OpenCV中神经网络介绍与使用

一：神经网络介绍

人工神经网络(ANN) 简称神经网络(NN)，最早它的产生跟并行计算有关系，主要是学习生物神经元互联触发实现学习、完成对输入数据的分类与识别。最基本的单元是神经元，有一个输入值，一个输出值，神经元本身根据激活函数来说决定输出值，最简单例子就是感知器

上述在开始的时候通过随机初始化生成权重，然后通过对数据X的训练迭代更新权重直到收敛，过程表示如下：

上述就是最简单的单个感知器工作原理。而在实际情况下，神经网络会有多个感知器，多个层级，我们把输入数据X的层称为输入层，最终输出结果的层称为输出层，中间各个层级统统称为隐藏层。一个典型的多层感知器(MLP)网络如下：

这个时候我们选择的激活函数就不能选择简单的二分类函数，OpenCV中支持的激活函数有三个：

上述网络中的权重值是未知的，只有通过训练我们才可以得到这些权重值，生成可用网络模型，OpenCV中支持的两种训练算法分别是：

反向传播算法
RPROP算法

二：OpenCV中创建神经网络

首先创建多层感知器的层数：

Mat_<int> layerSizes(1, 3);

layerSizes(0, 0) = data.cols;

layerSizes(0, 1) = 20;

layerSizes(0, 2) = responses.cols;

上面几行代码是创建一个三层的感知器，输入层跟数据维度有关系，隐藏层有20个神经元、最后是输出层，一般是类别表示。

Ptr<ANN_MLP> network = ANN_MLP::create();

network->setLayerSizes(layerSizes);

network->setActivationFunction(ANN_MLP::SIGMOID_SYM, 0.1, 0.1);

network->setTrainMethod(ANN_MLP::BACKPROP, 0.1, 0.1);

上述代码是创建神经网络，设置层数、激活函数、训练方法等参数。

Ptr<TrainData> trainData = TrainData::create(data, ROW_SAMPLE, responses);network->train(trainData);

上述代码是创建训练数据，执行网络训练

三：代码演示

OpenCV3.4中的sample的代码演示如下：

#include <opencv2/ml/ml.hpp>


using namespace std;

using namespace cv;

using namespace cv::ml;


int main()

{

    //create random training data

    Mat_<float> data(100, 100);

    randn(data, Mat::zeros(1, 1, data.type()), Mat::ones(1, 1, data.type()));


    //half of the samples for each class

    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;

    */


    //create the neural network

    Mat_<int> layerSizes(1, 3);

    layerSizes(0, 0) = data.cols;

    layerSizes(0, 1) = 20;

    layerSizes(0, 2) = responses.cols;


    Ptr<ANN_MLP> network = ANN_MLP::create();

    network->setLayerSizes(layerSizes);

    network->setActivationFunction(ANN_MLP::SIGMOID_SYM, 0.1, 0.1);

    network->setTrainMethod(ANN_MLP::BACKPROP, 0.1, 0.1);

    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;

}

四：基于神经网络的实现mnist数据集训练

Mat train_images = readImages(0);

normalize(train_images, train_images, -1.0, 1.0, NORM_MINMAX, -1);

Mat train_labels = readLabels(0);

printf("n read mnist train dataset successfully...n");

Mat response = Mat::zeros(Size(10, train_labels.rows), CV_32FC1);

for (int i = 0; i < train_labels.rows; i++) {

    int digit = train_labels.at<int>(i, 0);

    response.at<float>(i, digit) = 1;

}


//create the neural network

Mat_<int> layerSizes(1, 3);

layerSizes(0, 0) = train_images.cols;

layerSizes(0, 1) = 100;

layerSizes(0, 2) = 10;


Ptr<ANN_MLP> network = ANN_MLP::create();

network->setLayerSizes(layerSizes);

network->setActivationFunction(ANN_MLP::SIGMOID_SYM, 0.1, 0.1);

network->setTrainMethod(ANN_MLP::BACKPROP, 0.1, 0.1);

network->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, 1000, 1e-6));

Ptr<TrainData> trainData = TrainData::create(train_images, ROW_SAMPLE, response);

printf("start network trainning...n");

network->train(trainData);

if (network->isTrained())

{

    printf("ready to save network model data...n");

    network->save("D:/vcprojects/images/mnist/ann_knowledge.yml");

}


test_ann_minist();

waitKey(0);

return 0;