做行人检测时用SVM训练模型，读取文件时老是出现错误

#include <opencv.hpp>
#include <iostream>
#include <fstream>
#include <string>
#include <vector>
using namespace cv;
using namespace cv::ml;
using namespace std;
#define CENTRAL_CROP true
#define NUMPOS 2416//正样本
#define NUMNEG 12180//负样本
#define NUMHARD 0//负样本个数
int main()
{
int DescriptorDim;
string ImgName;
ifstream pos("POS.txt");
ifstream neg("NEG.txt");
Mat sampleFeatureMat;
Mat sampleabelMat;
for (int num = 0; num < NUMPOS&&getline(pos, ImgName); num++)
{
cout << "处理:" << ImgName << endl;
Mat src = imread(ImgName);
if (CENTRAL_CROP)
{
src = src(Rect(16, 16, 64, 128));
}
HOGDescriptor *hog = new HOGDescriptor(cvSize(64, 128), cvSize(16, 16), cvSize(8, 8), cvSize(8, 8), 9);
vector<float>descriptor;
hog->compute(src, descriptor, Size(8, 8));
if (0 == num)
{
DescriptorDim = descriptor.size();
sampleFeatureMat = Mat::zeros(NUMPOS + NUMNEG + NUMHARD, DescriptorDim, CV_32SC1);
sampleabelMat = Mat::zeros(NUMPOS + NUMNEG + NUMHARD, 1, CV_32SC1);
}
for (int i = 0; i < DescriptorDim; i++)
{
sampleFeatureMat.at<float>(num, i) = descriptor[i];
sampleabelMat.at<float>(num, 0) = 1;
}
}
for (int num = 0; num < NUMNEG&&getline(neg, ImgName); num++)
{
cout << "处理:" << ImgName << endl;
Mat src = imread(ImgName);
HOGDescriptor *hog = new HOGDescriptor(cvSize(64, 128), cvSize(16, 16), cvSize(8, 8), cvSize(8, 8), 9);
vector<float>descriptor;
hog->compute(src, descriptor, Size(8, 8));
for (int i = 0; i < DescriptorDim; i++)
{
sampleFeatureMat.at<float>(num + NUMNEG, i) = descriptor[i];
sampleabelMat.at<float>(num + NUMNEG, 0) = -1;
}
}
ofstream fout("SampleFeatureMat.txt");
for(int i=0; i<NUMPOS+NUMNEG;i++)
{
fout<<i<<endl;
for(int j=0; j<DescriptorDim; j++)
fout<<sampleFeatureMat.at<float>(i,j)<<" ";
fout<<endl;
}
cout << "Starting training process" << endl;
Ptr<SVM> svm = SVM::create();
CvTermCriteria criteria = cvTermCriteria(CV_TERMCRIT_ITER + CV_TERMCRIT_EPS, 1000, FLT_EPSILON);
svm->setType(SVM::Types::C_SVC);
svm->setKernel(SVM::KernelTypes::LINEAR);
svm->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, 100, 1E-6));
Ptr<TrainData> tData = TrainData::create(sampleFeatureMat, ROW_SAMPLE, sampleabelMat);
svm->train(tData);
cout << "训练完成" << endl;
svm->save("E:\\视觉检测\\SVM.xml");
system("pause");
}