PAROTO project: the benefit of infrared imagery for obstacle avoidance

Most road accidents are due to a lack of perception of the driver about the road situation. The PAROTO system´s objective is to help the driver in detecting obstacles and thus allow him a better and earlier reaction in dangerous situations. Data coming from radar and an infrared camera are merged in order to take advantage of complementary information. Eventually the driver can be alerted after a filtering step. This step consists in deciding which information is relevant for the driver amongst the results of the sensors analysis. We present a segmentation technique using the particular content of infrared images. We also present a motion analysis approach, based on features. Both processes lead to a first set of promising results.