继续坑自己，EmguCV之SVM.Train（二）

最新推荐文章于 2022-05-08 09:15:07 发布

southmay

最新推荐文章于 2022-05-08 09:15:07 发布

阅读量2.3k

点赞数 1

分类专栏： EmguCV C#

本文链接：https://blog.csdn.net/u010347277/article/details/69101259

版权

本文探讨了在使用EmguCV的SVM进行训练时遇到的问题，发现训练后的SVM模型与HOGDetector加载效果不一致。通过分析源代码，找到了将SVM结果转换为HOGDetector识别因子的过程，并解释了3780到3781的转变。然而，即使解决了识别率问题，高误报率（90%以上）依然存在，作者正在寻找解决方案。此外，文章还介绍了如何提取样本进行训练。

摘要由CSDN通过智能技术生成

想到哪写到哪，继续。

这是续上一篇的问题，训练了SVM，然后直接通过hog.SetSVMDetector加载进来，

发现内容有点奇怪，为什么这个值就是3781，而我通过HOG.compute计算出来的矩阵只有3780？而且检测的结果基本为0？

耍了点小心机，把OpenCV中的代码抠出来，嵌到自己的XML中：（这段代码是直接写在openCV的源代码中，具体在哪个文件我也忘了，可以自行搜索）

<support_vectors>
    <_>
      0.05359386  -0.14721455  -0.05532170  0.05077307 
           0.11547081  -0.04268804  0.04635834  -0.05468199  0.08232084 
           0.10424068  -0.02294518  0.01108519  0.01378693  0.11193510 
           0.01268418  0.08528346  -0.06309239  0.13054633  0.08100729 
           -0.05209739  -0.04315529  0.09341384  0.11035026  -0.07596218 
           -0.05517511  -0.04465296  0.02947334  0.04555536 
           -3.55954492e-003  0.07818956  0.07730991  0.07890715  0.06222893 
           0.09001380  -0.03574381  0.03414327  0.05677258  -0.04773581 
           0.03746637  -0.03521175  0.06955440  -0.03849038  0.01052293 
           0.01736112  0.10867710  0.08748853  3.29739624e-003  0.10907028 
           0.07913758  0.10393070  0.02091867  0.11594022  0.13182420 
           0.09879354  0.05362710  -0.06745391  -7.01260753e-003 
           5.24702156e-003  0.03236255  0.01407916  0.02207983  0.02537322 
           0.04547948  0.07200756  0.03129894  -0.06274468  0.02107014 
           0.06035208  0.08636236  4.53164103e-003  0.02193363  0.02309801 
           0.05568166  -0.02645093  0.04448695  0.02837519  0.08975694 
           0.04461516  0.08975355  0.07514391  0.02306982  0.10410084 
           0.06368385  0.05943464  4.58420580e-003  0.05220337  0.06675851 
           0.08358569  0.06712101  0.06559004  -0.03930482  -9.15936660e-003 
           -0.05897915  0.02816453  0.05032348  0.06780671  0.03377650 
           -6.09417039e-004  -0.01795146  -0.03083684  -0.01302475 
           -0.02972313  7.88706727e-003  -0.03525961  -2.50397739e-003 
           0.05245084  0.11791293  -0.02167498  0.05299332  0.06640524 
           0.05190265  -8.27316567e-003  0.03033127  0.05842173 
           -4.01050318e-003  -6.25105947e-003  0.05862958  -0.02465461 
           0.05546781  -0.08228195  -0.07234028  0.04640540  -0.01308254 
           -0.02506191  0.03100746  -0.04665651  -0.04591486  0.02949927 
           0.06035462  0.02244646  -0.01698639  0.01040041  0.01131170 
           0.05419579  -0.02130277  -0.04321722  -0.03665198  0.01126490 
           -0.02606488  -0.02228328  -0.02255680  -0.03427236 
           -7.75165204e-003  -0.06195229  8.21638294e-003  0.09535975 
           -0.03709979  -0.06942501  0.14579427  -0.05448192  -0.02055904 
           0.05747357  0.02781788  -0.07077577  -0.05178314  -0.10429011 
           -0.11235505  0.07529039  -0.07559302  -0.08786739  0.02983843 
           0.02667585  0.01382199  -0.01797496  -0.03141199  -0.02098101 
           0.09029204  0.04955018  0.13718739  0.11379953  1.80019124e-003 
