Abstract [eng] |
This thesis examines facial recognition and illumination normalization technologies. The new and more efficient lighting normalization algorithm, which is implemented by using \"OpenCV\" and \"MS Visual Studio C ++\" software is proposed. The effectiveness of the proposed method is investigated and the optimal PCA, LDA and LBP method parameters are determined. Experimental studies are carried out by using the Multimedia and computer design laboratory face database. The work consists of 5 parts: introduction, face recognition and lighting normalization algorithm analysis, the methodology of proposed method integration with \"OpenCV\" and description of the experimental results. Thesis summary is presented. Work scope - 77 p. of text, 23 pictures, 10 tables, 56 references. |