Apple may be looking to add facial recognition to iOS, indirectly backing up previous iTV reports that Apple wants users to be recognized across multiple devices.
As stated in previous reports, Apple insiders claim that the company wants iOS to recognize the end-user across multiple devices, whether its the iPhone or the current hot topic, Apple's alleged iTV. Now, on the heels of recent iTV reports, the US Patent & Trademark Office has conveniently published a patient application from Apple entitled "Low Threshold Face Recognition" which could allow the end-user to unlock an iOS device via facial recognition, even when it's in sleep mode.
According to the application diagrams, the face recognition system doesn't use correlation matching, but rather uses a weighted difference map. The system will reportedly first apply an orange-distance filter to determine variations in skin tone, and then determine the position of high-information areas like the eyes and mouth. It will then analyze the weighted differences between the normalized target face (image) and the normalized detected face (user). After that, both frames are compared on a whole. If successful, the security system will acknowledge the user and unlock the device. Throw in a possible voice recognition via Siri, and you have a two-way security measure that eliminates the need for the current Slide to Unlock and PIN features.
But this new system could do more than just unlock the device -- it could provide user customization. For example, specific settings could be loaded onto an iPad once the device detects a specific user, loading a special wallpaper, app arrangement, notification settings and more. What's more, it could be possible to lock media to a specific user's face, thus leading back to Apple's talks with media execs about recognizing the user across multiple devices. If the user's face is stored in the cloud, it may be possible that the user could share a movie with friends on their own iTV or iPad 2.
But as PatentlyApple points out, the feature is more or less bound to the user's device. "The methods disclosed in Apple's patent specification could adequately recognize a user associated with an iOS device without computing resources overhead that is characteristic of other face recognition techniques," the site reports. "Therefore, the face detection and recognition methods described in Apple's specification could be implemented in hardware, for example in graphical processing units (GPUs) of the iOS device. Apple clarifies that the new face detection and recognition system will apply to the iPhone, iPad, iPod touch and MacBook."
To learn more about Apple's new patent, head here.

Also, how is "low threshold face recognition" secure? If I mock up a mannequin head with the same general features in the correct (or close-enough to pass) positions, or even hold up a photograph of my face, isn't that enough to get past the facial recognition system in a "low threshold" configuration?
The Android face unlock was first shown in may 2011, and this patent was filled in June 2010. I wonder how much time was face unlock under development for Android. Also like you stated facial recognition isn't something new.
Google Goggles was Google's test bed for FR and other recongzation, GG has been out for about 3 years now and pre-dates this patent.
Sigh.... As you probably have hundreds of pictures on mySpace or Facebook....
Irrelevant.
Prior art has to actually be art. Not just a claimed 'in progress' art.
Was Google Goggles used to unlock smartphones?
To confuse this patent with facial recognition in general, as per 'Vladislaus', is to misunderstand what that patent is actually for.
No it isn't enough.
Next year they will "invent" widgets and after that power controls...
too late law enforcement/FBI/DOJ, the DOD, the DOHS and the TSA most likely already have your face on file and has been using it since the 90's.
my question is who did apple steal the software recognition from?