Method of numerical analysis of similarity and differences of face shape of twins
M Rychlik, W Stankiewicz, M Morzynski - 13th International Conference on …, 2009 - Springer
M Rychlik, W Stankiewicz, M Morzynski
13th International Conference on Biomedical Engineering: ICBME 2008 3–6 …, 2009•SpringerThis article presents application of the PCA (Principal Component Analysis) method for
analysis and computation of three dimensional biometric description of 3D objects. As the
data input the geometrical data sets (threedimensional points coordinates) of faces are
used. Authors apply 3D version of PCA method for numerical estimation of similarity and
extraction of difference of analyzed faces. PCA decompose set of 3D objects into mean face
and individual features (empirical modes), which describing deviations from mean value …
analysis and computation of three dimensional biometric description of 3D objects. As the
data input the geometrical data sets (threedimensional points coordinates) of faces are
used. Authors apply 3D version of PCA method for numerical estimation of similarity and
extraction of difference of analyzed faces. PCA decompose set of 3D objects into mean face
and individual features (empirical modes), which describing deviations from mean value …
Abstract
This article presents application of the PCA (Principal Component Analysis) method for analysis and computation of three dimensional biometric description of 3D objects. As the data input the geometrical data sets (threedimensional points coordinates) of faces are used.
Authors apply 3D version of PCA method for numerical estimation of similarity and extraction of difference of analyzed faces. PCA decompose set of 3D objects into mean face and individual features (empirical modes), which describing deviations from mean value. Obtained mean shape describe the similarity of faces, eigenmodes present geometrical differences between faces. Eigenvalues can be used to numerical (mathematical) comparison of study faces.
In this paper three sets of data of different type of twins (identical — monozygotic, fraternal-dizygotic) and thirteen typical faces are used and compared. The mean face and the features (eigenfaces) are presented and discussed.
Authors propose using the set of eigenfaces and corresponding coefficient values (computed from PCA) for security verification. As an example of authorization key the set of coefficient values for the faces are presented. Each key describes individual shape of face and can be decoded and compared with the original data of user to obtain access to restricted area or data.
Springer
以上显示的是最相近的搜索结果。 查看全部搜索结果