Регистрация / Вход
Прислать материал


Kumar Vivek
Электронная почта
Информационные технологии
Институт информационных технологий и автоматизированных систем управления (ИТАСУ)
Интеллектуальные системы управления
ФИО научного руководителя
Alexander Nalivayko
Академическая группа
Innovative Software design, Development and Implementation
Наименование тезиса

Objective of Thesis:-

Applicability is easier and working range is larger than other biometric information processing, i.e.; fingerprint, iris scanning, signature, etc. A face recognition system is designed, implemented and tested in this thesis study. The system utilizes a combination of techniques in two topics; face detection and recognition. The face detection is performed on live acquired images without any application field in mind. Processes utilized in the system are white balance correction, skin like region segmentation, facial feature extraction and face image extraction on a face candidate. Then a face classification method namely Principal Component analysis by means of MATLAB is deployed. The system is tested with a database and the number of images are not a specific limit. One can consider as large as database as per requirements and same is implementable for training set also. The tested system has acceptable performance to recognize faces within intended limits. To sensitivity of the algorithm depends upon the threshold value defined by the user, lesser the threshold would be, better the detection performance will be. It is also found that system is also capable of detecting and recognizing multiple faces in live acquired images as an extension work.

Working Process:-

The procedure of matching faces by this algorithm is as follows:-
-In the first step user selects a training image from the dtabase of test images.

-In second step, after running the algorithm, it extracts the featural deatails of of the image, like colour tone, gesture, contrast, position of face part etc.

-In the third step, the binary image based on extracted information  is generated.

-In the fourth step, the algorithms serach for the matching image from the database of different images. Depending upon the the parameters like threshold variation from 10-100 the selection depends. 

Result and Conclusions:- 

The results can be altered by varying the threshold for 10 to 100. This variation declares the selectivity of user on the basis of details of images. More harsh the threshold would be precise would be the result. Future scope of this work is next the system can be developed with webcam and an android device. This system could be useful for a deaf and dumb person carrying an android device or a system connected with webcam. Database of sign gesture is stored into binary form of size 60X80 pixels so that it takes less time and memory space during pattern recognition. 

Scientific adviser -  Alexander Nalivayko