Reflectance Analysis and Synthesis using the Tapered Kaleidoscope

2003年，首席研究员和他的合作者介绍了一种新技术，用于测量从不同方向观察或照明时纹理物体的外观。 该技术不需要机械运动，并且非常容易校准，基于使用锥形万花筒：多边形横截面的锥形中空管，其内壁衬有前表面反射镜。 这种布置用于模拟大量摄像机和投影仪的效果，所有摄像机和投影仪都从不同方向指向目标对象。这项技术有望改善测量和显示在许多领域的接地计算机生成的图像更彻底地在现实世界的数据，并使捕获手段更容易获得更多的研究人员。 该技术也有值得法医使用，如子弹和指纹识别，皮肤黑色素瘤的早期检测，以及模拟材料测量。 利用万花筒原理创建光场传感器的早期工作也有巨大的潜在好处，它有可能创建一种新型的三维显示器，显示出从根本上改善的角度特性。由于这种数据捕获方法不会引入机械误差，因此它有望实现对依赖于精确校准的多视图测量的问题进行调查的新领域，多照明和多光谱光学特性。 研究任务包括数学建模和分析以及以下领域的实际实验室实验和数据收集：捕获八维表面BSSRDF，捕获3D物体的全光场，使用可调波长滤波器捕获光场的全光谱，真实的实时多视图结构光深度扫描，通过组合使用多个相机和窄锥角来增加角分辨率，有效地利用图像片段，从稀疏和非矩形样本合成，以及对极端广角光场传感器的实验。

In 2003, the Principal Investigator and his collaborators introduced a new technique for measuring the appearance of a textured object as it is viewed or illuminated from different directions. The technique, which requires no mechanical movement and is very easy to calibrate, is based on use of a tapered kaleidoscope: a tapered hollow tube of polygonal cross-section, whose inner walls are lined with front-surface mirrors. This arrangement is used to simulate the effect of a large array of cameras and projectors, all pointing toward the target object from different directions. This technology promises to improve measurement and display in many areas by grounding computer generated imagery more thoroughly in real-world data and by making the means of capture more readily available to a greater number of researchers. The technology also has worthy forensic uses, such as bullet and fingerprint identification, and early detection of skin melanomas, as well as materials measurements for simulation. There is also enormous potential benefit from proposed earlier work on using the kaleidoscope principle to create a lightfield transducer, which has potential for creating a new type of three dimensional display exhibiting fundamentally improved angular characteristics.Because this approach to data capture introduces no mechanical error, it promises to enable a new field of inquiry into questions that rely on precisely calibrated measurement of multi-view, multi-illuminated, and multi-spectral optical characteristics. Research tasks include mathematical modeling and analysis as well as practical lab-bench experimentation and data collection in the following areas: Capturing eight dimensional surface BSSRDF, capturing full lightfield for 3D objects, capturing full spectrum of lightfield with tunablewavelength filters, real time multi-view structured light depth scanning, increasing angular resolution through combined use of multiple cameras and narrow taper angles, making effective use of image fragments, synthesis from sparse and non-rectangularsamples, and experiments toward extreme wide-angle lightfieldtransducers.