• The system can control two cameras at the same time. Each camera can be used with different function. For example, camera A can be used with surveillance function and camera B can be used with access control function.
• The system supports a special function: train memory. This function analyzes all enrolled users’ data and trains a special memory to improve accuracy of face recognition.
• The system uses a patented image tracking/analyzing function to make sure that the recorded facial images are always clear.
• The system automatically tracks the head motion and gives feedback to user. For example, the system detects the head size and view angle and then gives proper instructions to user such as “stand closer please” or “please look straight”.
• The system also supports a hardware I/O control box. This I/O control box can access a maximum of 8 devices.
• The system provides a CPU performance option to ensure that the system can run smoothly at different CPUs. For example, for a low performance CPU, the system can skip feedback drawings on preview window to prevent computer from locking up.

Hardware Specifications

• Package comes with application software, one capture card, one Industrial PC platform, and one IO box
• Hardware Requirements:
  • P4 2.8 GHz or faster CPU with at least 512MB memory, Industrial PC Grade platform recommended.
  • 1 or 2 CCTV cameras (used as indoor and /or outdoor camera) with a good background lighting control(BLC), and a fast digital shutter
  • Camera lens: recommend a 3~8mm zoom lens for ease of installation
  • Camera installation: suggest 1) install cameras at average height of people, 2) camera should avoid aiming at outdoor with extreme light change
  • Lighting requirements: No obvious shadow or highlight on face, at least 500lux (similar to office lighting)
  • Hard disk with at least 5GB free space for storing image data, selectable storage time then overwrite
  • DC 12V electric deadbolt
  • (optional) speakers – for announcing names or welcome messages

Software Specifications

• Enrollment:
  • Max Enrollment: 200 people
  • Single Enrollment: The system can enroll one person at a time in front of cameras.
  • Batch Enrollment: The system can have a period of “Training Mode” (user definable, 1 week for example) to collect facial data for batch enrollment. During this period, the system opens the door when a clear facial data is collected on either camera.
• Access Control
  • Access Control Modes:
    • Face detection: door opens whenever a clear face is captured and logged
    • Face recognition: door opens only when the face is recognized as one of the enrolled persons
• Time & Attendance (T&A)
  • In either Access Control Mode, the recognized face is logged for T&A purpose
  • A T&A record is stored as an image file with user ID, date and time as the file name. Any backend application can retrieve these T&A records and parses the filenames with ease.
• Surveillance
  • Either camera can detect up to 4 faces at a time. Faces are detected, recognized against database, and stored. Therefore any person appears near the cameras are captured – this is a more effective way of surveillance than a corner CCTV camera because the facial images are always clear and recognizable.
• Data Management
  • Allow system manager to query any records or logs, including T&A records and surveillance records
  • Allow system manager to delete data to make room
• Facial Detection/Recognition Software SDK
  • ISS-D02 Face Detection SDK
  • ISS-D02ATM,No Face Detection SDK
  • ISS-R02, Face Recognition SDK