Optimization Design And Implementation Of Paper Money Image Process System Based On Multi-core ARM Processor

Paper money as a type of banknote is widely used in many countries or regions and it has huge circulation and circulation.With the deterioration about counterfeiting of paper money,the traditional financial detection equipments which only use the magnetic signal and the UV characteristic can not meet the demand of identification.The paper money image processing system which is based on the image sensor technology and image processing technology has been widely favored by the market because of handling of paper money rapidly and stably.In the field of banknote security,this system has lots of applications.At present,the paper image processing system is widely used on the financial detection equipments including the clearing machine,the cash register.These machines collect the paper image data through the image acquisition device,and then use the image processing technology to realize the identification of the paper money.The paper image processing system has huge data calculations,complex image processing algorithms and strict real-time requirements so that high-performance DSP embedded solution is adopted.But the DSP solution also brings high commercial cost and development cost.Therefore,the solution which reduces costs and does not affect the real-time character needs to be raised.This paper proposes a solution based on the multi-core ARM processor to fix up above problems.In order to meet the real-time requirement,this solution focuses on solving software optimization problem.Firstly,this paper introduces the design scheme and the key technology of the paper image processing system.Secondly,this paper presents the standard input image and the processing flow.Thirdly,in order to improve the running speed of the system,we put forward different system optimization methods including compiling,multi-threading and OpenMP.Then,we use Qt framework to construct the debug software.Finally,on the basis of implementing the debugging software,different ways of optimization are tested and analyzed.This paper presents the paper money image processing solution based on multi-core ARM processor.Compared with the traditional DSP solutions,the cost is significantly reduced.After optimization,the time coset of paper image processing is reduced from 16.53 ms to 5.32 ms,which satisfies the technical requirements of the Class A currency detector.