Join or Sign in

Register for your free asmag.com membership or if you are already a member,
sign in using your preferred method below.

To check your latest product inquiries, manage newsletter preference, update personal / company profile, or download member-exclusive reports, log in to your account now!
Login asmag.comMember Registration
https://www.asmag.com/showpost/34324.aspx
INSIGHTS

"Body recognition" to enhance facial recognition technology

Body recognition analytics is increasingly being integrated with facial recognition technology to improve overall efficiency and security
Imagine a crime scene where the CCTV camera could not capture the face of the criminal. What can the police do with the visuals of the body alone? Until recently, a manual check of other footages for people with similar attributes such as clothes, height, etc., was the only answer. But now, body recognition analytics added to facial recognition technology systems makes things simpler, with algorithms quickly analyzing attributes such as body mass and clothing to narrow down the search to visuals where the face is available before using facial recognition.

Body recognition is increasingly used to enhance facial recognition technology solutions, according to Saurabh Gandhi, Technical Services Manager at AnyVision. There are several instances where a CCTV camera may not capture a person’s face clearly, making facial recognition solutions or face detection useless. With an integrated layer of body recognition, the system can search for people of similar physical attributes across footage from different cameras and find the face.

Also read: Top facial recognition access control systems 

Facial recognition technology enabler for legacy systems

A popular market for body recognition technology is customers using existing CCTV cameras for facial recognition. Gandhi points out that because these cameras have been installed way before facial recognition became so widely used, their positioning may not suit the analytics. With body recognition, the system can work with footage from multiple cameras and track down a person.

“Wherever you have to use an existing CCTV camera that was deployed without keeping a facial recognition use case in mind, the adoption of the analytics becomes a challenge,” Gandhi explained. “The camera could be repositioned, but then you may lose the details that were originally meant to be captured. This was a challenge we used to see across all verticals, but most often in city surveillance and casinos.”

Suited for human memory

Another distinct advantage of the body recognition solution is that human beings are wired to remember body attributes better than faces. For instance, when the police questions a witness about a crime and asks to describe the criminal they may have seen, attributes like height and shirt color are more easily remembered than facial features.

“When we look at a person’s body, there are a few attributes we commonly observe,” Gandhi said. “One is gender. Others would include the color of their shirt or bottom-wear. These are attributes that can help you shortlist suspects, and this is where body recognition can really help.”
So, when the police try to corroborate the witness’s description to CCTV videos captured, the system can quickly find people with the given descriptions.

Enhancing facial recognition systems 

The use of body recognition helps facial recognition when a camera may not capture the face of a person. There are certain acceptable degrees of angles (often expressed using terms yaw, pitch, and roll) that a facial recognition algorithm can use to identify a person. When these are not available, an additional layer of body recognition becomes a safety net to avoid security lapses. Such integration has always been an attractive option for customers. 

“This is why we are now using a multi-layer approach, where the face is one option, the body is another, and a third is the combination of face and body,” Gandhi explained. “One of these layers will help you track a person across multiple cameras more effectively compared to using just face.”

A solution to privacy concerns

Finally, body recognition would come as a relief to customers who cannot use facial recognition due to privacy-related concerns of laws forbidding facial data use. Especially in regions like Europe, where strict GDPR and other independent laws limit the use of facial recognition, body recognition serves to maintain security. Gandhi added that there are already quite a few customers making use of the integrated body recognition solution.

Combine for better results

When used together, face and body recognition create a completer and more accurate picture of what’s going on in your space, enhancing your ability to protect the people within it. As incidents like recent terror attacks in Europe continue to increase security concerns worldwide, this combination could help authorities protect people and assets. Also, this could be yet another criterion for customers to consider when purchasing facial recognition technology systems


Product Adopted:
Software
Subscribe to Newsletter
Stay updated with the latest trends and technologies in physical security

Share to: