Acta Optica Sinica, Volume. 39, Issue 7, 0710001(2019)

Multispectral Face Image Registration Based on T-Distribution Mixture Model

Wei Li1, Mingli Dong2、*, Naiguang Lü1,2, and Xiaoping Lou2
Author Affiliations
  • 1 Institute of Information Photonics and Optical Communications, Beijing University of Posts & Telecommunications, Beijing 100876, China;
  • 2 Key Laboratory of the Ministry of Education for Optoelectronic Measurement Technology and Instruments, Beijing Information Science & Technology University, Beijing 100192, China;
  • show less
    Figures & Tables(9)
    Infrared and visible face image registration process based on feature maps
    Diagrams of face silhouette feature extraction based on IDSC. (a) Bellman-Ford shortest path graph built using face silhouette landmark points; (b) four marked points; (c) their IDSC feature histograms
    Student's T distribution for various degrees of freedom
    Flow chart of proposed registration algorithm
    Registration results on synthetic Chinese character point sets. (a)-(c) Point sets before registration; (d)-(f) registration results of proposed algorithm; (g)-(i) recall curves of three registration algorithms
    Visible and infrared face registration results in UTK-IRIS database. (a)(b) Original visible and infrared images; (c) face edge maps; (d)(e) checkerboards of warping visible image into and infrared image
    Quantitative result comparison of multispectral face image pairs of different individuals. (a) Charles; (b) Heo; (c) Gribok; (d) Sharon
    Visible and infrared face image registration and fusion results captured by ourselves. (a) Original visible images; (b) original infrared images; (c) checkerboard images; (d) fusion images
    • Table 1. Comparison of average matching errors and running time with different registration algorithms

      View table

      Table 1. Comparison of average matching errors and running time with different registration algorithms

      AlgorithmAverage matching error /pixelMean running time /s
      CharlesHeoGribokSharon
      SMM[13]1.81.51.61.71.1
      RPM-GMM[10]1.51.41.41.21.5
      L-SMM1.21.11.00.90.9
    Tools

    Get Citation

    Copy Citation Text

    Wei Li, Mingli Dong, Naiguang Lü, Xiaoping Lou. Multispectral Face Image Registration Based on T-Distribution Mixture Model[J]. Acta Optica Sinica, 2019, 39(7): 0710001

    Download Citation

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

    Category: Image Processing

    Received: Jan. 7, 2019

    Accepted: Apr. 1, 2019

    Published Online: Jul. 16, 2019

    The Author Email: Dong Mingli (dongml@bistu.edu.cn)

    DOI:10.3788/AOS201939.0710001

    Topics