《一种基于全局阈值二值化方法的BP神经网络车牌字符识别系统_张坤艳》由会员分享，可在线阅读，更多相关《一种基于全局阈值二值化方法的BP神经网络车牌字符识别系统_张坤艳（4页珍藏版）》请在金锄头文库上搜索。
1、 CN43-1258/TPISSN 1007-130X9SCOMPUTER ENGINEERING & SCIENCE2010M322 Vol1 32,No1 2, 2010 cI|:1007-130X(2010)02-0088-03B=ZEBP*3MYd*A Plate-Character Identification System Based onGloba-l ValveBinarization and theBP Neural Networkf1 ,R2 ,1 ,I1ZHANG Kun-yan1 , ZHONG Y-i ya2 , MIAO Song-chi1 , WANG Gu-i
2、juan1(1.yv,62250101;2.65asOqaim4rXMYrTb+,BP*T3MY,P3MY,88*l:2008-09-19;:2008-12-13Te:f(1977 ),o,Q,=,Z_?EbY:250101862g7?uyv;Tel:13296403668;E-mail:Address:Schoolof InformationandElectrical Engineering,Shangdong Jianzhu University,Fengming Rd,Lingang Development Zone,Jinan,Shandong 250101,P.R.China7O1a
3、?;*y3MYd,?Y|1XTb2 d!9Z,sLVLMY#VbLV,n5|Vs3m),B,mqBqb4MYrT,emMY,1m=)bBBV?SMYb|V)sF,sYkbBP*L,r!,kkbTkT?L=1p,5?*bYVk*VmMYd3MYbVm1Ubm1 LVi,*VMYd,*HqM7M,1mdMa;vMbf/,MY1a?MYddmbm2LMY1aVmbm2 LMY#V3MYdTV,VYV1ZT|BtF,TLkL1a*bLk-,n5|-T*!z,i!z*L,V73MY*b-,d*OMbTVL,51a9Cb3 ELC3. 1 =)V4A,=|mMm,m=ca,m0255bm=1oa,P3?u
4、s7,=MTmA1!z,3-bbMYd1p)aav,=m),?vv4)rqb=|ZE,1s:EaEEbZE,MY+,Em=)b=ZEmbWsB,iLCm=mbK=|/:(1)|mC#Cqbb9C#CQbpCQs1(q)b=FBib,=ibCq,PPix (0, i)ibmi, PPix(1, i)ibiCqb(2)p|fsF( i) :F( i) = Eij = 0PPix(0, j )* PPix (1, j) (1), iV07Sm9b(3)p|-iCqPSum(i) :PSum(i) = Eij= 0PPix (1,j ) (2), iV07Sm9b(4)p8m(AGray ,mp,9b
5、(5)p|/W Valve( i) :WValve( i) = AGray* F( i) - PSum(i)F( i)* 1- F(i) (3), iV07Sm9b(6)pKvWValve( i)iTK=b3. 2 mjS#BBsm3vlivs,M,dBj3MYS.F+3MYJ.NNV,2006,27(7):698-701.2 C,.s3MY J.S,2002,15(4):75-77.3 O,f.*E9M.:bv, 2002:17-28.4 y.*L M.s:v,2001.5 f. Visual C+ +3mTMY/#LlM.:,2003.6 V,.*ZEJ.v(1S),2003,24(1):35-38.7 f,f.3s-m)J.9S,2008,30(4):40-42.134
