FACE IMAGE CANDIDATE AREA SEARCH METHOD, FACE IMAGE CANDIDATE AREA SEARCH SYSTEM, AND FACE IMAGE CANDIDATE AREA SEARCH PROGRAM

摘要

Internal portions of a search objective image G are successively selected for each of the predetermined areas and the image characteristic vector in the selected area is generated. After this, the image characteristic vector is inputted to a support vector machine (30) which has learned the image characteristic vector of the sample images for learning and it is judged whether a face image exists in the selected area according to the positional relationship with respect to the identification hyperplane. Thus, it is possible to rapidly and accurately search the area having the high possibility of existence of a personal face image in the search objective image G.