DCT변환과 AdaBoost 분류기를 이용한 얼굴 검출 알고리즘

• 영문명: Face Detection Algorithm Using the DCT Transformation and the AdaBoost Classifier
• 발행기관: 한국IT마케팅학회
• 저자명: 김현수(Hyun Soo Kim) 권영만(Young Man Kwon)
• 간행물 정보: 『The Journal of Information Technology and Management』Vol.1 No.1, 1~5쪽, 전체 4쪽
• 주제분류: 사회과학 > 무역학
• 파일형태: PDF
• 발행일자: 2013.12.31

영문 초록

In this paper, we proposed the new method for detecting a human face, which uses the DCT method and AdaBoost classifier. We used DCT method for extracting facial characteristics from image and used AdaBoost classifier for detecting the face from image. As a result of experiment, a satisfactory outcome has been obtained as for the front image of a face. However, a problem has been found out if an image is tilted in excess of a certain angle. In future, therefore, studies will be carried out to find out a method for making it possible to detect a face with a much higher accuracy regardless of its tilt angle.

목차

Abstract

1. 서론

2. 관련연구

3. 실험 및 실험 결과

4. 결론

5. References

키워드

Face Detection DCT AdaBoost Feature Pattern Recognition