A synthesis result of the proposed view sharing system. (a) The
original view of the preceding vehicle. (b) The original view of the
subject vehicle with a large portion of the image blocked by the
preceding vehicle. (c) The perspective of the preceding vehicle is
transferred to the corresponding view of the subject vehicle to
“disocclude” the blocked area as if the preceding vehicle becomes
transparent.
Visual obstruction caused by a preceding
vehicle is one of the key factors threatening driving safety. One
possible solution is to share the first-person-view of the preceding
vehicle to unveil the blocked field-of-view of the following vehicle.
However, the geometric inconsistency caused by the camera-eye
discrepancy renders view sharing between different cars a very
challenging task. In this paper, we present a first-person-perspective
image rendering algorithm to solve this problem. Firstly, we contour
unobstructed view as the transferred region, then by iteratively
estimating local homography transformations and performing
perspective-adaptive warping using the estimated transformations, we
are able to locally adjust the shape of the unobstructed view so that
its perspective and boundary could be matched to that of the occluded
region. Thus, the composited view is seamless in both the perceived
perspective and photometric appearance, creating an impression as if
the preceding vehicle is transparent. Our system improves the driver’s
visibility and thus relieves the burden on the driver, which in turn
increases comfort. We demonstrate the usability and stability of our
system by performing its evaluation with several challenging data sets
collected from real-world driving scenarios.
An overview of the proposed method. Given the target and
reference sequences, the occluded region (a) in the target image is
estimated and our system automatically finds the corresponding contour
(b) in the reference image. To transform the area inside the contour in
the reference image to match the occluded region in the target image,
the perspective of the region to be transformed are adapted to fit that
of the location in the target image by performing (c) perspective
adaptation through reference video volume and (d) a stitching process
between the two image frames. In the stage of perspective adaptation, a
novel view I′ is synthesized by performing local homography estimation
and perspective-aware warping. Finally, we stitch the synthesized view
and target image where the warped region is seamlessly blended into the
target image to make an impression that the vehicle is transparent (e).
Note that the “see-through” effect does not cover the entire occluded
region such that the viewers remain consciously aware of the existence
of the preceding vehicle, thus improving driving safety.
