PMD-Modeling, -Simulation, -Evaluation & Algorithmics

距离传感飞行时间(ToF)相机已经活跃地发展了大约15年。光子混合器件(PMD)是最早由西根大学的鲁道夫·施瓦特教授介绍并实现的方法之一。PMD概念是PMD Technologies GmbH的核心，PMD Technologies GmbH是世界领先的ToF相机生产商之一，拥有约65名员工。所请求的知识转移项目的中心思想是基于计算机图形学主席和工业合作伙伴PMD Technologies GmbH各自的发展状况，对TOF相机的传感器建模、仿真、评估和算法主题进行集成调查和概念集成。这一方法将促进过去十年提出的相关研究成果的进展，使其融入用于现实世界应用的商业产品。此外，我们将开发新的解决方案来解决ToF相机系统的扰动效应，到目前为止，这方面的研究还很少。在这个项目中，我们将在DFG以前和现在资助的研究项目的基础上，研究非常新的和实际的问题，同时，这些问题具有强烈的科学兴趣。这涉及传感器建模、模拟和评估的综合研究，允许对未来的传感器和算法进行定量评估，从芯片级到驱动器级的特定测量误差校正。此外，在算法方面，我们还将研究运动伪影、多路径问题等相关问题。找到稳健、高效的解决方案将有力地提高数据质量，从而为TOF相机开辟新的应用领域。

Range sensing Time-of-Flight (ToF) cameras have been actively developed for some 15 years now. The Photonic Mixing Devices (PMD) is one of the first approaches that was introduced and prototypically realized by Prof. Dr. Rudolf Schwarte from the University of Siegen. The PMD concept was the nucleus for the foundation of the pmd technologies gmbh, which is one of the world leading ToF camera producers, with a staff of about 65 employees.The central idea of the requested knowledge transfer project is the integrated investigation and conceptual integration of the topics sensor modeling, simulation, evaluation and algorithmics of ToF cameras based on the respective development status at the chair for Computer Graphics and the industrial partner pmd technologies gmbh. This approach will promote the advancement of relevant research results, presented in the last decade, towards integration into commercial products for real-world applications. Furthermore, we will develop new solution to perturbing effects of ToF camera systems, which have been little investigated until now.In this project we will, on the basis of prior and current DFG funded research projects, investigate very current and practical questions, which are, at the same time, of strong scientific interest. This involves an integrated investigation of sensor modeling, simulation and evaluation, which allows for quantitative assessment of future sensors and algorithms, from the chip level up to the correction of specific measurement errors at driver level. Furthermore, in the algorithmic context, we will research relevant questions such as motion artefacts and multi-path problems. Finding robust and efficient solutions to resolve these problems will strongly improve the data quality, and thus open up new application domains for ToF cameras.