OpenCV 3.1.0 SVM輪廓特徵的提取和分類 II

版權聲明：本文爲博主原創文章，未經博主允許不得轉載。 https://blog.csdn.net/PinkRiverside/article/details/72615077

對OpenCV 3.1.0 SVM輪廓特徵的提取和分類 I中提取到的輪廓使用OpenCV中的SVM進行分類。

首先對這些輪廓進行裁剪：

Mat getROI(Mat &src){
    int left, right, top, bottom;
    left = src.cols;
    right = 0;
    top = src.rows;
    bottom = 0;


    //得到區域
    for (int i = 0; i<src.rows; i++)
    {
        for (int j = 0; j<src.cols; j++)
        {
            if (src.at<uchar>(i, j) > 0)
            {
                if (j<left) left = j;
                if (j>right) right = j;
                if (i<top) top = i;
                if (i>bottom) bottom = i;
            }
        }
    }
    cout << "rows:" << src.rows << ",cols:" <<src.cols<< endl;
    cout << "left:" << left << ",right:" << right << ",top:" << top << ",bottom:"<<bottom<<endl;
    imshow("b", src);
    Mat dst(bottom - top, right - left, CV_8UC1);
    for (int j = 0; j < bottom - top; j++){
        for (int k = 0; k < right - left; k++){
            dst.at<uchar>(j, k) ='0';
        }
    }
    cout << right - left << "  "<<(int)src.at<uchar>(bottom , right) << endl;
    cout << bottom - top << endl;

    for (int j = 0; j < bottom - top; j++){
        for (int k = 0; k < right - left; k++){
            dst.at<uchar>(j, k) = src.at<uchar>(j+top, k+left);
        }
    }
    cout << "left:" << left << ",right:" << right << ",top:" << top << ",bottom:" << bottom << endl;
    imshow("a", dst);
    return dst;
}

將裁剪後的區域縮放至16*16的大小，準備樣本數據並進行訓練。

int main(int argc, char** argv)
{


    //訓練數據的分類 1.0蘋果-1.0香蕉
    int labels[8] = { 1,1,1,1,-1,-1,-1,-1 };
    Mat labelsMat(8, 1, CV_32SC1, labels);
    //訓練數據矩陣
    float colors[8][16*16];
    //蘋果樣本數據 
    for (int i = 0; i < 4; i++){
        char name[30];
        sprintf(name, "appleCanny/apple%d.jpg", i);

        Mat src = imread(name,-1);
        Mat tmp = getROI(src);
        Mat dst(16, 16, CV_8UC1);
        resize(tmp, dst, dst.size());

        for (int j = 0; j < 16; j++){
            for (int k = 0; k < 16; k++){
                colors[i][j*16+k] = dst.at<uchar>(j, k);
            }
        }
    }
    //香蕉樣本數據
    for (int i = 0; i < 4; i++){
        char name[30];
        sprintf(name, "bananaCanny/banana%d.jpg", i);

        Mat src = imread(name, -1);
        Mat tmp = getROI(src);
        Mat dst(16, 16, CV_8UC1);

        resize(tmp, dst, dst.size());

        for (int j = 0; j < 16; j++){
            for (int k = 0; k < 16; k++){
                colors[i+4][j * 16 + k] = dst.at<uchar>(j, k);
            }
        }
    }
    Mat trainingDataMat(8, 16*16, CV_32FC1, colors);

    //初始化SVM
    Ptr<SVM> svm = SVM::create();
    svm->setType(SVM::Types::C_SVC);
    svm->setKernel(SVM::KernelTypes::LINEAR);
    svm->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, 100, 1e-6));

    //訓練
    svm->train(trainingDataMat,SampleTypes::ROW_SAMPLE , labelsMat);
//保存訓練數據
    svm->save("SVM_CANNY_DATA.xml");
    waitKey();

    return 0;
}