OpenCV自带神经网络接口初探

Opencv 神经网络API

neural_network.cpp

#include <opencv2/ml/ml.hpp>

using namespace std;
using namespace cv;
using namespace cv::ml;

int main()
{
    //制作随机训练数据;
    Mat_<float> data(100, 100);
    randn(data, Mat::zeros(1, 1, data.type()), Mat::ones(1, 1, data.type()));

    //每个类别(预测的标签)平分一半的数据
    Mat_<float> responses(data.rows, 2);
    for (int i = 0; i<data.rows; ++i)
    {
        if (i < data.rows/2)
        {
            responses(i, 0) = 1;
            responses(i, 1) = 0;
        }
        else
        {
            responses(i, 0) = 0;
            responses(i, 1) = 1;
        }
    }

    /*
    //example code for just a single response (regression)
    Mat_<float> responses(data.rows, 1);
    for (int i=0; i<responses.rows; ++i)
        responses(i, 0) = i < responses.rows / 2 ? 0 : 1;
    */

    //创建神经网络;
    Mat_<int> layerSizes(1, 3);
    layerSizes(0, 0) = data.cols;       //输入层
    layerSizes(0, 1) = 20;              //隐含层
    layerSizes(0, 2) = responses.cols;  //输出层

    Ptr<ANN_MLP> network = ANN_MLP::create();
    network->setLayerSizes(layerSizes);
    network->setActivationFunction(ANN_MLP::SIGMOID_SYM, 0.1, 0.1);  //用sigmoid函数作激活函数;
    network->setTrainMethod(ANN_MLP::BACKPROP, 0.1, 0.1);            //用bp神经网络作训练;
    Ptr<TrainData> trainData = TrainData::create(data, ROW_SAMPLE, responses);

    network->train(trainData);
    if (network->isTrained())
    {
        printf("Predict one-vector:\n");
        Mat result;
        network->predict(Mat::ones(1, data.cols, data.type()), result);
        cout << result << endl;

        printf("Predict training data:\n");
        for (int i=0; i<data.rows; ++i)
        {
            network->predict(data.row(i), result);
            cout << result << endl;
        }
    }

    return 0;
}

编译

root@master:# g++ `pkg-config --cflags opencv` -o neural_network neural_network.cpp `pkg-config --libs opencv`

输出

root@master:# ./neural_network 
Predict one-vector:
[0.49815065, 0.49921253]
Predict training data:
[0.49815512, 0.4992235]
[0.49815148, 0.4992148]
[0.49816421, 0.49922886]
[0.49815488, 0.49921444]
[0.4981553, 0.49921337]
[0.49815708, 0.49922335]
[0.49815473, 0.49921861]
[0.49815747, 0.49921554]
[0.49815911, 0.49922082]
[0.4981544, 0.49921104]
[0.49815312, 0.49920845]
[0.49815834, 0.4992134]
[0.4981578, 0.499221]
[0.49815252, 0.49921796]
[0.49815798, 0.49920991]
[0.4981553, 0.49921247]
[0.49815372, 0.49922264]
[0.49815732, 0.49922493]
[0.49815607, 0.49922168]
[0.49814662, 0.49921048]
[0.49815735, 0.49921829]
[0.49815637, 0.4992227]
[0.49815381, 0.49921602]
[0.49815139, 0.49921939]
[0.49815428, 0.49921471]
[0.4981533, 0.49922156]
[0.49815425, 0.49922514]
[0.49815568, 0.4992153]
[0.49816135, 0.49922994]
[0.49815443, 0.49921793]
... ...