Vehicle detection in close-up range with a fisheye camera

摘要

In modern cars, driver assistance systems support the motorist to handle the ever increasing complexity of therndriving task. These systems aid to avoid accidents and alleviate their effects, thus reducing the number of casualties. Tornreduce the amount of rear-end collisions the ACC (Adaptive-Cruise-Control) was introduced a few years ago. It takes overrnthe longitudinal guiding, by controlling the distance to the vehicle ahead. If the driving-lane is free, the car accelerates to thernchosen velocity. The system uses radar-sensors, monitoring an area of up to 200 meters in front of the vehicle. There are twornproblems facing system, which use solely radar-sensors: With a narrow angle of beam of typically sixteen degrees objects inrnclose distance can only be detected right in front of the sensor. In addition radar sensors cannot distinguish different kinds ofrnobjects. Therefore no object-specialized vehicle behaviour can be implemented. A supporting fisheye camera solves thesernproblems. Via specialized algorithms all kinds of objects are distinguishable.rnIn a visual driver-assistance system, obstacle detection during driving is one of the major tasks. This paper presents a newrnvehicle detection method based on multi-features fusion in the images acquired by a fisheye camera. The vehicle detectionrnalgorithm can be divided into three main steps: fisheye image calibration, generation of candidates with respect to a vehiclernand verification of the candidates. In this paper vehicles are detected through typical features of their front and rearrnperspective, like shadow and symmetry. The use of a fisheye camera with an angle of beam exceeding 180° allows therndetection of objects not only for ACC or forward collision warning systems, but also enables close-up lane and pedestrianrndetection, as well as sensing overtaking cars earlier.rnIn a first step the car detection capabilities of similar systems using non-wide-angle cameras shall be reached. Due to thernlower resolution in the central part of the camera picture the distance of detection is expected to be smaller. The vehiclerndetection was accomplished using the well established concept of hypothesis generation and verification. Whereby a largerrnnumber of vehicle hypotheses is generated in the first and filtered in the second step.rnThe system was tested in various weather and road conditions. Experimental results in different conditions, including sunny,rnrainy, snowy demonstrates that most vehicles can be detected and recognized with a high accuracy and a frame rate ofrnapproximately 16 frames per second on a standard PC.