Article
Optics
Xinrui Zhan, Chunli Zhu, Jinli Suo, Liheng Bian
Summary: This letter presents a weighted optimization technique for sampling-adaptive single-pixel sensing. By introducing a weighting scheme, the method achieves efficient sensing and only requires training the network once to adapt to different sampling rates.
Article
Computer Science, Artificial Intelligence
Sidi Lu, Xin Yuan, Aggelos K. Katsaggelos, Weisong Shi
Summary: In this work, reinforcement learning is applied to video compressive sensing to adapt the compression ratio. The gap in previous studies of how to adapt B in the video SCI system is filled using RL. An RL model and various convolutional neural networks are employed to achieve adaptive sensing of video SCI systems. Additionally, the performance of an object detection network is utilized for RL-based adaptive video compressive sensing. This proposed adaptive SCI method can be implemented in low cost and real time, and takes the technology one step further towards real applications of video SCI.
ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY
(2023)
Article
Optics
Zixin Tang, Jianwei Zhang, Zhenyu Xu, Mao Ye, Yiguang Liu
Summary: This study introduces the Stockwell transform into the field of Single-Pixel Imaging (SPI). By optimizing the basis selection and designing an adaptive differential modulation technique, the imaging efficiency and resolution have been improved.
Article
Optics
Rui Zhang, Wenyi Ren, Zhilong Xu, He Wang, Jiangang Jiang, Yuanyuan Wang, Xibo Luo
Summary: This paper proposes a single pixel echelle spectrometer system based on compressed sensing theory. The system utilizes Digital Micromirror Devices as the Coding board and high sensitivity photomultiplier point detector as the signal receiver. The optical system is designed according to the structural characteristics of the system, achieving 70% energy concentration in the spot of a single micro mirror. Furthermore, the spectral information reconstruction algorithm is studied, where the original two-dimensional spectrum can be reconstructed with a compression factor of 20% and transformed into the required one-dimensional spectrum through a spectrum reduction algorithm.
Article
Physics, Applied
Jun-Tian Ye, Chao Yu, Wenwen Li, Zheng-Ping Li, Hai Lu, Rong Zhang, Jun Zhang, Feihu Xu, Jian-Wei Pan
Summary: We demonstrate photon-counting single-pixel imaging in the ultraviolet region using a 4H-SiC single-photon avalanche diode (SPAD) as a high-performance compact single-photon detector. By developing a tailored readout circuit with active hold-off time, we restrain detector noise and operate the SPAD in the free-running mode. With structured illumination, we reconstruct compressed images at a 4 fps frame rate, demonstrating the capability of ultraviolet imaging applications.
APPLIED PHYSICS LETTERS
(2023)
Article
Computer Science, Information Systems
Hadi Hadizadeh, Ivan V. Bajic
Summary: This paper presents a soft video multicasting system using adaptive block-based compressed sensing method, in which each block in each frame of the input video is adaptively sampled based on texture complexity and visual saliency. The system achieves better reconstruction quality at the decoder side using an iterative algorithm and adaptive processing, exploiting temporal similarity between adjacent frames. Extensive experimental results demonstrate the superiority of this system over existing soft video multicasting systems.
IEEE TRANSACTIONS ON MULTIMEDIA
(2021)
Article
Physics, Multidisciplinary
Jianming Wang, Jianhua Chen
Summary: An adaptive rate Compressive Sensing method for video signals is proposed, which utilizes Blocked Compressive Sensing scheme and adaptive rate sampling. Experimental results show that the method achieves accurate sparsity estimation and better image reconstruction quality compared to the latest adaptive Compressive Video Sensing methods.
Article
Optics
Rui Li, Jiaying Hong, Xi Zhou, Qin Li, Xiao Zhang
Summary: Single-pixel imaging based on FRFT offers advantages such as wide operation wavelength range, low radiation dose, and insensitivity to distortion. The method utilizes structured illumination and inverse FRET for object image reconstruction.
Article
Optics
Haocun Qi, Shu Zhang, Zhuang Zhao, Jing Han, Lianfa Bai
Summary: This paper proposes a dual optical path spectral imaging system based on single-pixel imaging, which reduces sampling time by lowering imaging resolution. The system does not require spectral unmixing and employs a lightweight deep learning algorithm for image fusion.
OPTICS COMMUNICATIONS
(2022)
Article
Optics
Canwen Xu, Aiping Zhai, Wenjing Zhao, Peng He, Dong Wang
Summary: An adaptive sampling method is proposed for single-pixel imaging to improve imaging speed by reducing the sampling rate while sacrificing as little as possible of the imaging fidelity. The method is based on sparse coefficient matrix obtained from orthogonal transforms, and adaptively stops sampling based on real-time variance curve changes.
OPTICS COMMUNICATIONS
(2021)
Article
Engineering, Electrical & Electronic
Ning Jing, Chen Yu, Kewu Li, Rui Zhang, Zhibin Wang, Chao Wang
Summary: This new compressive sensing absorption spectroscopy technique uses a photoelastic modulator (PEM) to modulate the input spectrum and create multiple harmonic orders for measurement channels, achieving a high compression ratio. By randomly selecting elements to form a sensing matrix, it successfully reconstructs the infrared spectrum with CO2 absorption.
IEEE SENSORS JOURNAL
(2021)
Article
Multidisciplinary Sciences
Yuning Guo, Baowen Li, Xiaobo Yin
Summary: This study proposes a dual-compressed photoacoustic imaging method with single-pixel detection, which enables high-efficiency surface tomography with 3D spatial resolution unlimited by the acoustics, revealing the potential for dynamic imaging in a scattering medium.
NATIONAL SCIENCE REVIEW
(2023)
Article
Engineering, Electrical & Electronic
Markus Hofbauer, Christopher B. Kuhn, Goran Petrovic, Eckehard Steinbach
Summary: The article proposes a preprocessing filter concept that allows for individual rate/quality adaptation while using a single encoder. By designing analytical and machine learning-based bitrate models, various factors can be considered when estimating video bitrate, which outperforms previous models significantly. Experimental results show that the preprocessing approach can achieve similar bitrates to individual encoders in a teledriving scenario with a smaller total bitrate required.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Sameh Ghanem
Summary: This paper introduces a proposed adaptive compressive measurement method for decoding frequency-hopping receivers, enhancing detection performance of received frequencies and decoding accuracy. The proposed method effectively recovers transmitted symbols by relying on the instantaneous sparsity of the spectra of FHSS signals.
INTERNATIONAL JOURNAL OF COMMUNICATION SYSTEMS
(2021)
Article
Optics
Anthony T. Giljum, Kevin F. Kelly
Summary: This study utilizes foveation to achieve a 4x improvement in L1 compressive sensing reconstruction speed for hyperspectral images and video, with the advantage of flexibility and adjustability.