Abstract

We propose and experimentally demonstrate a novel, to the best of our knowledge, hybrid optoelectronic system that utilizes mode-selective frequency upconversion, single-pixel detection, and a deep neural network to achieve the reliable reconstruction of two-dimensional (2D) images from a noise-contaminated database of handwritten digits. Our system is designed to maximize the multi-scale structural similarity index measure (MS-SSIM) and minimize the mean absolute error (MAE) during the training process. Through extensive evaluation, we have observed that the reconstructed images exhibit high-quality results, with a peak signal-to-noise ratio (PSNR) reaching approximately 20 dB and a structural similarity index measure (SSIM) of around 0.85. These impressive metrics demonstrate the effectiveness and fidelity of our image reconstruction technique. The versatility of our approach allows its application in various fields, including Lidar, compressive imaging, volumetric reconstruction, and so on.

© 2023 Optica Publishing Group

Single-pixel pattern recognition with coherent nonlinear optics

Ting Bu, Santosh Kumar, He Zhang, Irwin Huang, and Yu-Ping Huang
Opt. Lett. 45(24) 6771-6774 (2020)

Robust and efficient single-pixel image classification with nonlinear optics

Santosh Kumar, Ting Bu, He Zhang, Irwin Huang, and Yuping Huang
Opt. Lett. 46(8) 1848-1851 (2021)

Masked autoencoder for highly compressed single-pixel imaging

Haiyan Liu, Xuyang Chang, Jun Yan, Pengyu Guo, Dong Xu, and Liheng Bian
Opt. Lett. 48(16) 4392-4395 (2023)

Self-supervised learning for single-pixel imaging via dual-domain constraints

Xuyang Chang, Ze Wu, Daoyu Li, Xinrui Zhan, Rong Yan, and Liheng Bian
Opt. Lett. 48(7) 1566-1569 (2023)

Research on data-driven low-sampling-rate single-pixel imaging method

Shaosheng Dai, Ziqiang He, and Jinsong Liu
Opt. Lett. 48(23) 6132-6135 (2023)