摘要: |
The parking assistance system evaluated consisted of four outward facing cameras whose images could be presented on a monitor on the center console. The images presented varied in the location of the virtual eye point of the camera (the height above the vehicle and forward depression angle) and the number of camera images (one or multiple) appearing together. In the experiment, 16 drivers (eight under age 30, eight over age 65) parked a 2002 Infiniti Q45 in an 8.5-foot-wide stall. Each subject parked 24 times, using the various image combinations or no assistance at all. Parking clearances (the closest approach distance on the driver and passenger side, distance to a wall in front) declined with practice, stabilizing to 4 inches laterally and 13 inches in front after about three unrecorded practice trials and 15 test trials. Overall, age and sex differences were not large, though older drivers were more variable than younger drivers in how far they parked from the end of a stall. Finally, there were no statistically significant differences among camera image combinations, or differences from no-assistance conditions. Only half of the subjects made significant use of the assistance system, which is not an unusual outcome given its early state of system development. Of the image combinations examined, those with a single image field resulted in subjects parking 2 inches closer to the wall than those with multiple fields. Overall, subjects preferred scenes presenting what appeared to be a single image. Subjects reported that the aerial view was the most helpful. |