Seite - 18 - in Integration of Advanced Driver Assistance Systems on Full-Vehicle Level - Parametrization of an Adaptive Cruise Control System Based on Test Drives

2. Adaptive Cruise Control object lens imager picture xs S rdld object left camera right camera d xs (b)(a) Sys zs Figure 2.10.: (a)Projectionofanobjectontotheimagerand(b)stereoprinciple,adapted from [SBD09] Stereo Camera Instereocameras systems, twocamerasaremountedatadefineddistanced, seefig.2.10(b). The picture of one object is not at the same place on the two imagers. There is a lateral offset of ∆d=dl−dr, which is called a disparity. So for every pixel on the imager, the depth information sx can be found because ∆d∼ sx. Using this so-called range map, the feature extraction is performed to find the objects in the picture. Due to the large number of pixels in a picture, this process requires a high computation power, [SBD09]. Stein et al. showed in [SGS10] that it is possible to reduce the required computation power if the order of the signal processing process is changed. In the Stereo-Assist ap- proach, they use two mono cameras which perform the feature extraction. Afterwards, the stereo principle is used to find the distance for every object instead of for every pixel, which significantly reduces the number of operations. 2.2.3. LIDAR Sensor The Light Detection and Ranging (LIDAR) sensor is an optical sensor that radiates electromagnetic waves, with a wave length of about 850 to 1000nm. This is in the area of infrared light on the border to visible light. If too much power is emitted, human eyes willbehurt. Therefore, the impulsehastobeveryshort. Forautomotiveapplications,an impulse lengthofabout30ns is commonlyused. Therangemeasurement isperformedas described ineq. (2.1) for thePulseDopplerRADAR.ForLIDARsensors, it ispossible to use the Doppler Effect to measure the relative speed, but it is too expensive for vehicles. Therefore, r˙ is estimated with the difference quotient r˙= r2−r1t2−t1 , where r2−r1 describes the change of the distance between two measurements, and t2− t1 describes the time between the measurements. The angle can be measured in different ways. For the fixed multi-beam LIDAR, the sensor emits different beams at fixed position, see fig. 2.11(a). The angular resolution of this type corresponds roughly to the width of the beam. The 18