Learn something new every day More Info... by email
Image processing face recognition is a computerized technique that uses an algorithm to locate and recognize a face in an image, and this technology has several uses. While there are many different facial recognition algorithms available, most programs use edge or eye detection to locate a face. Not only does image processing face recognition find a face, but most algorithms also show the user the exact pixel location of the face within the image, making it easier to find a certain face in a crowded or blurred image. In the entertainment industry, facial recognition is used for face tracking in motion capture, which is used for more realistic animation. Another common use for this technique is in facial recognition security; the algorithm is typically specialized in this case to recognize only certain faces.
Creating an image processing face recognition algorithm can be done in many ways, but the two most common components are edge and eye detection. Edge detection is usually supplementary, because it is typically not as reliable as eye detection. With edge detection, the algorithm searches for edges and curves that are common on the human face. Eye detection looks for eyes, though some algorithms may only be able to recognize them if the image shows a person looking forward with both eyes exposed.
Aside from finding a face, this recognition algorithm usually displays the exact pixel coordinates of the face. For most users, especially in clear photos in which the face is easily found, this may not be very important. If users are searching through blurry and cluttered images, then knowing the exact pixel location can be important in finding and identifying a person in the image.
Another use for face recognition algorithms is in entertainment, especially with motion capture and performance capture animation. In this technique, actors and actresses act out scenes in front of a blue screen as they read their lines. This allows facial and body movements to be captured, which makes it easier to model the movements and increase realism. The actors and actresses wear special electrodes on their faces, and a facial recognition algorithm tracks all facial movements used for expressions. This algorithm is specialized to translate real-life facial images into animated images.
Specialized image processing face recognition algorithms are used in security systems that depend on authorizing visitors by their face. Unlike other algorithms, which just find a face, these algorithms are programmed to find a certain face. Typically, these algorithms have to find the common elements of a certain person's face — such as the eyes, jaw, ears and cheekbones — and then compare the face to one in its image database. If the faces are the same, then the visitor is granted access.