tensorflow 2 中的分类对象检测产品计数

问题描述

我们可以在 Tensorflow 2 对象检测 API 类中计算检测到的对象吗？

因为我是新手，所以很难根据我下面描述的用例来操作对象检测模型的输出

假设您在超市货架库存中有番茄和马铃薯两个类别，我想明智地计算每个对象类别

例如;马铃薯数量：5 番茄数量：3。

作为参考，我的以下输出是这样的，只是为了提供一个想法：-

{'raw_detection_scores': array([[3.8237274e-03,3.1729043e-03,5.1983595e-03,...,1.0126382e-02,4.1468740e-03,3.5721064e-03],[3.7932396e-03,1.9723773e-03,2.3661852e-03,3.4036636e-03,9.3266368e-04,4.3996871e-03],[3.2063425e-03,2.9956400e-03,3.9784312e-03,5.5939257e-03,2.3936033e-03,2.7040839e-03],[4.1239262e-03,3.9246678e-04,4.5044391e-05,1.2922287e-04,2.8958917e-04,2.2355914e-03],[1.4656782e-03,4.8859119e-03,1.4899671e-03,2.8479993e-03,2.8250813e-03,2.1298528e-03],[1.8135607e-03,2.2478402e-03,1.1820495e-03,9.5197558e-04,1.3802052e-03,2.2761822e-03]],dtype=float32),'detection_anchor_indices': array([44692.,44710.,44728.,44818.,39652.,44674.,39670.,40036.,40018.,44800.,39634.,39371.,44830.,38090.,44731.,10078.,39796.,27838.,37604.,16933.,24833.,39778.,44659.,45058.,38084.,44791.,44692.,30244.,5284.,38204.,33593.,38192.,37982.,6635.,33118.,24389.,44910.,33112.,39601.,16133.,3845.,39918.,48370.,19204.,44740.,16792.,6629.,25763.,38150.,48187.,15839.,38180.,23524.,44914.,1438.,38078.,35992.,38012.,39888.,44578.,51075.,15833.,37976.,40258.,48751.,39906.,31684.,16453.,38054.,5140.,42568.,36484.,38202.,37946.,14024.,2404.,40002.,5764.,39870.,48823.,26878.,38198.,39430.],'detection_multiclass_scores': array([[1.6726255e-03,3.2217724e-07,2.8865278e-02,2.6032329e-04,2.0583175e-05,4.5886636e-04],[1.0796189e-03,5.8811463e-07,3.6984652e-02,6.6033006e-04,3.2279208e-05,4.3705106e-04],[6.1860681e-04,2.4921978e-06,6.0835034e-02,2.0631850e-03,5.2474130e-05,3.7664175e-04],[2.0163953e-03,1.0121465e-03,1.6601086e-03,5.1327944e-03,1.6998947e-03,9.6607208e-04],[1.2855232e-03,5.4006279e-03,1.0573506e-02,1.3051391e-02,1.0753423e-02,1.3659596e-03],[8.1962347e-04,9.5169604e-01,1.5044212e-03,5.1358938e-03,8.4767938e-03,3.2877922e-04]],'detection_classes': array([4,4,1,6,2,7,9,3,5,8,7]),'detection_boxes': array([[5.92004418e-01,1.69490814e-01,7.45701075e-01,2.46565759e-01],[5.89631081e-01,2.46157080e-01,7.39599228e-01,3.18454713e-01],[5.87109149e-01,3.14503819e-01,7.36972034e-01,3.85336846e-01],[5.87837219e-01,7.05797434e-01,7.28340387e-01,7.74214983e-01],[6.35630414e-02,1.69735521e-01,2.04962432e-01,2.47269154e-01],[5.92036664e-01,9.53008384e-02,7.46890843e-01,1.73706189e-01],[6.85142130e-02,2.45773658e-01,2.10277155e-01,3.21099281e-01],[9.23785418e-02,7.77337551e-01,2.25108251e-01,8.45668435e-01],[9.24619362e-02,7.06092656e-01,2.26126671e-01,7.77547657e-01],[5.85118353e-01,6.37673438e-01,7.26098835e-01,7.05848277e-01],[5.94619289e-02,9.38070714e-02,2.03622460e-01,1.72308475e-01],[3.44502553e-02,7.54546002e-03,2.08990484e-01,9.39401984e-02],[5.87913811e-01,7.74250209e-01,7.28712976e-01,8.34733903e-01],[9.84132528e-01,3.18221241e-01,9.96858120e-01,3.95583898e-01],[5.89539468e-01,3.49524260e-01,7.35065162e-01,4.21008408e-01],[1.87163889e-01,9.88169909e-01,3.71130943e-01,9.98176932e-01],[9.36717317e-02,7.77330160e-01,2.24804163e-01,8.45728278e-01],[6.63008153e-01,9.89469707e-01,8.10642183e-01,1.00000000e+00],[9.70665693e-01,3.16653520e-01,9.95440483e-01,3.85887355e-01],[3.70503038e-01,2.54840344e-01,4.76123840e-01,3.14984292e-01],[5.87433934e-01,7.05650687e-01,7.27492571e-01,7.73511648e-01],[9.28397924e-02,7.06507027e-01,2.26004675e-01,7.77664006e-01],[5.82323313e-01,1.54982358e-02,7.39678025e-01,1.03125945e-01],[5.87077260e-01,7.05565095e-01,7.27602482e-01,7.74259925e-01],[9.83516991e-01,3.11883837e-01,9.97174442e-01,3.89778942e-01],[5.88727355e-01,6.20116591e-01,7.30183959e-01,6.90428734e-01],[5.92004418e-01,[7.42125034e-01,0.00000000e+00,8.81110668e-01,8.84985179e-03],[5.44907227e-02,2.06223458e-01,1.28744319e-02],[9.84484971