           -0.04577610  -1.11108483e-003  -0.09470536  -0.11596080 
           0.04489342  0.01784211  3.06850672e-003  0.10781866 
           3.36498418e-003  -0.10842580  -0.07436839  -0.10535070 
           -0.01866805  0.16057891  -5.07316366e-003  -0.04295658 
           -5.90488780e-003  8.82003549e-003  -0.01492646  -0.05029279 
           -0.12875880  8.78831954e-004  -0.01297184  -0.07592774 
           -0.02668831  -6.93787413e-004  0.02406698  -0.01773298 
           -0.03855745  -0.05877856  0.03259695  0.12826584  0.06292590 
           -4.10733931e-003  0.10996531  0.01332991  0.02088735  0.04037504 
           -0.05210760  0.07760046  0.06399347  -0.05751930  -0.10053057 
           0.07505023  -0.02139782  0.01796176  2.34400877e-003  -0.04208319 
           0.07355055  0.05093350  -0.02996780  -0.02219072  0.03355330 
           0.04418742  -0.05580705  -0.05037573  -0.04548179  0.01379514 
           0.02150671  -0.02194211  -0.13682702  0.05464972  0.01608082 
           0.05309116  0.04701022  1.33690401e-003  0.07575664  0.09625306 
           8.92647635e-003  -0.02819123  0.10866830  -0.03439325 
           -0.07092371  -0.06004780  -0.02712298  -7.07467366e-003 
           -0.01637020  0.01336790  -0.10313606  0.04906582  -0.05732445 
           -0.02731079  0.01042235  -0.08340668  0.03686501  0.06108340 
           0.01322748  -0.07809529  0.03774724  -0.03413248  -0.06096525 
           -0.04212124  -0.07982176  -1.25973229e-003  -0.03045501 
           -0.01236493  -0.06312395  0.04789570  -0.04602066  0.08576570 
           0.02521080  0.02988098  0.10314583  0.07060035  0.04520544 
           -0.04426654  0.13146530  0.08386490  0.02164590  -2.12280243e-003 
           -0.03686353  -0.02074944  -0.03829959  -0.01530596  0.02689708 
           0.11867401  -0.06043470  -0.02785023  -0.04775074  0.04878745 
           0.06350956  0.03494788  0.01467400  1.17890188e-003  0.04379614 
           2.03681854e-003  -0.03958609  -0.01072688  6.43705716e-003 
           0.02996500  -0.03418507  -0.01960307  -0.01219154 
           -4.37000440e-003  -0.02549453  0.02646318  -0.01632513 
           6.46516960e-003  -0.01929734  4.78711911e-003  0.04962371 
           0.03809111  0.07265724  0.05758125  -0.03741554  0.01648608 
           -8.45285598e-003  0.03996826  -0.08185477  0.02638875 
           -0.04026615  -0.02744674  -0.04071517  1.05096330e-003 
           -0.04741232  -0.06733172  8.70434940e-003  -0.02192543 
           1.35350740e-003  -0.03056974  -0.02975521  -0.02887780 
           -0.01210713  -0.04828526  -0.09066251  -0.09969629  -0.03665164 
           -8.88111943e-004  -0.06826669  -0.01866150  -0.03627640 
           -0.01408288  0.01874239  -0.02075835  0.09145175  -0.03547291 
           0.05396780  0.04198981  0.01301925  -0.03384354  -0.12201976 
           0.06830920  -0.03715654  9.55848210e-003  5.05685573e-003 
           0.05659294  3.90764466e-003  0.02808490  -0.05518097  -0.03711621 
           -0.02835565  -0.04420464  -0.01031947  0.01883466 
           -8.49525444e-003  -0.09419250  -0.01269387  -0.02133371 
