Automatic feature detection and age classification of human faces in digital images

A method for automatically categorizing the age of facial images found in digital images using a computer, comprising the steps of:

• (a) detecting a two-dimensional digital image by a computer vision system;
• (b) locating facial features of a human face by:
• (i) dropping and aligning first snakelet-curves to the digital image the snakeletcurves using a Hough Transform voting process to find recordable positions corresponding to a chin and sides of the human face;
• (ii) using a template on the two-dimensional digital image to determine recordable positions to eyes, mouth and nose corresponding to the human face;
• and
• (c) classifying the recordable positions of the chin, the sides of the face, the eyes, the mouth, and the nose of the human face, into one of at least two age categories.

A computer vision detection system for detecting facial images in a two-dimensional digital image and categorizing age ranges of the detected facial image, comprising:

• (a) a video-camera for creating a two-dimensional image containing digital data;
• (b) a computer for reading the digital data to find facial features of a human face;
• (c) a facial feature detection program, for execution in the computer for, dropping and aligning snakelet-curves to the digital image the first snakelet-curves using a Hough Transform voting process to find recordable positions corresponding to a chin and sides of a human face, using a template on the two-dimensional digital image to determine recordable positions to eyes, mouth and nose corresponding to the human face;
• (d) a ratio program, for execution in the computer for determining ratio values between each of the recordable positions;
• and
• (e) a classification threshold program, for execution in the computer for, comparing the ratio values of the human face, to a preselected age category.