Skip to Main Content
 

Global Search Box

 
 
 

ETD Abstract Container

Abstract Header

High-speed Imaging with Less Data

Baldwin, Raymond Wesley
http://orcid.org/0000-0002-2380-6416
http://rave.ohiolink.edu/etdc/view?acc_num=dayton1626353398699931

Abstract Details

2021, Doctor of Philosophy (Ph.D.), University of Dayton, Electrical Engineering.
A primary bottleneck in video processing is the readout of large sensor arrays. Typical video contains highly correlated information, which goes unexploited in traditional imaging devices. This research focuses on two revolutionary hardware designs that eliminate the need for large data handling and bypass the readout of sparse information in large arrays. First, this research proposes a novel representation for event cameras called TORE volumes and demonstrates several advantages over current methods (e.g. prioritized encoding, low computational cost, and temporal consistency). This makes the proposed method an ideal replacement for any machine learning solution that struggles to encode sparse event data into a meaningful dense tensor. TORE volumes are evaluated using several public datasets and achieve state-of-the-art performance for human pose estimation, image reconstruction, event denoising, and classification. Second, this research designs and constructs a prototype Fourier camera that compresses high-speed video in real time. Furthermore, this research evaluates several design parameters, and processing algorithms necessary to capture high-speed video including camera calibration, temporal demosaicking, and frame reconstruction. Fourier cameras perform real-time, hardware-based encoding during a single camera integration via spatial light modulation and use temporal filter arrays to sample time-related information (similar to how color filter arrays sample spectral information in standard cameras). A prototype design is constructed and evaluated against a traditional high-speed camera—achieving 4,000fps with 16× compression. The prototype design serves as an excellent proof of concept for future designs such as on-chip temporal filter arrays.
Vijayan Asari (Advisor)
Keigo Hirakawa (Committee Member)
Theus Aspiras (Committee Member)
Bryan Steward (Committee Member)
117 p.
Computer Engineering; Scientific Imaging
event camera; high-speed; Fourier camera; spatial light modulation; denoising

Recommended Citations

Citations

  • Baldwin, R. W. (2021). High-speed Imaging with Less Data [Doctoral dissertation, University of Dayton]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=dayton1626353398699931

    APA Style (7th edition)

  • Baldwin, Raymond. High-speed Imaging with Less Data. 2021. University of Dayton, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=dayton1626353398699931.

    MLA Style (8th edition)

  • Baldwin, Raymond. "High-speed Imaging with Less Data." Doctoral dissertation, University of Dayton, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=dayton1626353398699931

    Chicago Manual of Style (17th edition)

dayton1626353398699931
50
© 2021, all rights reserved.
This open access ETD is published by University of Dayton and OhioLINK.