What exactly is “face detection technology”, and why is it considered a valuable asset in a digital compact camera?  Part of the confusion about this feature is caused by its rather vague name—“face detection”—that sounds more like something out of a James Bond movie than a photography store.  This article will explain what face detection is, how it works, and why it’s useful in a camera.

First, let me describe the problem that face detection technology in a digital compact camera solves.  Have you ever taken a picture of a person in which the flash from the camera washed out their face, to the point where you can only make out a pale blob with a pair of eyes and lips in it?  This phenomenon usually occurs where taking a picture indoors or outdoors in dim lighting.  When the picture is taken, the flash is too bright.  The light reflects off the person’s skin (usually the face) back at the camera, through the lens, and onto the film or image sensor where, because the flash was too bright, the image of the person’s face ends up being overexposed and consequently washed out.  It is possible to avoid this, however, by manually adjusting the exposure and flash levels.  This takes time, however, and thus is only really practical for posed pictures—and a lot of amateur photographers like snapping quick, candid photos that capture a second of true life.

What face detection does is identify the face of the person in a framed shot, calculate the proper exposure levels to get the best image of the subject’s face in the picture, and then adjusts the flash level accordingly.  It does all this in less than a second, of course.  The result is a photograph where the person’s face is clearly visible in all its lifelike color and detail.

How does a digital compact camera even know part of the framed image is a face in the first place?  The answer is to that is complicated, involving technical computer jargon and mathematical algorithms.  Put simply, a computer program in the digital camera defines a face as a pre-defined set of shapes and colors.  When you frame a shot through the viewfinder, the computer program scans the image looking for the “face pattern” of shapes and colors until it identifies one.  Of course, when you turn your head, the shape changes.  Consequently, digital cameras are best able to identify faces when the subject it looking straight at the camera, and has more trouble with faces in profile or looking up, down, or to the sides. 

Considering how impressive and useful this technology is, one would expect to find it as an advanced, high-end feature in expensive, professional cameras with price tags running into the thousands of dollars.  Surprisingly, this is not the case.  Face detection technology is becoming increasingly common in inexpensive digital compact cameras aimed at the family market.  The average compact camera targeting the amateur family photographer usually runs between $100 and $350 dollars. 

Face detection technology starts becoming commonly available in the midrange cameras around the $200 level, such as the Sony Cybershot series or the Casio Exilim series.  Of course, face detection is becoming a standard element in the more expensive higher-end compact cameras, such as the Canon PowerShot G10 retailing for around $500.  Even some of the inexpensive low-end digital compact cameras come with face-detection technology, such as the Samsung S860, which sells for under $90.