Deprecated: mysql_connect(): The mysql extension is deprecated and will be removed in the future: use mysqli or PDO instead in /home/virtualw/public_html/Resources/Hosted/BigImage.php on line 9
Canine Facial Recognition
Large Image Display:Canine Facial Recognition

This image is taken from the immeasurably silly 2010 film Cats And Dogs: The Revenge Of Kitty Galore. Whilst the rest of the film holds nothing new, and very little that isn't either done better elsewhere, or is extremely predictable, one moment of the film does give pause.

That's the moment when Diggs, the canine protagonist and star of the film is recruited. The dog agency already know who they want, and where to find him. This periscopic sensor pops up in the middle of the New York police dog kennels, and narrows in on the desired dog by means of facial recognition technology, drawing a contour match of each dog's face in order to find the one that matches.

It is not a silly idea. Using facial detection techniques on non-human subjects is something that is already occurring with great apes in the wild. So, its not a great stretch to apply it to identifying individuals of other species as well.

As with humans, the trick is to have the machine vision system identify key aspects of the face - the nose, eyes, and mouth as were done here, and track the distortions to determine the probability of it being the desired individual. It is unlikely to be 100% accurate as the film tries to portray. Rather, all you can hope for is a very strong probability of a match, heading towards 100% as the same face is seen by the system, multiple times.

Even with humans, we still get the occasional case of mistaken identity, after all.


Warning: Unknown: write failed: Disk quota exceeded (122) in Unknown on line 0
503 Service Unavailable

Service Unavailable

The server is temporarily unable to service your request due to maintenance downtime or capacity problems. Please try again later.

Additionally, a 503 Service Unavailable error was encountered while trying to use an ErrorDocument to handle the request.