Abstract

In this study, a 3D salient object detection model is built at the acquisition step in the full-color holographic system, and a deep network architecture ${{\rm U}^2}$-reverse attention and residual learning (RAS) algorithm is proposed for salient object detection to obtain more efficient and accurate point cloud information. In addition, we also use the point cloud gridding method to improve the hologram generation speed. Compared with the traditional region of interest method, RAS algorithm, and ${{\rm U}^2}$-Net algorithm, the computational complexity is significantly reduced. Finally, the feasibility of this method is proved by experiments.

© 2023 Optica Publishing Group

Implementation of a full-color holographic system using RGB-D salient object detection and divided point cloud gridding

Yu Zhao, Jing-Wen Bu, Wei Liu, Jia-Hui Ji, Qin-Hui Yang, and Shu-Feng Lin
Opt. Express 31(2) 1641-1655 (2023)

Implementation of full-color holographic system using non-uniformly sampled 2D images and compressed point cloud gridding

Yu Zhao, Ki-Chul Kwon, Munkh-Uchral Erdenebat, Seok-Hee Jeon, Mei-Lan Piao, and Nam Kim
Opt. Express 27(21) 29746-29758 (2019)

Cauchy graph embedding based diffusion model for salient object detection

Yihua Tan, Yansheng Li, Chen Chen, Jin-gang Yu, and Jinwen Tian
J. Opt. Soc. Am. A 33(5) 887-898 (2016)

Salient object detection using coarse-to-fine processing

Qiangqiang Zhou, Lin Zhang, Weidong Zhao, Xianhui Liu, Yufei Chen, and Zhicheng Wang
J. Opt. Soc. Am. A 34(3) 370-383 (2017)

Performance of a U²-net model for phase unwrapping

Liangzhe Liao, Zhenkun Lei, Chen Tang, Ruixiang Bai, and Xiaohong Wang
Appl. Opt. 62(34) 9108-9118 (2023)