| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide an image processing method, system, and program for improving accuracy and robustness in discriminating the shape of the predetermined object included in an image. <P>SOLUTION: When the facial frame model is configured by deforming the average frame model Sav in such a way that an average frame model Sav is set in an actual facial image and the position of each landmark on the average frame model Sav is moved to match the position of the corresponding landmark detected from the face, the positions of the points showing the landmarks are detected from the facial image by using the discriminators and the discriminating conditions for each discriminator obtained through learning in accordance with a machine learning method the luminance profiles at a plurality of points known as predetermined landmarks on a sample image and the luminance profiles of a plurality of points known as not the landmarks on the sample image. <P>COPYRIGHT: (C)2006,JPO&NCIPI |
