Facial Feature Localizer Essay

727 Words 3 Pages
3 Locating Eye Features Using LPT
Eye feature extraction consists of two stages. In the first stage, also known as eye detection, the eye perimeter is extracted from the face image. In the next stage, the interest region has been searched for localizing the feature point(s). Many appropriated algorithms have been proposed in the literatures [16], [17] for eye detection. Here, we concentrate on localizing features in this perimeter. Fig. 5. Result of performing LPT on the left eye image (59×91 pixels). Two maximums are placed on the top and bottom eyelid.
In order to locate the exact iris, first, the eye searching perimeter is extracted from the face image. Afterward, this area is divided into the left and right eye searching
…show more content…
Choose the eigenvector corresponding to the highest eigenvalue and name it “Phigh”. max_e = max(eigenvalue); max_e_p = find(eigenvalue==max_e);
Phigh = P[:, max_e_p];
4. Map eye on the new basis, Phigh, and name it “LPT”.
LPT = PhighT×eye;
5. Calculate derivation of LPT and their extrema. extremum = Ø; f(x) = LPT’; extremum = extremum U {x| f(x)=0};
6. There are three features as follows: eyelid_top = first maximum of extremum set; eyelid_bottom = second maximum of extremum set; iris = (eyelid_top + eyelid_bottom) / 2;
Choice of the interest region and its neighborhood is crucial for the performance of many eye features localization algorithms. If this neighborhood is small enough, a high accuracy is achieved. In contrast, if it is too big, a pixel in the hair, background or glasses can be given an unacceptable accuracy. Furthermore, change of the conditions like illumination and pose can influence on the performance of the system. Nevertheless, the proposed method is free from these limitations. Fig. 6. Five extracted features in the eye searching perimeter.
For instance, we perform LPT on the area which includes left eye, eyebrow, part of ear and hair. The obtained results were very favorable, and besides three previously mentioned features, we could also extract top and bottom of the eyebrow line. The result is evident in Fig. 6.
The proposed method yields low computational cost allowing real time processing. Time complexity

Related Documents