Research On Palmprint Recognition Algorithm Based On Digital Image Technology

In the fast-developing information age,identity recognition technology has transformed from hard-to-use high-end technologies into mature products.Taking payment methods as an example,payment from paper money to electronic money is an improvement,and payment from a password to payment by a fingerprint(face)is another improvement.In view of this,more and more researchers are investing in the research of wider biometric technology.As an important biometric technology,the advantages and disadvantages of palmprint recognition are very clear.Advantages: Like fingerprints,palm print is a human biological feature with superior performance,easy access,obvious classification,and stable information.Disadvantages: Unlike mainstream fingerprint facial recognition,palmprint recognition is late and lacks a comprehensive research system.The physical equipment that has been put into use is often bulky,expensive,and the market penetration rate is obviously not high.Therefore,how to implement a better,more convenient and cheaper palmprint recognition system has become the focus of current researchers.The main contributions and work of this paper are summarized as follows: Based on the morphological characteristics of the palm,a method for obtaining and preprocessing palm print images is proposed,which eliminates the problem of uneven brightness of the original palm image.Further extracted palmprint features in the ROI region;researched palmprint recognition algorithms based on probabilistic templates and fuzzy logic,which can eliminate the rotation interference caused by image collection,and analyzed the extraction,fusion and storage of sample palmprint features.The extracted palmprint feature lines can be stored in a set template library.Finally,the feature matching method based on fuzzy logic is used to determine the category assignment of palmprints by comparing the size of the membership function.The mainline and double-tree complex wavelet are proposed.Hierarchical palmprintrecognition method can effectively remove the influence of some interference options,extract the palmprint main line more completely,and retain the original penetrability to a certain extent;use translation,rotation,and scale-invariant palmprint recognition Algorithm,using DoG operator to locate key corner points for feature extraction,using gradient histogram,Analysis of the direction key palmprint pattern.A multi-modal biometric recognition system based on hand characteristics was completed,and a palmprint authentication system was implemented to verify the accuracy and practicability of the proposed algorithm.To sum up,this paper takes palm print recognition as the research object,based on the existing theoretical basis of biometric recognition technology,combined with the research needs of palm print recognition systems,and takes the main line and double-tree complex wavelet hierarchical palm print recognition method as the core.A multi-modal biometric recognition system based on hand characteristics was constructed and applied to the office attendance system.In addition to developing the system,a sample experiment was also conducted in this research to verify the practical effect of the system,so as to provide some thinking and reference for the research of related practitioners.