ALL_TII-19-1583 (002).pdf (1.53 MB)
Download fileAssembling convolution neural networks for automatic viewing transformation
journal contribution
posted on 2019-09-27, 12:27 authored by Haibin Cai, Lei Jiang, Bangli Liu, Yiqi Deng, Qinggang MengQinggang MengImages taken under different camera poses are
rotated or distorted, which leads to poor perception experiences.
This paper proposes a new framework to automatically transform
the images to the conformable view setting by assembling
different convolution neural networks. Specifically, a referential
3D ground plane is firstly derived from the RGB image and
a novel projection mapping algorithm is developed to achieve
automatic viewing transformation. Extensive experimental results
demonstrate that the proposed method outperforms the state-ofthe-art vanishing points based methods by a large margin in
terms of accuracy and robustness.
Funding
YOBAN project (Newton Fund/Innovate UK, 102871)
EPSRC CDT-EI
SukeIntel Co., Ltd
History
School
- Science
Department
- Computer Science