• DocumentCode
    2526541
  • Title

    Towards Recognition of Humans and their Behaviors from Space and Airborne Platforms: Extracting the Information in the Dynamics of Human Shadows

  • Author

    Stoica, Adrian

  • Author_Institution
    Jet Propulsion Lab., California Inst. of Technol., Pasadena, CA
  • fYear
    2008
  • fDate
    4-6 Aug. 2008
  • Firstpage
    125
  • Lastpage
    128
  • Abstract
    This paper proposes the use of shadows and shadow dynamics for the recognition of humans and human behavior. ldquoShadow biometricsrdquo may be a useful addition to the biometrics arsenal; shadows can be used in isolation, or in addition to the dasiadirectpsila (body) observation. Shadows may offer significantly more information could be obtained (larger image areas and more behavioral dynamics details) than in a direct observation of humans from above. Thus, shadow biometrics enables dasiaabove-the-headpsila remote biometrics, based on which future space/airborne imaging systems may be able to recognize and track suspects (such as wanted terrorists) by analyzing shadow dynamics (e.g. gait). Beyond the recognition of humans, the movement dynamics may offer information on specific types of behavior, such as those of suicide bombers, rioters or other aggressive individuals. The paper describes the image processing and feature analysis steps and illustrates them with examples.
  • Keywords
    biometrics (access control); feature extraction; image recognition; above-the-head remote biometrics; airborne imaging systems; biometrics arsenal; feature analysis; human behavior; human recognition; human shadows; image processing; shadow biometrics; shadow dynamics; space imaging systems; Biometrics; Data mining; Head; Humans; Image analysis; Image recognition; Information security; Intelligent systems; Space technology; Terrorism;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Bio-inspired Learning and Intelligent Systems for Security, 2008. BLISS '08. ECSIS Symposium on
  • Conference_Location
    Edinburgh
  • Print_ISBN
    978-0-7695-3265-3
  • Type

    conf

  • DOI
    10.1109/BLISS.2008.36
  • Filename
    4595808