细节展示放大镜特效

Lin'an Smart Face Recognition

  • No:Lin'an Intelligent Face Recognition
  • Quantity:

      Recommend

    Product introduction
    Design highlights: dynamic video face grab smaller omissions, recognition speed. Dynamic portrait system: Lin'an's dynamic portrait system is a face recognition product based on dynamic video flow, which integrates the functions of face tracking capture and recognition, real-time control, and passing path query. Static portrait system: static portrait ratio support s1,1,1:N,N:N match, widely used in public security, finance and other fields, can be built on the urban population and key population, through portrait comparison can quickly and accurately confirm the identity of the person being queried.
    Product features
    Biometric intelligent recognition for real-time capture of faces
    Intelligently judge biological behavior, such as opening, blinking, nodding, etc.
    With accurate identification, fast judgment, smooth ness and so on
    Performance in low-light situations
    Effectively avoidhuman photo-spoofing system
    Replace manual verification, resulting in significant labor cost savings
    Workflow
    Lin's intelligent face recognition system makes real-time face capture on the video stream, and compares the captured face to the image library quickly, and presents the results to the user in real time.
    Technical parameters
    Lin's intelligent face recognition system can accurately and quickly identify faces, face feature extraction rate is almost no missing, millions of face feature data N:N comparison time is only about 5 milliseconds.
    Name Description
    Moving object snaps Automatic tracking of moving faces, detects and captures still and moving portraits, automatically detects from video streaming signals, and captures multiple portrait signals simultaneously.
    Photo collection Minimum acquisition illumination is less than 10Lux (black and white); portrait acquisition frame rate of 10 frames (black and white), s6 frames (color); maximum detection of portraits in the same frame screen is s.8
    Capture portrait size The pitch is greater than 60 pixels
    Identify portrait sizes Portrait size at least 120 x 180, the more ideal size is: 240 x 360 or more
    Identify distances 0.5-150M (recognition distance is related to camera lens)
    Identify angles The front avatar is the best recognizable location. The change of posture in the range of about 15 degrees or less 15 degrees does not affect the recognition effect.
    Change factors Light and background do not have special requirements, and it works best when the surrounding light source is diffuse. Try to avoid strong backlight and side light.
    Library speed Less than 1S/sheet
    Identify the response Misperception rate: less than 1%
    Rejection rate: less than 1%
    Alarm response: less than 1 second