| Vein recognition technology has become a research hotspot in the field of biometrics due to its high accuracy,safety,convenience and fast recognition speed.A complete biometrics system should have the function of in vivo identification.The study of in vivo identification parameters of vein recognition is of great significance to the security of its identification.According to the principle of vein recognition,due to the absorption effect of hemoglobin on a specific band of near-infrared light(850nm),a darker line will appear in the subcutaneous vein.At the same time,hemoglobin is closely related to the blood oxygen saturation,so it is proposed to detect the blood oxygen saturation in human body.The detection of oxygen saturation can provide in vivo identification for the existing biometrics.The combination of blood oxygen in vivo detection index and vein recognition technology will inevitably increase the security of vein recognition.A multi-wavelength reflection non-contact oxygen detection system is designed.Current problems of blood oxygen detection and vein recognition technology:(1)the accuracy of blood oxygen detection based on imaging photoelectricplethometry(IPPG)technology is low,and the range of detection is small;(2)non-contact detection of blood oxygen is susceptible to interference by ambient light;(3)vein recognition technology requires in vivo assistance.Considering the above technical problems,this paper takes IPPG technology as the core technology of blood oxygen saturation detection,selects three wavelengths of 810nm,850nm and 660nm,adopts non-contact blood oxygen detection method,and realizes the stable in vivo auxiliary function of dorsal hand vein recognition system under complex ambient light.Specifically,the main research contents of this paper are as follows:1.Analyzed the main research framework of in vivo detection and further analyzes the application of in vivo detection technology in the field of dorsal hand vein recognition.This paper analyzes,classifies and summarizes the current spoofing techniques.The model of combining multi-wavelength blood oxygen detection with non-contact physiological parameter detection is perfected to improve the stability of non-contact detection,and then a sequential infrared image optical detection method is proposed to provide in vivo assistance for vein recognition technology by detecting human blood oxygen saturation,heart rate and other vital physiological parameters.2.Improve the model by combining multi-wavelength blood oxygen detection with non-contact physiological parameter detection to improve the accuracy of non-contact detection.3.Designed and developed the prototype of multi-wavelength reflection non-contact blood oxygen detection system.In terms of hardware,the tower construction of the system platform is completed through light source module,signal control module and signal acquisition module.In the software part,the distribution of light is tracked and analyzed by Zemax simulation,and the uniformity can reach more than 75%.In terms of signal processing,on the premise of eliminating high frequency components and baselines based on Mallat wavelet decomposition and reconstruction method,a combined method of the empirical mode decomposition method(EMD)algorithm and adaptive filtering was proposed to extract signal waveforms with high signal noise.4.Experiments were designed to prove the feasibility,anti-interference and repeatability of the system,and the results proved the feasibility of non-contact blood oxygen detection in the dorsal hand region.Under different light intensity,the process of obtaining R value from the dorsal hand shows that the extraction of R value under ambient light has little influence.At the same time,the system is used to detect the blood oxygen value under different light intensity,and it is found that it has good anti-interference. |
PDF Full Text Request