摘要 |
PROBLEM TO BE SOLVED: To provide an image processing method that can reduce computational complexity when detecting a human head area from an identification photograph image. SOLUTION: The method for detecting a head area from an identification photograph image comprises a background color identification step (S102) of identifying a background color in the image, a step (S103) of scanning an upper image part with respect to the background color to detect a human vertex point, a step (S104) of detecting left and right head ends corresponding to the part that has the widest non-background-color area, a step (S105) of scanning a lower image part with respect to the background color to detect the narrowest neck part corresponding to the part that has the narrowest non-background-color area, and a step (S106) of outputting the area enclosed by the detected vertex point, left and right head ends and narrowest neck part as a head area. COPYRIGHT: (C)2005,JPO&NCIPI
|