Laser & Optoelectronics Progress, Volume. 57, Issue 16, 161006(2020)

Improved ViBe Moving Object Detection Algorithm to Eliminate Ghost and Shadow

Pengxiang Cui* and Fengqin Yu
Author Affiliations
  • School of Internet of Things Engeneering, Jiangnan University, Wuxi, Jiangsu 214122, China
  • show less
    Figures & Tables(6)
    Schematic diagram of foreground target extraction based on ViBe algorithm
    Flow chart of improved ViBe algorithm
    Contrast diagram of simulation results of ghost detection and elimination. (a) Original video images; (b) traditional ViBe algorithm; (c) algorithm in Ref.[11]; (d) algorithm in Ref.[12]; (e) proposed algorithm
    Contrast diagram of simulation results of shadow detection and elimination. (a) Original video images; (b) traditional ViBe algorithm; (c) algorithm in Ref.[11]; (d) algorithm in Ref.[12]; (e) proposed algorithm
    • Table 1. Performance comparison of average accuracy for different algorithms

      View table

      Table 1. Performance comparison of average accuracy for different algorithms

      VideoTraditional ViBe algorithm[2]Method in Ref. [11]Method in Ref. [12]Our algorithm
      Baseline0.92750.92580.89300.9281
      Dynamic background0.54860.78610.83390.7516
      Camera jitter0.53850.81200.74830.8056
      Intermittent object motion0.65600.74890.71320.7542
      Shadow0.84390.80450.83510.8579
      Thermal0.93450.95360.89300.9458
    • Table 2. Comparison of processing speed of the four algorithms

      View table

      Table 2. Comparison of processing speed of the four algorithms

      VideoTraditional ViBe algorithm[2]Method in Ref. [11]Method in Ref. [12]Our algorithm
      HighwayII35.1228.2730.5833.27
      Cubicle31.5224.7327.1529.36
      BusStation31.1424.2526.8028.97
      PeopleInShade30.7523.8326.0228.26
      Bungalows31.2024.3126.8729.14
    Tools

    Get Citation

    Copy Citation Text

    Pengxiang Cui, Fengqin Yu. Improved ViBe Moving Object Detection Algorithm to Eliminate Ghost and Shadow[J]. Laser & Optoelectronics Progress, 2020, 57(16): 161006

    Download Citation

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

    Category: Image Processing

    Received: Nov. 22, 2019

    Accepted: Jan. 6, 2020

    Published Online: Aug. 5, 2020

    The Author Email: Pengxiang Cui (2858197960@qq.com)

    DOI:10.3788/LOP57.161006

    Topics