Face Recognition

            Face Recognition functionality is unique to cameras such as the Panasonic G series that have live view sensors. It is only available for pictures and can slow down your camera’s operating time. It is such an interesting feature that we think you will want to play with it—even if it is at a later date after you have gained more familiarity with the camera. It seems uncanny when you raise the camera to your eye, it has correctly identified a specific person. You will see this even before you take your picture. The camera will have the recognized person’s name on display as you frame the image. If you determine at a later time you don’t need Face Recognition functionality, you can always turn it off. Also, if you do not think you need to use Face Recognition with the camera, feel free to jump to the Intelligent Auto Mode section.

Remember that Panasonic distinguishes between facial identification and facial recognition. The former describes the identification of a generic face. If a face is detected, the camera will prioritize its action to make sure the face is the point of focus. Face Recognition goes a step further in that the camera executes software to determine if the detected face is stored among the registered faces. If the camera recognizes the detected face, you will be informed of it while framing the image and the person’s name and age will be stored with the saved picture.

For the software to work, you will have to compile a library of images of a person’s face and associate those images with a name and birth date. There are limitations. You can register only up to a total of six people for Face Recognition. Each person’s registration will include their name, age, and up to three different sample images of their face. These images should all be different but cover typical looks for the person. In addition, the images must be of the subject’s full face. The software bases its determination on recognizing two eyes, a nose, and a mouth at typical human proportions and locations. So, if the registered person is not facing the camera, but instead presents his profile to be photographed, facial recognition will fail. Also, certain conditions, such as being too distant from the camera (face appears too small) or harsh shadows falling on the persons face, may interfere with correct recognition.

To make the image library, you will have to decide whom you wish to register and place images of them in the Face Recognition system. Suppose you wish the camera to recognize your best friend. This is accomplished by taking up to three different sample images of her full face, showing a range of typical facial expressions, such as a big grin outdoors, a serious contemplative look while indoors, and a subtle, Mona Lisa smile taken with an electronic flash. This way the camera’s software will have a variety of your best friend’s typical facial expressions under different lighting conditions for reference. We recommend that you take the maximum of three sample images for each person you register. This improves the software’s accuracy in recognizing and identifying the person.

The Face Recognition software works in group shots, recognizing up to three people in a single picture, even when they are scattered among a group of people. Of course, to be recognized, each registered person must be facing toward the camera so that his or her facial features are clearly presented. Also, if a group is too large and, as a result, a registered face occupies too small an area on the sensor, it may not be recognized.

When you raise your camera to take a picture, the camera software will determine if it can identify the face belonging to any of the six Face Recognition registered people. If so, the saved registered name will be superimposed on the image below the framed face.