Fig 1.
Masks creation for feature extraction.
(a) Landmarks distribution. (b) Mask created from landmarks. (c) Thickened mask. (d) Independent masks for each feature. (e) Right eye. (f) Mirrored left eye. (g) Extracted nose. (h) Original mouth. (i) Shaved mouth.
Fig 2.
Pseudo-code of the algorithm to extract the features from the whole face photographs.
Fig 3.
Dataset of 580 images of eyes obtained using the extraction algorithm.
Table 1.
Percentages of variance explained by 45 eigenfaces for each dataset.
Fig 4.
Original and reconstructed mouths before de-normalization using 45 eigenfaces.
(a) Original mouths. (b) Reconstructed mouths.
Fig 5.
Dunn’s Index and clusters with a single element per number of clusters for white mouths.
Fig 6.
The name, the representative feature, and the membership percentage of each cluster are shown for each ethnic group.
Fig 7.
The name, the representative feature, and the membership percentage of each cluster are shown for each ethnic group.
Fig 8.
The name, the representative feature, and the membership percentage of each cluster are shown for each ethnic group.
Fig 9.
Clusters of Black, White, Latino and Asian eyes.
Fig 10.
Clusters of Black, White, Latino and Asian noses.
Fig 11.
Clusters of Black, White, Latino and Asian mouths.
Table 2.
Number of clusters for each feature for each ethnic group.
Fig 12.
Codification of faces composed using the representative facial features of the most populated clusters for each ethnic group.
Fig 13.
Stages 1 and 2 of the survey procedure.
Table 3.
Results of the validation survey.