本文整理汇总了C++中Tensor::IDFT方法的典型用法代码示例。如果您正苦于以下问题：C++ Tensor::IDFT方法的具体用法？C++ Tensor::IDFT怎么用？C++ Tensor::IDFT使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类Tensor的用法示例。
在下文中一共展示了Tensor::IDFT方法的1个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的C++代码示例。
示例1: main
int main(void)
{
Tensor<double,1> ts("/home/guoxin/Projects/MTC/data/texture1.png");
//ts.Display();
Tensor<double,2> cts = ts.ToComplex();
double time = (double)getTickCount();
Tensor<double,2> f;
for (int i=0; i<1000; i++)
{
f = cts.DFT();
cout<<i<<endl;
}
f = f.IDFT();
time = 1000*((double)getTickCount() - time)/getTickFrequency();
cout<<"time = " <<time<<"ms"<<endl;
Mat temp = f.GetFrame(0);
vector<Mat> tv;
cv::split(temp,tv);
Tensor<double,1>(tv[0]).Display();
Cube c(0,0,0,10,10,1);
cout<<c.area()<<endl;
Steerable sp;
sp.buildSCFpyr(f,3,4,1,false);
vector<Tensor<double,2> >& pyr = sp.getSpaceDomainPyr();
cv::split(pyr[12],tv);
Tensor<double,1>(tv[0]).Display();
Tensor<double,1> ts2("/home/guoxin/Projects/MTC/data/texture2.png");
cout<<ComputeMSE(ts,ts2)<<endl;
ts2.Print("t2",true);
//test LocalMean
Mat flatker = Mat::ones(Size(16,16),CV_64F)/256;
//verify local mean
BufferGPU gbuf;
Size3 subWin(16,16,1);
auto mu1 = ts.LocalMean(flatker,subWin);
auto mu2 = ts.LocalMeanGPU(flatker,gbuf,subWin);
(mu1-mu2).Print();
mu1.Print();
auto var1 = ts.LocalVariance(mu1,flatker,subWin);
auto var2 = ts.LocalVarianceGPU(mu2,flatker,gbuf,subWin);
var1.Print();
var2.Print();
(var1-var2).Print();
// mylib::DisplayMat(flatker);
int TIMES=10000;
time = (double)getTickCount();
for (int i=0; i< TIMES; i++)
{
Tensor<double,1> mu=ts.LocalMean(flatker,Size3(16,16,1));
ts.LocalVariance(mu,flatker,Size3(16,16,1));
}
time = 1000*((double)getTickCount() - time)/getTickFrequency();
time /= TIMES;
cout << "Time of CPU (averaged for " << TIMES << " runs): " << time << " milliseconds."<<endl;
time = (double)getTickCount();
for (int i=0; i< TIMES; i++)
{
Tensor<double,1> mu=ts.LocalMeanGPU(flatker,gbuf,Size3(16,16,1));
ts.LocalVarianceGPU(mu,flatker,gbuf,Size3(16,16,1));
}
time = 1000*((double)getTickCount() - time)/getTickFrequency();
time /= TIMES;
cout << "Time of GPU (averaged for " << TIMES << " runs): " << time << " milliseconds."<<endl;
return 0;
}