OpticFusion
收藏资源简介:
OpticFusion数据集由浙江大学CAD&CG国家重点实验室创建,旨在通过融合白光干涉仪(WLI)和光学显微镜(OM)的多模态数据,实现微结构的三维重建与自然色彩纹理的获取。该数据集包含多种微尺度样本的多视角WLI和OM图像,样本表面特征尺寸从数十微米到数百微米不等,涵盖蝴蝶翅膀、花种、电路板和微传感器等。数据集通过两阶段数据关联方法获取WLI和OM图像的姿态,并利用神经隐式表示技术进行多模态数据融合,最终生成具有自然色彩纹理的微结构三维模型。该数据集的应用领域包括微尺度表面粗糙度测量和生物样本的详细分析,为微尺度研究提供了有效的工具。
The OpticFusion dataset was developed by the State Key Laboratory of CAD&CG, Zhejiang University. It is designed to achieve 3D reconstruction of microstructures and acquisition of natural color textures by fusing multimodal data from White Light Interferometers (WLI) and Optical Microscopes (OM). This dataset includes multi-view WLI and OM images of various micro-scale samples, whose surface feature sizes range from tens to hundreds of micrometers, covering specimens such as butterfly wings, flower seeds, circuit boards, and microsensors. The dataset obtains the poses of WLI and OM images via a two-stage data association method, and leverages neural implicit representation technology to fuse multimodal data, ultimately generating 3D models of microstructures with natural color textures. Its application scenarios include micro-scale surface roughness measurement and detailed analysis of biological samples, providing an effective tool for micro-scale research.

- 1OpticFusion: Multi-Modal Neural Implicit 3D Reconstruction of Microstructures by Fusing White Light Interferometry and Optical Microscopy浙江大学CAD&CG国家重点实验室 · 2025年



