Opencv--ann Neural Network

Ann--artificial Neural Networks Artificial Neural network

//defining artificial Neural networksCVANN_MLP BP; //Set up bpnetwork ' s parametersCvann_mlp_trainparamsparams; params. train_method=Cvann_mlp_trainparams::backprop; params. bp_dw_scale=0.1; params. bp_moment_scale=0.1; //Params.train_method=cvann_mlp_trainparams::rprop; //params.rp_dw0 = 0.1; //Params.rp_dw_plus = 1.2; //Params.rp_dw_minus = 0.5; //params.rp_dw_min = Flt_epsilon; //Params.rp_dw_max =.;

Two methods of training: Backprop and Rprop

Two parameters of Backprop:

Four parameters of Rprop:

//Training Data    floatlabels[3][5] = {{0,0,0,0,0},{1,1,1,1,1},{0,0,0,0,0}}; Mat Labelsmat (3,5, CV_32FC1, labels); floattrainingdata[3][5] = { {1,2,3,4,5},{111, the,113, the, the}, { +, A, at, -, -} }; Mat Trainingdatamat (3,5, CV_32FC1, trainingdata);//layersizes set up a network structure with three hidden layers: input layer, three hidden layer, output layer. Both input and output layer nodes are 5, with two nodes per layer in the middle of the hidden layerMat layersizes= (mat_<int> (1,5) <<5,2,2,2,5);//Create the second parameter can set the activation function of each nerve node, the default is Cvann_mlp::sigmoid_sym, that is, the SIGMOID function//Other activation functions provided are Gauss (Cvann_mlp::gaussian) and step Functions (cvann_mlp::identity). 

Bp.create (layersizes,cvann_mlp::sigmoid_sym);   //
params);

// predicting new nodes Mat Samplemat = (mat_<float> (1,5) << i,j,0,0,0 );              Mat Responsemat;              Bp.predict (Samplemat,responsemat);  

Full code:

#include <opencv2/core/core.hpp>#include<opencv2/highgui/highgui.hpp>#include<opencv2/ml/ml.hpp>#include<iostream>#include<string>using namespacestd; using namespaceCV; intMain () {CVANN_MLP BP; Cvann_mlp_trainparamsparams; params. Train_method=cvann_mlp_trainparams::backprop;//(back PROPAGATION,BP) reverse propagation algorithm    params. bp_dw_scale=0.1; params. bp_moment_scale=0.1;     floatlabels[Ten][2] = {{0.9,0.1},{0.1,0.9},{0.9,0.1},{0.1,0.9},{0.9,0.1},{0.9,0.1},{0.1,0.9},{0.1,0.9},{0.9,0.1},{0.9,0.1}}; //Here the sample is labeled 0.1 and 0.9 instead of 0 and 1, mainly considering that the output of the sigmoid function is usually between 0 and 1, and only when the input approaches-∞ and +∞ is gradually approaching 0 and 1, and it is impossible to achieve. Mat Labelsmat (Ten,2, CV_32FC1, labels); floattrainingdata[Ten][2] = { { One, A},{111, the}, { +, A}, {211,212},{Wuyi, +}, { in, the}, {441,412},{311,312}, { A, +}, {Bayi, the} }; Mat Trainingdatamat (Ten,2, CV_32FC1, Trainingdata); Mat layersizes= (mat_<int> (1,5) <<2,2,2,2,2);//Layer 5: Input layer, 3 layer hidden layer and output layer, each layer is two perceptronBp.create (LAYERSIZES,CVANN_MLP::SIGMOID_SYM); Bp.train (Trainingdatamat, Labelsmat, Mat (), Mat (),params);intwidth = +, height = +; Mat Image=Mat::zeros (height, width, cv_8uc3); VEC3B Green (0,255,0), Blue (255,0,0);      for(inti =0; i < image.rows; ++i) { for(intj =0; J < Image.cols; ++j) {Mat Samplemat= (mat_<float> (1,2) <<i,j);              Mat Responsemat;              Bp.predict (Samplemat,responsemat); float* p=responsemat.ptr<float> (0); //            if(p[0] > p[1]) {image.at<Vec3b> (j, i) =Green; }             Else{image.at<Vec3b> (j, i) =Blue; }        }      }    //Show the training data    intThickness =-1; intLinetype =8; Circle (Image, point (111, the),5, Scalar (0,0,0), thickness, linetype); Circle (Image, point (211,212),5, Scalar (0,0,0), thickness, linetype); Circle (Image, point (441,412),5, Scalar (0,0,0), thickness, linetype); Circle (Image, point (311,312),5, Scalar (0,0,0), thickness, linetype); Circle (Image, point ( One, A),5, Scalar (255,255,255), thickness, linetype); Circle (Image, point ( +, A),5, Scalar (255,255,255), thickness, linetype); Circle (Image, point (Wuyi, +),5, Scalar (255,255,255), thickness, linetype); Circle (Image, point ( in, the),5, Scalar (255,255,255), thickness, linetype); Circle (Image, point ( A, +),5, Scalar (255,255,255), thickness, linetype); Circle (Image, point (Bayi, the),5, Scalar (255,255,255), thickness, linetype); Imwrite ("Result.png", image);//Save the imageImshow ("BP Simple Example", image);//Show it to the userWaitkey (0); return 0;} 

Opencv--ann Neural Network