The Research Of Pattern Recognition Arithmetic Base On Digital

As the promotion of "Internet +" strategy,and the rapid expansion of the mobile Internet,the mobile terminal based on digital identification technology gradually has applied to every field of daily life,It has become an important branch of pattern recognition.Digital recognition can be applied to the medical treatment,transportation,logistics,finance and other life areas.Algorithms differ in different fields for digit recognition,the corresponding time of storage and identification accuracy is different,there isn't a unified optimal algorithm that can satisfy identification work of different applications and different environment.In this paper,aiming at the above problem,based on the character recognition at home and abroad,handwritten character recognition,pattern recognition and related digital identification algorithms mentioned in the literature,It carries through the induction,analysis,summary,makes every effort to find a relatively optimized and more efficient,more accurate digital recognition algorithm with a wider application field.The research of this paper make a concrete analysis in the needed specific process of realizing digital identification technology,On this basis,it also compares a variety of different algorithms,through simulation experiment,analyzes their respective advantages and disadvantages,improves the existing algorithm,and has implemented a more practical and efficient digital recognition algorithm,and provided the feasibility of the algorithm for relevant applications of digital identification in related fields.For the current feature extraction algorithm and classifier contrast analysis have been conducted,finally the advantages and disadvantages of the above algorithm have been analysed and summarized,the innovation of the new feature extraction algorithm has been proposed and combined with the Euclidean distance,finally we got Euclidean distance of the pattern recognition algorithm based on six line feature extraction.The simulation experiments proved that Euclidean distance number recognition algorithm based on six line feature extraction,which is united by six line feature extraction method and the minimum distance classifier,its operation speed is faster,identification accuracy is higher,the overall design achieved the optima,accords with the requirement of digital recognition,and has achieved thefirst research goal of this article,it also provided the beneficial reference for the application of pattern recognition based on digital.