In this paper, a topographic data stitching method is firstly introduced for compensating and optimizing the tilt errors derived from the lateral guide of the stage of a confocal laser scanning microscope. Afterwards, a novel evaluation process, including both simulated verification and experimental verification, is proposed to analyze the stitching precision and its applicability. At last, the evaluation is conducted with multiscale measured datasets, using five typical engineering surfaces with different surface textures. The results show that the developed method can compensate for the tilt uncertainty without the help of the precision apparatus and exhibits improved performance in comparison with conventional methods. In addition, the results of the comprehensive verification show the stitching precision is mainly influenced by concrete surface textures, surface roughness, acquired data resolution as well as the magnitude of tilt uncertainties.

本文首先介绍了一种地形数据拼接方法，用于补偿和优化从共聚焦激光扫描显微镜镜台的侧向引导所产生的倾斜误差。然后，提出了一种新颖的评估方法，包括模拟验证和实验验证，以分析缝合精度及其适用性。最后，使用五个具有不同表面纹理的典型工程表面，使用多尺度测量数据集进行评估。结果表明，与传统方法相比，所开发的方法无需精密仪器就可以补偿倾斜不确定性，并具有改进的性能。此外，