Products
Solutions
Featured Analytics
About
References
intuVision in Action

intuBlog

(781) 497-1015

intuVision VA Event - Face Detection

 

Face Detection alerts on people faces seen in the video frame, tracks them in the video and creates a catalogue of detected faces with thumbnail images.

intuVision VA Face Detection

Face Detection is available in intuVision VA’s Face Detection & Matching module.

intuVision VA Face Detection, unlike face recognition, does not require a special camera and a close view, and is not a positive ID tool. Any surveillance camera with a full view of people’s faces can be used for face detection. Face Detection alerts on people faces seen in the video frame, tracks them in the video and creates a catalogue of detected faces with thumbnail images. Catalogued faces are stored for easy access and review.

The ability to detect and track multiple faces in each scene makes intuVision VA Face Detection a very versatile tool where using Face Recognition is not possible or proves too expensive. In a given scene such as an existing lobby or a hallway camera all faces are detected and catalogued, as long as they are clearly visible within the +/- 30 degrees of viewing range.

If people have been enrolled into intuVision VA (see Face Matching), each detected face is labeled by the name or ID number of the person by matching the detected faces to those enrolled in the system.

For security and manufacturing domains, intuVision VA Face Detection is typically used to record a catalogue of people entering into selected areas. For instance, in the security domain, it can create a record of employees who entered a restricted room, for later verification. In the manufacturing domain, it can track when employees entered for their shift, went on break, or started specific tasks.

For the retail domain, intuVision VA Face Detection can be used for people counting in scenes where conventional camera views are not viable or practical for people counting. For example, if only a lower camera placement is possible, through which people may not be visible individually, rendering counting detected faces as the most accurate estimate.

Optional settings are available to fine-tune the event, including:

  • Sensitivity, to detect faces at varying levels of visibility or clarity in the scene.
  • Resurrection time, allowing reassociation with between faces if someone temporarily turned away from the camera or left the scene.

When combined with compound events and output trigger options, Face Detection can be further refined, to be used for meeting custom video analytics requirements with our off-the-shelf product without the need for any development time. For example, Face Detection can be used along with a person tailgating alert to record the culprit. If using enrollment, the employee ID number could be easily identified, or if no matches are found raising the alarm level for a potential intrusion for quick investigation.

Check out the intuBlog!