Abstract
In this paper, we propose a deep neural network to detect 3D planar surfaces from single plenoptic images captured by a Lytro Illum camera. Different from learning methods based on a single RGB image, we train a network to exploit the light distribution information both in spatial and angular dimensions from multi-sub-aperture images. The features from each sub-aperture image are extracted by using parameter-sharing convolutional layers and then fused to jointly infer the parameters of planes, depths, and segmentation masks. The experiments demonstrate that our approach outperforms the existing state-of-the-art methods with significant margins in the plane detection metrics.
© 2023 Optica Publishing Group
Full Article | PDF ArticleMore Like This
Tianyu He, Wenyi Ren, Yang Feng, Ruoning Yu, Dan Wu, Rui Zhang, Yanan Cai, Yingge Xie, and Jian Wang
Opt. Express 31(20) 33387-33400 (2023)
Qingsong Liu, Xiaofang Xie, Xuanzhe Zhang, Yu Tian, Jun Li, Yan Wang, and Xiaojun Xu
Appl. Opt. 59(14) 4209-4219 (2020)
Andrew-Hieu Nguyen, Brian Sun, Charlotte Qiong Li, and Zhaoyang Wang
Appl. Opt. 61(34) 10105-10115 (2022)