Research of Image Preprocessing Methods for Enhancement of Information Technology for Person Identification

Research presented in this paper describes the experiments conducted on image preprocessing methods, such as anisotropic diffusion and image histogram equalization, with an aim to enhance the performance of the information technology for person identification based on Gabor wavelet transform, local binary patterns in 1-dimensional space and histogram of oriented gradients for image processing, feature vector extraction, and classification. The experimental research of the algorithm was performed on three face image databases with the use of program software created in Python. Experiments were divided into three sets: the original set of the methods without preprocessing, set with the method of anisotropic diffusion, set with the method of image histogram equalization, and set with both these methods. Analysis of obtained results of the experimental research has shown that chosen methods haven’t provided the improvement of the algorithm, therefore other image preprocessing methods must be considered for further exploring.