           -0.10190815  -0.07844430  2.43644323e-003  -4.09610150e-003 
           0.01202551  -0.06452291  -0.10593818  -0.02464746  -0.02199699 
           -0.07401930  0.07285886  8.87513801e-004  9.97662079e-003 
           8.46779719e-003  0.03730333  -0.02905126  0.03573337  -0.04393689 
           -0.12014472  0.03176554  -2.76015815e-003  0.10824566  0.05090732 
           -3.30179278e-003  -0.05123822  5.04784798e-003  -0.05664124 
           -5.99415926e-003  -0.05341901  -0.01221393  0.01291318 
           9.91760660e-003  -7.56987557e-003  -0.06193124  -2.24549137e-003 
           0.01987562  -0.02018840  -0.06975540  -0.06601523  -0.03349112 
           -0.08910118  -0.03371435  -0.07406893  -0.02248047  -0.06159951 
           2.77751544e-003  -0.05723337  -0.04792468  0.07518548 
           2.77279224e-003  0.04211938  0.03100502  0.05278448  0.03954679 
           -0.03006846  -0.03851741  -0.02792403  -0.02875333  0.01531280 
           0.02186953  -0.01989829  2.50679464e-003  -0.10258728 
           -0.04785743  -0.02887216  3.85063468e-003  0.01112236 
           8.29218887e-003  -0.04822981  -0.04503597  -0.03713100 
           -0.06988008  -0.11002295  -2.69209221e-003  1.85383670e-003 
           -0.05921049  -0.06105053  -0.08458050  -0.04527602 
           8.90329306e-004  -0.05875023  -2.68602883e-003  -0.01591195 
           0.03631859  0.05493166  0.07300330  5.53333294e-003  0.06400407 
           0.01847740  -5.76280477e-003  -0.03210877  4.25160583e-003 
           0.01166520  -1.44864211e-003  0.02253744  -0.03367080  0.06983195 
           -4.22323542e-003  -8.89401045e-003  -0.07943393  0.05199728 
           0.06065201  0.04133492  1.44032843e-003  -0.09585235  -0.03964731 
           0.04232114  0.01750465  -0.04487902  -7.59733608e-003  0.02011171 
           0.04673622  0.09011173  -0.07869188  -0.04682482  -0.05080139 
           -3.99383716e-003  -0.05346331  0.01085723  -0.03599333 
           -0.07097908  0.03551549  0.02680387  0.03471529  0.01790393 
           0.05471273  9.62048303e-003  -0.03180215  0.05864431  0.02330614 
           0.01633144  -0.05616681  -0.10245429  -0.08302189  0.07291322 
           -0.01972590  -0.02619633  -0.02485327  -0.04627592 
           1.48853404e-003  0.05514185  -0.01270860  -0.01948900  0.06373586 
           0.05002292  -0.03009798  8.76216311e-003  -0.02474238 
           -0.05504891  1.74034527e-003  -0.03333667  0.01524987  0.11663762 
           -1.32344989e-003  -0.06608453  0.05687166  -6.89525274e-004 
           -0.04402352  0.09450210  -0.04222684  -0.05360983  0.01779531 
           0.02561388  -0.11075410  -8.77790991e-003  -0.01099504 
           -0.10380266  0.03103457  -0.02105741  -0.07371717  0.05146710 
           0.10581432  -0.08617968  -0.02892107  0.01092199  0.14551543 
           -2.24320893e-003  -0.05818033  -0.07390742  0.05701261 
           0.12937020  -0.04986651  0.10182415  0.05028650  0.12515625 
           0.09175041  0.06404983  0.01523394  0.09460562  0.06106631 
           -0.14266998  -0.02926703  0.02762171  0.02164151 
           -9.58488265e-004  -0.04231362  -0.09866509  0.04322244 
           0.05872034  -0.04838847  0.06319253  0.02443798  -0.03606876 
