A.D. Varlamov - Ph.D. (Eng.), Vladimir State University named after Alexander and Nikolay Stoletovs. E-mail: varlamov_aleks@mail.r

Among the biometric characteristics of a person, used to identify an individual, particularly noteworthy face. Since the biometric features are unique, it is possible to achieve a very high reliability of the identification, provided accurate estimates of their values. Singularity of face details segmentation (as the face segmentation) task is that, unlike many other segmentation tasks, these objects are arranged in certain areas over the boundaries of a rectangular face region. Therefore, among the main image features, which are used for face segmentation is considered a priori information about its location. This information should be used in a segmentation algorithm as an additional feature to make a decision for each pixel of segmented image. Experimental work on assessing the impact of a priori information about the location of the person on the accuracy of its segmentation used 146 images. Formal neural network is selected to implement the machine learning. The characteristic values obtained by treating a large number of labeled images are used for training of neural network. The results show that the system is automatically selected faces do not differ much from the results of the expert work, and the feature exclusion from the test set has led to unsatisfactory results. Similar results were obtained after segmentation of objects: lips, pupils, nostrils, chin and eyebrows. Thus, we have experimentally proved the necessity of using a priori information about the objects location to assess face biometric features.

 

1. Varlamov A.D., Makarova E.E. Metod avtomaticheskogo vydelenija obektov na UZI snimkakh serdca // Algoritmy, metody i sistemy obrabotki dannykh. 2011. Vyp. 2(17). S. 49-54.
2. Gorbacevich V.S., Vizilter JU.V., ZHeltov S.JU., KHain S.A. Algoritmy predobrabotki i postobrabotki dannykh dlja biometricheskikh sistem raspoznavanija lic // Vestnik kompjuternykh i informacionnykh tekhnologijj. 2014.  № 6(120). S. 9-14.
3. Dmitriev D.V., Kapranov S.N., Markov E.V. Issledovanie algoritmov predvaritelnojj obrabotki biometricheskikh obrazov dlja celejj verifikacii // Nejjrokompjutery: razrabotka, primenenie. 2014. № 3. S. 52-55.
4. Koljuchkin V.JA., Nguen K.M., CHan T.KH. Algoritmy obrabotki izobrazhenijj v sistemakh mashinnogo zrenija robotizirovannykh proizvodstvennykh linijj //Nejjrokompjutery: razrabotka, primenenie. 2014. № 3. S. 44-51.
5. Noskov A.A., Aminova E.A., Priorov A.L., Trapeznikov I.N. Sistema opredelenija povtornogo detektirovanija obektov na osnove nejjronnojj seti // Nejjrokompjutery: razrabotka, primenenie. 2014. № 3. S. 36-43.
6. KHrulev A. Sistemy raspoznavanija lic. Sostojanie rynka. Perspektivy razvitija // Sistemy bezopasnosti. № 1, 2012.
7. CHizhov V.S., Kovalev JU.A. , Varlamov A.D. Razrabotka metoda povyshenija kachestva poiska lic na izobrazhenijakh analizom ikh biometricheskikh priznakov // Algoritmy, metody i sistemy obrabotki dannykh. 2014. Vyp. 2(27).  S. 55-63.
8. SHarapov R.V., Varlamov A.D.Sravnitelnyjj analiz sistem poiska graficheskikh dannykh // Sovremennye naukoemkie tekhnologii. 2013. № 1. S. 27-31.
9. Li H., Toh K.-A., Li L. Advanced Topics in Biometrics. World Scientific. 2012.500 p.