Various forms of night vision systems, using a variety of sensing technologies, are being developed to further reduce the risk of drivig at night. Two major sensing technologies are receiving particular development interest and are both currently available on new vehicles: near infrared (NIR) systems, which actively illuminate the scene in the near infrared spectrum and capture the reflected radiation, and far infrared (FIR) systems, which generate images by passively detecting thermal emissions from objects and surfaces in the road scene. In a recent laboratory study of two such systems it is stated that detection distances of pedestrians were substantially greater with an FIR system than with NIR for each of 30 road/pedestrian combinations tested. On average, detection distances were 165 m with FIR and 59 rn with NIR. In this paper the authors demonstrate, that visual clutter was a primary contributing factor for the difference in detection distance between FIR and NIR systems. Visual clutter is associated with the nontarget information in an image. An inherent attribute of FIR images is that such nontarget information is usually minimal because there are not many objects in the environment that can be confused with pedestrians, cyclists, and animals. In contrast, NIR images usually include many nontarget objects that may interfere with the detection of pedestrians or be confused with pedestrian targets. These nontargets include oncoming traffic, traffic lights, streetlights, road signs, lane delineators, etc. To support the claim that visual clutter contributes to the decreased detection distance in NIR systems relative to FIR systems, it is shown that: (1) visual clutter in NIR images is higher than in FIR images, and that (2) as visual clutter increases detection distance decreases, independent of which night vision system is used. Image processing is a promising technology that is likely to improve the quality of night vision images.
Visual clutter in active night vision systems reduces detection distance
Visuelle Störsignale in aktiven Nachtsichtsystemen verringern den Erkennungsabstand
2005
9 Seiten, 4 Bilder, 8 Quellen
Conference paper
English
Tema Archive | 2002
|SAE Technical Papers | 2002
|Seeing pedestrians at night: Visual clutter does not mask biological motion
Online Contents | 2009
|Seeing pedestrians at night: Visual clutter does not mask biological motion
Online Contents | 2009
|