Opto-Electronic Advances, Volume. 2, Issue 9, 190019-1(2019)

Multifunctional inverse sensing by spatial distribution characterization of scattering photons

Lianwei Chen1, Yumeng Yin2, Yang Li1, and Minghui Hong1、*
Author Affiliations
  • 1Department of Electrical and Computer Engineering, National University of Singapore, 4 Engineering Drive 3, 117576, Singapore
  • 2Department of Computer Science, School of Computing, National University of Singapore, 117576, Singapore
  • show less
    Figures & Tables(4)
    Mechanisms to detect information from scattered photon. Optical images of (a) Rayleigh scattering by small molecules, and (b) scattering at the interfaces between air and reflective surfaces. Laser: 532 nm, CW laser. (c) Schematic of the setup to analyze the scattered photon from the laser beam. Power meter is the conventional instrument used for the calibration and training data collection. (d) Diagram showing the detection of the direction of the incident laser beam. Data presented in the spherical coordinates with the green line that denotes the incident laser (polar angle denoted by θ and azimuthal angle φ). (e) Experimental results to show the accuracy of the CNN DL to measure the intensity of the incident laser after every training epoch. Blue curve denotes the accuracy of the model to measure the training data and red curve denotes the accuracy to measure the independent testing data different from the training data. The accuracy of the model is defined as the peak value as 0.99 on the test data curve.
    Schematic illustration of the convolutional neuron network. The details of the pre-processing, architecture, and parameter settings are described in the supplementary materials. For this neuron network, the input is a 150×150×3 image pre-processed from optical images taken from a CMOS. A 2 layers' convolutional network, 1 fully connected layer and one output are used to calculate the light properties of interests.
    Accuracy of the inverse sensing in different conditions.(a) Schematic of the CNN DL method to measure the intensity of the laser beam by analyzing the scattered photon on the surrounding objects. The incident light beam is completely beyond the field of view; Experimental results of the accuracy corresponding to the (b) measurement without the incident light beam in the field of view (peak value 0.99), and (c) simultaneous monitoring of two incident laser beams with one single CMOS (peak value 1). (d) Schematic of the CNN DL to measure two incident laser beams. (e, f) Experimental results of the accuracy corresponding to the remote detection and noise test, respectively.
    Applications of the inverse sensing.(a-c) Illustration of detecting the light which "goes through the wall". (a) The object is completed hidden from the sight of the conventional detector. (b) The scattered photon from the detection laser reveals information of the hidden object. (c) The analysis of the scattered photon re-constructs and identifies the hidden object. (d, e) Dynamic monitoring of the high energy pulsed laser intensity. (d) Accuracy of the CNN DL after each training epoch for the dynamic laser ablation monitoring. (e) Monitoring of the laser intensity at the focal spot during the laser fabrication. (f-h) Light emission intensity characterization of the fluorescence molecules. Optical image of the Rhodamine B patterns irradiated by 532 nm pumping light at different intensities (from left to right: 8, 16, and 29 mW, respectively, scale bar: 50 μm).
    Tools

    Get Citation

    Copy Citation Text

    Lianwei Chen, Yumeng Yin, Yang Li, Minghui Hong. Multifunctional inverse sensing by spatial distribution characterization of scattering photons[J]. Opto-Electronic Advances, 2019, 2(9): 190019-1

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Original Article

    Received: Jun. 10, 2019

    Accepted: Aug. 29, 2019

    Published Online: Nov. 20, 2019

    The Author Email: Minghui Hong (elehmh@nus.edu.sg)

    DOI:10.29026/oea.2019.190019

    Topics