- 最后登录
- 2017-9-18
- 注册时间
- 2011-1-12
- 阅读权限
- 90
- 积分
- 12276
  
- 纳金币
- 5568
- 精华
- 0
|
Abstract
This work presents a method for obtaining perceptually natural
panoramic videos, that is, videos composed of wide-angle frames
where straight lines and object shapes are preserved. The shape
preservation for moving objects has a special treatment that ensures
temporal coherence. Energies describing these properties are
obtained for the case of a fixed omni-directional camera, and an optimization
procedure is presented. Our optimization works per-frame,
which makes the method scalable for arbitrarily long scenes.
Keywords: 2D Morphing and Warping, Panoramic Video.
1 Introduction
Recent developments in computational photography gave rise to
affordable omni-directional cameras, which are output video in the
equirectangular format. Applications in navigation, immersion, and
surveilance are the key driving forces of omni-directional video.
We consider a new use for the spherical videos captured by these
cameras: videos with wide-angle frames. We improve on recent
developments in the understanding of the distortions involved in
wide-angle images [Zorin and Barr 1995; Kopf et al. 2009; Carroll
et al. 2009], and study the implications of the introduction of the
time dimension.
Our method is divided into three phases: pre-processing, optimization
and post-processing. As in the work of [Carroll et al. 2009],
our pre-processing interface allows the user to select lines that are
to be straight in the final result video, as well as their final orientations.
The user can also specify the field-of-view (FOV) to be
considered in the input video. While these operations are performed,
our system detects the moving objects in the video using background
subtraction. Using this input, an optimization phase is performed
(see following sections). Finally, the post-processing stage consists
of specifying the rectangular area of the optimization results to store
in the output video. |
|