quanzhi.txt

描述了关于OCR的各种模式识别方法

隐含结点权值:
w[0][0]:11.039247
w[0][1]:8.710914
w[0][2]:15.031054
w[0][3]:-2.288993
w[0][4]:-8.880827
w[0][5]:18.404771
w[0][6]:-15.163876
w[0][7]:7.674364
w[0][8]:0.298254
w[0][9]:10.025167

w[1][0]:15.649881
w[1][1]:-50.077873
w[1][2]:1.779786
w[1][3]:10.391259
w[1][4]:4.907315
w[1][5]:-10.229421
w[1][6]:7.935352
w[1][7]:-10.155918
w[1][8]:17.905507
w[1][9]:-5.333561

w[2][0]:-5.829491
w[2][1]:-0.287583
w[2][2]:-7.958734
w[2][3]:17.423870
w[2][4]:2.458339
w[2][5]:-4.569799
w[2][6]:13.235220
w[2][7]:-0.229848
w[2][8]:11.677962
w[2][9]:-10.950258

w[3][0]:-21.150496
w[3][1]:25.171555
w[3][2]:-51.735313
w[3][3]:3.744356
w[3][4]:-15.293458
w[3][5]:-2.111470
w[3][6]:-24.500488
w[3][7]:-5.503322
w[3][8]:17.240699
w[3][9]:14.068985

w[4][0]:13.978108
w[4][1]:3.392197
w[4][2]:-11.800169
w[4][3]:-5.254983
w[4][4]:21.634837
w[4][5]:9.373469
w[4][6]:16.682957
w[4][7]:-2.583139
w[4][8]:-26.742201
w[4][9]:-1.976853

输出结点权值:
w1[0][0]:10.178228
w1[0][1]:-18.921714
w1[0][2]:-7.710249
w1[0][3]:16.645776

w1[1][0]:1.447877
w1[1][1]:-7.915255
w1[1][2]:9.364420
w1[1][3]:-13.554686

w1[2][0]:18.785592
w1[2][1]:-6.992396
w1[2][2]:8.453642
w1[2][3]:-14.223234

w1[3][0]:29.611479
w1[3][1]:-16.820418
w1[3][2]:17.023352
w1[3][3]:-7.457812

w1[4][0]:-26.665326
w1[4][1]:-23.729949
w1[4][2]:-7.131371
w1[4][3]:0.646520

w1[5][0]:9.968961
w1[5][1]:25.155499
w1[5][2]:-23.901567
w1[5][3]:9.220070

w1[6][0]:10.556350
w1[6][1]:15.369381
w1[6][2]:-7.396715
w1[6][3]:-11.215042

w1[7][0]:-30.225878
w1[7][1]:14.556611
w1[7][2]:5.967444
w1[7][3]:5.693746

w1[8][0]:-13.494102
w1[8][1]:-22.813812
w1[8][2]:-17.052535
w1[8][3]:13.120183

w1[9][0]:1.789480
w1[9][1]:3.095970
w1[9][2]:-11.491727
w1[9][3]:-23.038051