Fusion of Perception and V2P Communication Systems for the Safety of Vulnerable Road Users
The performance of environment perception sensors plays an important roll for the effectiveness of advanced driver assistance systems (ADAS). This paper analyzes different methods for sensor model generation for effectiveness simulations and derives a new sensor model unifying the main advantages of common deterministic and stochastic approaches. As an example, the paper utilizes the new sensor model to investigate the benefits of cooperative sensor systems for pedestrian safety. Cooperative sensors are a relatively new technology which allows for unique object classification and localization regardless of a sight obstruction. To generate a model, real world measurements are collected with a cooperative sensor system operating at 2.4 GHz. Finally, effectiveness simulations are performed to analyze the benefit of cooperative sensor systems in comparison to a single sensor system based on a monocular camera and a high-end fusion system with respect to a set of pedestrian accidents.