Dynamic Obstacles Avoidance Based on Image-based Dynamic Window Approach for Human-Vehicle Interaction

This paper presents an approach for the development of Advanced Driving Assistance System (ADAS) based on the human-vehicle interaction using Image-based Dynamic Window Approach (IDWA). The IDWA is associated to a method for dynamic obstacles avoidance in order to prevent human driving errors, in the context of intelligent robotic vehicles. The human-vehicle interaction is presented by the correction of the Human Driving Behavior (HDB) controller for driving defaults of human drivers, with respect to referential paths that intimate the average driving path in real circumstances. The performance of the proposed human-vehicle interaction methodology, based on autonomous embedded functionalities, is simulated and verified in different bypass scenarios.