Yong-hee HONG1, Hwan-ik CHUNG2, Hern-soo HAHN1
1. Dept. of Electronic Engineering, Soongsil University, Seoul 15 6-743, Korea;2. Dept. of Internet Information, Kyungbok College, Pocheon 487-717 Korea
Abstract-The paper proposes a new method of “Separated Same R ectangle Feature (SSRF)” for face detection. Generally, Haar-like feature is u sed to make an Adaboost training algorithm with strong classifier. Haar-like fe ature is composed of two or more attached same rectangles. Inefficiency of the H aar-like feature often results from two or more attached same rectangles. But t he proposed SSRF are composed of two separated same rectangles. So, it is very f lexible and detailed. Therefore it creates more accurate strong classifier than Haar-like feature. SSRF uses integral image to reduce executive time. Haar-lik e feature calculates the sum of intensities of pixels on two or more rectangles. But SSRF always calculates the sum of intensities of pixels on only two rectan gles. The weak classifier of Adaboost algorithm based on SSRF is faster than one based on Haar-like feature. In the experiment, we use 1 000 face images and 1 000 non-face images for Adaboost training. The proposed SSRF shows about 0 .9% higher accuracy than Haar-like features.
Key words-seperated same rectangle feature; Haar-like; discrete adaboost; feature
Manuscript Number: 1674-8042(2010)02-0121-04
dio: 10.3969/j.issn.1674-8042.2010.02.05
References
[1]Paul Viola, Michael J. Jones, 2004. Robust real-time face detection . Inter. Jour. of Computer Vision, 57(2): 137-154.
[2]Rainer Lienhart, Alexander Kuranov, Vadim Pisarevsky. Empirical Anal ysis of Detection Cascades of Boosted Classifiers for Rapid Object Detection. MR L Technical Report, p.297-304.
[3]Luka Furst, Sanja Fidler, Ales∨ Leonardis, 2008. Selec ting features for object detection using an ada boost-compatible evaluation fun ction. Pattern Recognition Letters, 29(11): 1603-1612.
[4]Takeshi Mita, Toshimitsu Kaneko, Osamu Hori, 2005. Joint Haar-like Features for Face Detection. IEEE International Conference on Computer Vision (I CCV’05), p.1619-1626.
[5]Gary Bradski, Adrian Kaehler, Vadim Pisarevsky, 2005. Learning-base d computer vision with intel’s open source computer vision library. In Intel Technologies Journal, 9(2): 121-128.
[6]J. Miteran, J. Matas, E. Bourennane, M. Paindavoine, J. Dubois, 2005 . Automatic hardware implementation tool for a discrete adaboost-based decision algorithm. Eurasip Journal on Applied Signal Processing, p.1035-1046.
[full text view]
