• Laser & Optoelectronics Progress
  • Vol. 62, Issue 3, 0312005 (2025)
Mingqiao Liu1,*, Shenglong Zhang1, Yuanxin Zhong1, Zheng Liu1..., Yanli Gong1,2 and Chunming Gao1|Show fewer author(s)
Author Affiliations
  • 1School of Optoelectronic Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, Sichuan , China
  • 2College of Electronic Engineering (College of Meteorological Observation), Chengdu University of Information Technology, Chengdu 610225, Sichuan , China
  • show less
    DOI: 10.3788/LOP240975 Cite this Article Set citation alerts
    Mingqiao Liu, Shenglong Zhang, Yuanxin Zhong, Zheng Liu, Yanli Gong, Chunming Gao. Extraction of Sound from Secondary Speckle Patterns in Real Time[J]. Laser & Optoelectronics Progress, 2025, 62(3): 0312005 Copy Citation Text show less
    Laser speckle image
    Fig. 1. Laser speckle image
    Principle of laser speckle vibration measurement
    Fig. 2. Principle of laser speckle vibration measurement
    Schematic of real-time audio detection system based on detector and Fresnel lens
    Fig. 3. Schematic of real-time audio detection system based on detector and Fresnel lens
    Laser speckle vibration measurement system
    Fig. 4. Laser speckle vibration measurement system
    MFCC speech quality evaluation process
    Fig. 5. MFCC speech quality evaluation process
    PESQ speech quality evaluation process
    Fig. 6. PESQ speech quality evaluation process
    Indoor experimental scenes
    Fig. 7. Indoor experimental scenes
    Experimental results. (a) Original time-domain signal; (b) spectrogram of original time-domain signal; (c) reconstructed time-domain signal; (d) spectrogram of reconstructed time-domain signal
    Fig. 8. Experimental results. (a) Original time-domain signal; (b) spectrogram of original time-domain signal; (c) reconstructed time-domain signal; (d) spectrogram of reconstructed time-domain signal
    Signal time-domain diagrams at different distances. (a) 20 m; (b) 40 m; (c) 60 m; (d) 80 m
    Fig. 9. Signal time-domain diagrams at different distances. (a) 20 m; (b) 40 m; (c) 60 m; (d) 80 m
    Outdoor experimental scenes
    Fig. 10. Outdoor experimental scenes
    Experimental result. (a) Original time-domain signal; (b) spectrogram of original time-domain signal; (c) reconstructed time-domain signal; (d) spectrogram of reconstructed time-domain signal
    Fig. 11. Experimental result. (a) Original time-domain signal; (b) spectrogram of original time-domain signal; (c) reconstructed time-domain signal; (d) spectrogram of reconstructed time-domain signal
    Schematic of laser irradiation on back of a person's neck
    Fig. 12. Schematic of laser irradiation on back of a person's neck
    Experimental result. (a) Original time-domain signal; (b) spectrogram of original time-domain signal; (c) reconstructed time-domain signal; (d) spectrogram of reconstructed time-domain signal
    Fig. 13. Experimental result. (a) Original time-domain signal; (b) spectrogram of original time-domain signal; (c) reconstructed time-domain signal; (d) spectrogram of reconstructed time-domain signal
    Actual frequency /HzMeasurement frequency /HzAbsolute error /HzRelative error /%Actual frequency /HzMeasurement frequency /HzAbsolute error /HzRelative error /%
    5500400400.020.020.00500
    101000600600.100.100.01600
    303000800799.700.300.03000
    50500010001000.380.380.03800
    70700020002000.540.540.02700
    90900030002999.570.430.02150
    1001000040003999.450.550.01375
    2002000050004999.200.800.02000
    Table 1. Frequency error
    ItemSegSNR/dBMFCCLLRPESQ
    Result9.6540.99604.643
    QualityGoodExcellentExcellentExcellent
    Table 2. Evaluation results of reconstructed speech based on four criteria
    Distance /mSegSNR /dBMFCCLLRPESQ
    207.450.92303.412
    409.630.95603.689
    609.520.95003.874
    806.870.9190.033.523
    Table 3. Evaluation results of speech reconstruction at different distances
    ItemSegSNR /dBMFCCLLRPESQ
    Result6.6540.9280.14.218
    QualityMediumGoodGoodGood
    Table 4. Evaluation results of reconstructed speech based on four criteria
    ItemSegSNR /dBMFCCLLRPESQ
    Result4.490.8960.23.612
    QualityMediumGoodMediumMedium
    Table 5. Evaluation results of reconstructed speech based on four criteria
    Mingqiao Liu, Shenglong Zhang, Yuanxin Zhong, Zheng Liu, Yanli Gong, Chunming Gao. Extraction of Sound from Secondary Speckle Patterns in Real Time[J]. Laser & Optoelectronics Progress, 2025, 62(3): 0312005
    Download Citation