Multi-ToF-Plattform von Becom kombiniert vier Sensoren
Becom's Multi-ToF platform is an ecosystem for parallel ToF sensors that allows up to four different sensors to be integrated and combined in different constellations.
The central processing unit is based on the Deep-Learning capable Nvidia-Jetson-TX2 module with Pascal-GPU as well as Dual-Core Nvidia Denver 2 and Quad-Core CPUs ARM Cortex-A5. In principle, any imaging sensor system can be integrated. The platform can thus be used in human-machine interaction, collaborative robot interaction, automotive applications and obstacle detection for autonomous systems. The platform consists of the ToF-Hub and the ToF-Frontend. The hub hosts connections for four sensor front ends. The front end accommodates the lighting and the sensor chip. Both components are connected via a two-wire serial connection, which also provides the power supply. The first ToF sensor front end is a module for close-range detection and has a 110-degree wide-angle lens and a sensor with 320 x 240 pixel resolution. For example, the front ends in lightweight robots or grippers enable a 360-degree view.