Regresar

Robot vision for intelligent autonomous vehicles

Abstract:

Road traffic injuries are a major global public health problem, requiring concerted efforts for effective and sustainable prevention. The worldwide consequences are devastating: The number of people killed in road traffic crashes each year is estimated at almost 1.2 million and injuring or disabling between 20 million and 50 million more. In economic terms, the cost of road crash injuries is estimated at roughly 1% of gross national product (GNP) in low-income countries, 1.5% in middle-income countries and 2% in high-income countries. Among others, human errors are the cause of most of traffic accidents, being driver's inattention and wrong driving decision the two main sources of errors. Due to their very nature, these mistakes can be attenuated with infrastructure improvement or reduced with educational campaigns, but not completely eliminated. Because of this, the research on security was firstly done on a mechanical level to later introduce internal sensors and actuators to improve vehicle control like Antilock Brake System (ABS) or Electronic Stability Program (ESP). It is interesting to note that something similar happened before within the field of Industrial Robotics, where at the beginning the stress was in the use of internal sensors like encoders in order to improve speed and position control of the robot and now the goal is to perceive the environment, being visual servoing a good example of that. Next, a further step in order to increase road security is the introduction in the vehicles of sensors able to analyze the environment. Robot Vision is playing a crucial role in that as it will be shown in this chapter where many techniques and approaches used in robotics applications are translated to the road environments. These perception-based systems receive the name of Advanced Driver Assistance Systems (ADASs) and the research on some of them is presented here in the framework of a research platform for the implementation of systems based on computer vision: Lane Keeping System to perceive the distance to the lane borders and the time to cross them; Adaptive Cruise Control that perceives the surrounding of the vehicle looking for others and calculating trajectories and possible conflicts; Pedestrian Protector to detect the most vulnerable users in road environment; Speed Supervisor that checks the vehicle's speed is correct taking into account the detected traffic signs, and Driver Monitoring which detect the attention status of the driver and warns when needed. These robotic technologies will be able to reduce the number, danger and severity of traffic accidents.