Illumination invariance face recognition using complex wavelet packet transform

  N.Adinarayana,  B.Sreenivasa Ganesh,  N.Durga Parameswara Rao

The illumination is one of the major problems in face recognition systems. The illumination effect is caused due to same face exposed under different lighting variations. In this paper, we propose a novel method for Illumination invariance face recognition called as complex wavelet packet transform. Wavelet is extracting very good features from the images, but it extracts features from the low frequency region only. But the features are also present in the high frequency region. To overcome this limitation going to the wavelet packet transform. In wavelet packet transform, extract features in both low and high frequency regions. In this paper, we concentrate only on high frequency regions, because the features are the strong points that mean high frequency components in the image and each decomposition level we combine all high frequency regions that means detailed coefficients. In wavelet packet based feature extraction not give the directional features for this limitation we use complex wavelet packet transform for getting the directional features. Correlation method is used to find the matching points in the two images. To overcome wrongly correlated point we use Homography estimation by using RANSAC. The experimental result shows that our method is simple comparing dual tree complex wavelet transforms based feature extraction.

Keywords- illumination variation, face recognition, complex wavelet packet transform.

Unique Identification Number - IJEDR1601087
Page Number(s) - 525-528
Pubished in - Volume 4 | Issue 1 | March 2016
DOI (Digital Object Identifier) -    PACE INSTITUTE OF TECHNOLOGY AND SCIENCES
Publisher - IJEDR (ISSN - 2321-9939)

  N.Adinarayana,  B.Sreenivasa Ganesh,  N.Durga Parameswara Rao,   "Illumination invariance face recognition using complex wavelet packet transform", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.4, Issue 1, pp.525-528, March 2016, Available at :http://www.ijedr.org/papers/IJEDR1601087.pdf