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

Learning based compressive snapshot spectral light field imaging with RGB sensors

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)

Stepwise calibration of plenoptic cameras based on corner features of raw images

Qingsong Liu, Xiaofang Xie, Xuanzhe Zhang, Yu Tian, Jun Li, Yan Wang, and Xiaojun Xu
Appl. Opt. 59(14) 4209-4219 (2020)

Different structured-light patterns in single-shot 2D-to-3D image conversion using deep learning

Andrew-Hieu Nguyen, Brian Sun, Charlotte Qiong Li, and Zhaoyang Wang
Appl. Opt. 61(34) 10105-10115 (2022)

Accurate calibration of standard plenoptic cameras using corner features from raw images

Yuxuan Liu, Fan Mo, Mitko Aleksandrov, Sisi Zlatanova, and Pengjie Tao
Opt. Express 29(1) 158-169 (2021)

Learning feature fusion for target detection based on polarimetric imaging

Sihao Gao, Yu Cao, Wenjing Zhang, Qian Dai, Jun Li, and Xiaojun Xu
Appl. Opt. 61(7) D15-D21 (2022)