• Chinese Optics Letters
  • Vol. 19, Issue 12, 121101 (2021)
Liyu Zhou, Xianwei Huang, Qin Fu, Xuanpengfan Zou, Yanfeng Bai*, and Xiquan Fu**
Author Affiliations
  • College of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, China
  • show less
    DOI: 10.3788/COL202119.121101 Cite this Article Set citation alerts
    Liyu Zhou, Xianwei Huang, Qin Fu, Xuanpengfan Zou, Yanfeng Bai, Xiquan Fu. Fine edge detection in single-pixel imaging[J]. Chinese Optics Letters, 2021, 19(12): 121101 Copy Citation Text show less
    Setup of LFSI.
    Fig. 1. Setup of LFSI.
    Simulation results. (a) ‘HNU’ binary image. (b) Original edge image of (a). (c)–(e) The results of edge detection based on Scheme I, Scheme II, and LFSI, respectively.
    Fig. 2. Simulation results. (a) ‘HNU’ binary image. (b) Original edge image of (a). (c)–(e) The results of edge detection based on Scheme I, Scheme II, and LFSI, respectively.
    Simulation results after adding noise. (a)–(c) Results of edge detection based on Scheme I, Scheme II, and LFSI. (d) Morphological processing results of (c). First row: adding SNR = 10 dB detection noise and no light source noise; second row: adding SNR = 10 dB light source noise and no detection noise; third row: adding SNR = 10 dB detection noise and light source noise.
    Fig. 3. Simulation results after adding noise. (a)–(c) Results of edge detection based on Scheme I, Scheme II, and LFSI. (d) Morphological processing results of (c). First row: adding SNR = 10 dB detection noise and no light source noise; second row: adding SNR = 10 dB light source noise and no detection noise; third row: adding SNR = 10 dB detection noise and light source noise.
    Simulation results after adding noise. (a1), (a2) Original image and edge image of ‘windmill’ grayscale image. (b)–(d) Results of edge detection based on Scheme I, Scheme II, and LFSI, respectively. (e) Morphological processing results of (d). First row: adding SNR = 10 dB detection noise and no light source noise; second row: adding SNR = 10 dB light source noise and no detection noise; third row: adding SNR = 10 dB detection noise and light source noise.
    Fig. 4. Simulation results after adding noise. (a1), (a2) Original image and edge image of ‘windmill’ grayscale image. (b)–(d) Results of edge detection based on Scheme I, Scheme II, and LFSI, respectively. (e) Morphological processing results of (d). First row: adding SNR = 10 dB detection noise and no light source noise; second row: adding SNR = 10 dB light source noise and no detection noise; third row: adding SNR = 10 dB detection noise and light source noise.
    Experimental results. (a1), (b1) Original image and edge of ‘G’ binary image. (a2), (b2) Original image and edge of ‘ghost’ image. (c)–(e) Experimental results of edge detection based on Scheme I, Scheme II, and LFSI, respectively. (f) Morphological processing results of (e).
    Fig. 5. Experimental results. (a1), (b1) Original image and edge of ‘G’ binary image. (a2), (b2) Original image and edge of ‘ghost’ image. (c)–(e) Experimental results of edge detection based on Scheme I, Scheme II, and LFSI, respectively. (f) Morphological processing results of (e).
    Liyu Zhou, Xianwei Huang, Qin Fu, Xuanpengfan Zou, Yanfeng Bai, Xiquan Fu. Fine edge detection in single-pixel imaging[J]. Chinese Optics Letters, 2021, 19(12): 121101
    Download Citation