Finger Veins Feature Extraction Algorithm Based on Image Processing Methods

摘要

Recently more interest in the recognition algorithms based on human veins is observable. In the literature we can find results confirm that this trait provide huge accuracy level. This feature is used for instance in cash machines. In the last years, more financial institutions took into consideration vein-based identification technology. Its popularity is connected with ease of use and analyzed trait uniqueness. A method to extract finger veins features with image processing algorithms is presented in this paper. In the preliminary stage of the research, the device to collect finger veins images was created. The second part of the work is implementation of the algorithm to process input images. The authors used soft computing algorithm that is artificial neural network to find specific structures on the image. The last stage of the work is connected with confirmation of the results obtained with artificial neural network.