A simple and effective real-time eyes detection human detection without training procedure
SSIP'06 Proceedings of the 6th WSEAS International Conference on Signal, Speech and Image Processing
| Hi-index | 0.00 |
This paper presents an automatic method for facedetection and eye location using a modified version of theALSI Texture Module.ALISA (Adaptive Learning Imageand Signal Analysis) is an adaptive classification enginebased on collective learning systems theory.Using supervised training, the ALISA engine builds a set ofmulti-dimensional feature histograms that estimate thejoint PDF of the feature space for the trained class(es).Inthe current research, 4 to 6 general-purpose texture andcolor features are used, which require only a fewthousands bins (unique feature vectors) to represent facesfor several different ethnic groups by allocating thefeature regions using the ALISA texture module and thenlocates the eyes inside these regions.A preliminarycomparison with a widely-used parametric approach formodeling color information in the presence of changingillumination conditions, demonstrates that the ALISAtexture regions of skin.The proposed method also offers competitive speed and is thus feasible for real-timeapplications to both still images and video sequences