高次局所自己相関特徴を用いた2方向顔画像認識法の検討

概要

論文の詳細を見る
Recently, biometrics has been used as a password for enhancing security. A variety of biometrics such as fingerprints, voices, and muscular currents are used as individual passwords. The purpose of this study is to develop the software for recognizing individuals by using their facial images. In the first step, a person's picture was taken with a CCD camera. Then, the edge extraction processing and the mosaic processing were given to the image to get the edge information and the grayscale information. After that, the values of the higher-order limited part autocorrelation function were calculated by using these images. Finally, the Euclid distance from the registered dictionary data to the input image was calculated, and the software recognized the images by using the Euclid distance. The recognition rate of the software was 70% of 10 images when the brightness of the images was equalized. And when the images were taken almost at the same time, 90% of 10 images could be recognized in the system. This result indicates that the new technique improves the overall performance of the software for recognizing the facial images effectively.
岐阜工業高等専門学校の論文
2007-03-01