Ethnicity Identification based on Fusion Strategy of Local and Global Features ExtractionDownload PDF
Human facial identification become an active and challenging subject in the area of computer vision, that provides the demographic information such as ethnicity, age, gender, etc. However, even with all the research that has been conducted in ethnic identification, it is still seems to be a difficult and largely problem. In this paper, we propose an approach for robust ethnic identification by fusing two complementary local and global descriptors. In this approach, uniform Local Binary Pattern (ULBP) has been used to extract the local features while Discrete Cosine Transform (DCT) has been performed to extract global features from facial images. The classification of the selected feature vector is computed using the k-Nearest Neighbor (KNN) classifier with the city block distance. We evaluate the performance of the proposed fusion approach, through the extensive experiments conducted on our collected dataset. Our experimental results indicate clearly that the proposed fusion strategy gives a high level of 98.03% ethnicity identification accuracy rate.
Keywords: Fusion strategy, feature extraction, ethnicity identification, uniform local binary pattern(ULBP), discrete cosine transform (DCT), k-nearest neighbor(KNN)