Research On Non-contact Palmprint Images Acquisition And Recognition System

Along with the progress of the society, identity recognition becomes more and more important, at the same time for identification technology requirements are constantly improved. Palm image recognition technology, because its characteristics, rich and stable collection convenient, not easy be replicated, users easy to accept characteristics by unanimous esteem.Through the domestic and international palm image acquisition and recognition system of correlative literature consultation, and on the current status of further analysis. Palm image acquisition from off-line collection developing to online collection, from single feature recogning method to multiple aspects combining direction. But there are collecting in the ways of contact, easy to cause disease crossover, and collecting strict spread. Identification method based on library research from practical mostly gathering and identify requirements.This system of collecting ways for dispute contact. Through the body's blood and skin to absorb and resonance scattering spectra of the analysis and research of the light spectrum selection, this paper uses the 430nm wavelength of blue light LED. Through the light emitting diode (LED) photoelectric characteristics, and multiple LED space and the arrangement of research, the system selects 10 grains of straw hat, with head LED to the positive side 70mm MHD triangle. Using perspective for 120°CMOS camera to collect palm images. In order to validate the rationality of this design, manufacture three distinct illuminant, use evaluation function of four light same palm acquisition of the image definition of the evaluation.This system is based on TM320DM6437, using TVP5146 video decoder chip gain acquisition 23ms high contrast between the palm of the hand form prominent image and texture clear palm of the image. Opponents of fractal image preprocessing prominent palm thereby accurate positioning palmprint effective area. Through the mapping relationship in texture clear palm extract palmprint image characteristics, use gray surface matching method of realization identification. The paper presents the design of the light to the collected palm image mainline texture clearly, image contrast highly and accord with recognition programs use. System identification method is suitable for system itself requirements. The whole system design meet the practical requirements.