• Laser & Optoelectronics Progress
  • Vol. 57, Issue 16, 161013 (2020)
Lifeng He1、2, Yanling Liu1、*, Yan Zhong1, and Bin Yao1
Author Affiliations
  • 1College of Electrical Information and Artificial Intelligence, Shaanxi University of Science & Technology, Xi'an, Shaanxi 710021, China;
  • 2Faculty of Information Science and Technology, Aichi Prefectural University, Nagakute, Aichi 480- 1198, Japan
  • show less
    DOI: 10.3788/LOP57.161013 Cite this Article Set citation alerts
    Lifeng He, Yanling Liu, Yan Zhong, Bin Yao. Video Foreground Target Extraction Algorithm in Complex Background[J]. Laser & Optoelectronics Progress, 2020, 57(16): 161013 Copy Citation Text show less
    Model of pixel classification model
    Fig. 1. Model of pixel classification model
    Detection results of ViBe algorithm. (a) People; (b) vehicles
    Fig. 2. Detection results of ViBe algorithm. (a) People; (b) vehicles
    Changing trend of different types of pixels. (a) Static background pixel; (b) foreground target pixel (c) dynamic background pixel
    Fig. 3. Changing trend of different types of pixels. (a) Static background pixel; (b) foreground target pixel (c) dynamic background pixel
    Office video results obtained by different algorithms. (a) Original image; (b) GMM algorithm; (c) ViBe algorithm; (d) algorithm of Ref.[7]; (e) algorithm of Ref.[8]; (f) our algorithm
    Fig. 4. Office video results obtained by different algorithms. (a) Original image; (b) GMM algorithm; (c) ViBe algorithm; (d) algorithm of Ref.[7]; (e) algorithm of Ref.[8]; (f) our algorithm
    RiverSide video results obtained by different algorithms. (a) Original image; (b) GMM algorithm; (c) ViBe algorithm; (d) algorithm of Ref.[7]; (e) algorithm of Ref.[8]; (f) our algorithm
    Fig. 5. RiverSide video results obtained by different algorithms. (a) Original image; (b) GMM algorithm; (c) ViBe algorithm; (d) algorithm of Ref.[7]; (e) algorithm of Ref.[8]; (f) our algorithm
    Fall video results obtained by different algorithms. (a) Original image; (b) GMM algorithm; (c) ViBe algorithm; (d) algorithm of Ref.[7]; (e) algorithm of Ref.[8]; (f) our algorithm
    Fig. 6. Fall video results obtained by different algorithms. (a) Original image; (b) GMM algorithm; (c) ViBe algorithm; (d) algorithm of Ref.[7]; (e) algorithm of Ref.[8]; (f) our algorithm
    AlgorithmGMMViBeRef. [7]Ref. [8]Ours
    Precision0.72310.76250.82650.86420.8736
    Recall0.62670.67320.70360.73150.7465
    FNR0.38260.32460.21980.19880.1834
    PWC3.64323.33652.66321.90361.8963
    FM0.63120.65360.75760.77490.7942
    Table 1. Performance comparison of different algorithms
    Lifeng He, Yanling Liu, Yan Zhong, Bin Yao. Video Foreground Target Extraction Algorithm in Complex Background[J]. Laser & Optoelectronics Progress, 2020, 57(16): 161013
    Download Citation