Regarding the application of the neural network to automate the processes of human face recognition.
Abstract
The article deals with the use of neural network to automate the processes of human face recognition. It is emphasized that today, it is possible to distinguish at least two broad categories of facial recognition systems: the need to find a person in a large database of persons (for example, in the police database); the need to identify specific people in real time (for example, in a security monitoring system, location tracking system, etc.), or to allow access to a group of people and deny access to everyone else (for example, access to a building, computer, etc.). It is emphasized that when solving various problems, the only stable features of the compared images are contour features. This situation is especially typical for the case of obtaining a photo of the same person in different parts of the electromagnetic spectrum. The algorithm of the Roberts operator or the operator of selection of contour lines 2 × 2 is offered based on an estimation and a choice of fragments of the image with a high gradient level. Describes the operator of the selection of contour lines 3 × 3 when estimating the value of the gradient of a particular image element takes into account the influence of eight adjacent elements. The essence of the Canney algorithm is given. It is noted that the sequential application of the mask filter of the Canney operator and the statistical filter of the search box managed to generate a bitmap image of the image while preserving the contours of the human face, which significantly improved the filtering results in contrast to the standard comparison with the threshold value. It is emphasized that the detection of boundaries occurs when determining the local maximum and minimum gradient of the brightness of the object. The block diagram of a high level system for face recognition is given and the principle of operation of the system is described. It is emphasized that full automation of the process of recognizing a person's face is quite possible, but requires an additional mechanism to eliminate possible errors at the stage of filtering the contour, which is the basis for further research
References
Grother Р. Face Recognition Vendor Test (FRVT). Performance of Face Identification Algorithms. / Patrick Grother, Mei Ngan. – Information Access Division National Institute of Standards and Technology. – May 26, 2014 – р. 138.
M. Lades, J. Vorbruggen, J. Buhmann, “Distortion invariant object recognition in the dynamic link architecture”, IEEE Transactions on computers, 1993, vol. 42, no. 3, pp. 300 -310, March 1993
L.Juwei, N. P.Konstantinos, A. Venetsanopoulos, “Face recognition using kernel direct discriminant analysis algorithms”, IEEE Transactions On Neural Networks, vol.14, no. 1, pp.117–126, January 2003.
M. Lades, J. Vorbruggen, J. Buhmann, “Distortion invariant object recognition in the dynamic link architecture”, IEEE Transactions on computers, 1993, vol. 42, no. 3, pp. 300 -310, March 1993
P. Viola, “Robust realtime face detection”, International Journal of Computer Vision, 2004, vol. 57, no. 2, pp. 137-154, 2004
S.Lawrence, C.L. Giles., C. Tsoiта, “Face Recognition: A Convolutional Neural Network Approach”, IEEE Transactions on Neural Networks, Special Issue on Neural Networks and Pattern Recognition, vol. 8, no 1, pp.98–113, 1997.
Y.Taigman, M.Yang, M.Ranzato, “DeepFace: Closing the Gap to Human-Level Performance in Face Verification”
Joo Er Meng, W.Chen, Wu Shiqian, “High-speed face recognition based on discrete cosine transform and RBF neural networks”, IEEE Transactions on Neural Networks, vol. 16, no. 3, pp. 679 – 691,2005
Abstract views: 0 PDF Downloads: 0
