3D camera technology trade-off and breadboard demonstration for space applications

Abstract

In previous decades, the vision-based navigation problem based on 2D imaging has been largely studied and applied in space, for rendezvous and docking, as well as rover navigation, or entry, descent and landing. By providing measurement of the third dimension (range), 3D camera technology looks a promising alternative for many applications. Stereoscopic camera is one option to measure the third coordinate, but relies on significant CPU capabilities, which are generally not available for space applications. Scanning LIDAR (LIght detection and ranging) is also an existing solution, but it is relatively large and heavy and the refresh rate, lifetime and reliability are mainly determined by moving parts. 3D time-of-flight (TOF) technology (including flash LIDARs) offers a reliable alternative. By illuminating a whole scene at a time and thus providing a whole array image, there is no need for complex processing nor moving mechanisms, which clearly appears as an advantage for space applications. This paper presents the ongoing study conducted under ESA contract in the field of 3D TOF technology. Its goal is to evaluate the suitability of a 3D TOF camera for space applications, to derive requirements and a preliminary design, and finally to create and test a breadboard model. Performance budget, cost, and a development plan of a versatile spatialized 3D TOF camera are also outputs of the study, in addition to a high-fidelity simulator, allowing further studies by generating representative images and depth maps. To fulfill this project, a European team has been created, gathering Thales Alenia Space, Terma and SINTEF.