Robust depth estimation on real-world light field images using Gaussian belief propagation

作者：

Highlights：

• A novel Gaussian belief propagation (GaBP) based algorithm is proposed for light field depth estimation.

• The disparity maps between every adjacent sub-aperture image pair propagate to each other via GaBP.

• The proposed method is robust on noisy real-world light field images and requires no training stage.

• The method accurately predicts the depth of far distance objects, e.g., 3.5–9.5 m.

摘要

Highlights•A novel Gaussian belief propagation (GaBP) based algorithm is proposed for light field depth estimation.•The disparity maps between every adjacent sub-aperture image pair propagate to each other via GaBP.•The proposed method is robust on noisy real-world light field images and requires no training stage.•The method accurately predicts the depth of far distance objects, e.g., 3.5–9.5 m.

论文关键词：Light field,Depth estimation,Optical flow,Real-world,Gaussian belief propagation

论文评审过程：Received 13 October 2021, Revised 1 March 2022, Accepted 1 April 2022, Available online 8 April 2022, Version of Record 28 April 2022.

论文官网地址：https://doi.org/10.1016/j.imavis.2022.104447