in microscopy fields. These applications depend on a quantization mathematical relationship between the object space and the image space. In this article, a novel, practical space quantization method was proposed. It was derived from two defined base cell matrices. First, it captured a series of stereo images of markers moving in various directions through the object space. The world and image coordinates of markers were acquired, stored and used to create the two base cell matrices. Next, a space quantization relationship between the object space and the image space was derived from analysis of the two base cell matrices. Finally, if two matching points located in the image space were given, the world coordinates of their object point could be calculated by the above space quantization relationship. The results show that this method of quantizing the two spaces could achieve a positioning precision of approximately ±5 μm in the horizontal direction and approximately ±10 μm in the vertical direction. It has a very simple expression, but it is more practical and efficient than other traditional projection methods. In addition, it has a better performance and can be widely used in several microscopy fields, such as microscopic shape reconstruction, micromanipulation, microassembly, and microinjection.
带有立体光学显微镜的微观立体视觉系统已被用于显微镜领域的量化观察和测量。这些应用依赖于物空间和像空间之间的量化数学关系。在本文中，提出了一种新颖、实用的空间量化方法。它源自两个定义的基元矩阵。首先，它捕捉了一系列标记物在物空间中沿不同方向移动的立体图像。获取、存储标记物的世界坐标和图像坐标，并用于创建两个基元矩阵。接下来，通过对两个基元矩阵的分析得出物空间和像空间之间的空间量化关系。最后，如果给出像空间中的两个匹配点，其物点的世界坐标可通过上述空间量化关系计算得出。结果表明，这种对两个空间进行量化的方法在水平方向上可实现约±5μm的定位精度，在垂直方向上约为±10μm。它具有非常简单的表达式，但比其他传统投影方法更实用、高效。此外，它具有更好的性能，可广泛应用于多个显微镜领域，如微观形状重建、微操作、微装配和微注射。