           9.38737206e-003  0.04289991  -0.01027411  0.08156885  0.08751175 
           -0.13191354  8.16054735e-003  -0.01452161  0.02952677  0.03615945 
           -2.09128903e-003  0.02246693  0.09623287  0.09412123  -0.02924758 
           -0.07815186  -0.02203079  -2.02566991e-003  0.01094733 
           -0.01442332  0.02838561  0.11882371  7.28798332e-003  -0.10345965 
           0.07561217  -0.02049661  4.44177445e-003  0.01609347  -0.04893158 
           -0.08758243  -7.67420698e-003  0.08862378  0.06098121  0.06565887 
           7.32981879e-003  0.03558407  -0.03874352  -0.02490055 
           -0.06771075  0.09939223  -0.01066077  0.01382995  -0.07289080 
           7.47184316e-003  0.10621431  -0.02878659  0.02383525  -0.03274646 
           0.02137008  0.03837290  0.02450992  -0.04296818  -0.02895143 
           0.05327370  0.01499020  0.04998732  0.12938657  0.09391870 
           0.04292390  -0.03359194  -0.06809492  0.01125796  0.17290455 
           -0.03430733  -0.06255233  -0.01813114  0.11726857  -0.06127599 
           -0.08677909  -0.03429872  0.04684938  0.08161420  0.03538774 
           0.01833884  0.11321855  0.03261845  -0.04826299  0.01752407 
           -0.01796414  -0.10464549  -3.30041884e-003  2.29343961e-004 
           0.01457292  -0.02132982  -0.02602923  -9.87351313e-003 
           0.04273872  -0.02103316  -0.07994065  0.02614958  -0.02111666 
           -0.06964913  -0.13453490  -0.06861878  -6.09341264e-003 
           0.08251446  0.15612499  2.46531400e-003  8.88424646e-003 
           -0.04152999  0.02054853  0.05277953  -0.03087788  0.02817579 
           0.13939077  0.07641046  -0.03627627  -0.03015098  -0.04041540 
           -0.01360690  -0.06227205  -0.02738223  0.13577610  0.15235767 
           -0.05392922  -0.11175954  0.02157129  0.01146481  -0.05264937 
           -0.06595174  -0.02749175  0.11812254  0.17404149  -0.06137035 
           -0.11003478  -0.01351621  -0.01745916  -0.08577441  -0.04469909 
           -0.06106115  0.10559758  0.20806813  -0.09174948  7.09621934e-004 
           0.03579374  0.07215115  0.02221742  0.01827742  -7.90785067e-003 
           0.01489554  0.14519960  -0.06425831  0.02990399  -1.80181325e-003 
           -0.01401528  -0.04171134  -3.70530109e-003  -0.09090481 
           0.09520713  0.08845516  -0.02651753  -0.03016730  0.02562448 
           0.03563816  -0.03817881  0.01433385  0.02256983  0.02872120 
           0.01001934  -0.06332260  0.04338406  0.07001807  -0.04705722 
           -0.07318907  0.02630457  0.03106382  0.06648342  0.10913180 
           -0.01630815  0.02910308  0.02895109  0.08040254  0.06969310 
           0.06797734  6.08639978e-003  4.16588830e-003  0.08926726 
           -0.03123648  0.02700146  0.01168734  -0.01631594  4.61015804e-003 
           8.51359498e-003  -0.03544224  0.03571994  4.29766066e-003 
           -0.01970077  -8.79793242e-003  0.09607988  0.01544222 
           -0.03923707  0.07308586  0.06061262  1.31683104e-004 
           -7.98222050e-003  0.02399261  -0.06084389  -0.02743429 
           -0.05475523  -0.04131311  0.03559756  0.03055342  0.02981433 
           0.14860515  0.0176