e-01,5.55115521e-01,9.97788608e-01,6.32665694e-01],[7.99783111e-01,9.75158930e-01,9.83929038e-01,9.97444630e-01],[9.85278428e-01,5.28306305e-01,9.97080624e-01,6.08543932e-01],[9.88123000e-01,9.33963358e-02,9.99226749e-01,1.72955215e-01],[9.08265784e-02,7.78585851e-01,2.25109786e-01,8.43916476e-01],[7.94785440e-01,9.86553550e-01,9.70936120e-01,9.98435020e-01],[5.84929466e-01,7.75964439e-01,7.24675894e-01,8.34971726e-01],3.21474910e-01]],'raw_detection_boxes': array([[-0.0132168,-0.00798112,0.03437265,0.02366759],[-0.01795438,-0.01333077,0.04313567,0.03091241],[-0.00845873,-0.01297706,0.02555573,0.02979016],[-0.01206583,-0.01901898,0.03632494,0.04061931],[-0.01634497,-0.00570066,0.04027664,0.01987169],[-0.02299639,-0.01094626,0.05078602,0.02601441],[-0.01034649,-0.00047059,0.03106559,0.04336115],[-0.01548673,-0.00679935,0.03944379,0.05214766],[-0.00469762,-0.00637354,0.02257038,0.05068764],[-0.00889431,-0.01532986,0.03383063,0.06445184],[-0.01338234,0.00258018,0.03299785,0.03899822],[-0.02030504,-0.00274394,0.04193052,0.04610612],[-0.0114202,0.00825354,0.0315875,0.05609718],[-0.01720474,0.00155611,0.03969076,0.06473814],[-0.0055348,0.00137738,0.02347516,0.06321988],[-0.0093858,-0.00954537,0.03353771,0.0789085 ],[-0.01528691,0.0120711,0.03230394,0.05128276],[-0.02242971,0.00611713,0.04139108,0.0590462 ],[-0.01265933,0.01957938,0.03226281,0.06821183],[-0.0190082,0.01264081,0.04051029,0.07676097],[-0.00625486,0.01262659,0.02384217,0.07535952],[-0.01057751,0.00036938,0.03408406,0.09211845],[-0.01712188,0.02387175,0.03272626,0.0631646 ],[-0.02457684,0.01729448,0.04191976,0.07130254],[-0.01416131,0.03209703,0.03322188,0.08013913],[-0.02092581,0.02524993,0.04159252,0.08845924],[-0.00731821,0.02507119,0.02447667,0.08743346],[-0.01213621,0.01294496,0.03459452,0.10395688],[-0.01857999,0.0361888,0.03388733,0.07542843],[-0.02637036,0.02969162,0.04293538,0.08341235],[-0.01507254,0.04520991,0.03351783,0.09184141],[-0.02222046,0.03861695,0.04212021,0.10008947],[-0.00780608,0.03797973,0.02448018,0.09932629],[-0.01303079,0.02687315,0.03459996,0.1151351 ],[-0.0191509,0.04890272,0.03473954,0.08777986],[-0.02749499,0.04277577,0.04370061,0.09534387],[-0.01489433,0.05867497,0.03314201,0.10344677],[-0.02239214,0.05207732,0.04205906,0.11197228],[-0.00734611,0.05139816,0.02392033,0.11116292],[-0.01289164,0.0412713,0.03449183,0.12679553],[-0.01872004,0.06203329,0.03483813,0.09988385],[-0.02761277,0.05606709,0.04412681,0.10715124],0.2496243 ]],'detection_scores': array([0.9957284,0.9954956,0.9948391,0.9935589,0.9928843,0.9922596,0.99091065,0.9904872,0.9904753,0.9836049,0.97076845,0.76198786,0.11483946,0.08861226,0.06485316,0.06403089,0.06083503,0.05606595,0.05304798,0.05192479,0.05068725,0.0497607,0.04650801,0.04170695,0.04141748,0.0396772,0.03875464,0.03834933,0.03700855,0.03698465,0.03656569,0.03464538,0.03429574,0.03408125,0.033981,0.03356522,0.03337869,0.03140217,0.03058183,0.02957818,0.02886528,0.02712101,0.02674139,0.02655837,0.02634463,0.02611795,0.02595255,0.02580112,0.0251711,0.02473494,0.02423027,0.02406707,0.02352765,0.02347961,0.02342641,0.02327773,0.02312759,0.0229713,0.02272761,0.02240831,0.02240023,0.02203956,0.02200234,0.02167007,0.02112213,0.0210447,0.02079707,0.02007249,0.01999336,0.01993376,0.01986268,0.0196887,0.01967749,0.01877454,0.01874545,0.01856974,0.01855248,0.01853141,0.01839408,0.01838818,0.01830906,0.01829055,0.01759666,0.01758116,0.01747909,0.01745978,0.01728415,0.01719788,0.0171611,0.01715598,0.01704106,0.01684934,0.01672551,0.01663077,0.01645952,0.01627839,0.01607156,0.01592609,0.01579505,0.01570672],'num_detections': 100}

请伙计们帮我解决这个问题 提前致谢

解决方法

查看他的链接：https://www.tensorflow.org/api_docs/python/tf/data/Dataset

在这里，您可以找到如何使用“.as_numpy_iterator()”迭代数据集，以及如何使用不同的方法来操作输入数据集。 希望这会有所帮助。