EmguCV检测图像模糊

Question

我试图使用EmguCV检测图像的模糊。现在我试图做一个灰色图像的FFT，但我被卡住了，因为我的程序在调用CvInvoke.cvCopy后挂起。见下：

_capturedFrames.ForEach(x => 
{ 
    using(var gray = x.Convert<Gray, float>()) 
    { 
     IntPtr complexImage = CvInvoke.cvCreateImage(gray.Size, IPL_DEPTH.IPL_DEPTH_32F, 2); 

     CvInvoke.cvSetZero(complexImage); // Initialize all elements to Zero 
     CvInvoke.cvSetImageCOI(complexImage, 1); 
     CvInvoke.cvCopy(gray.Ptr, complexImage, IntPtr.Zero); 
     CvInvoke.cvSetImageCOI(complexImage, 0); 

     var dft = new Matrix<float>(gray.Rows, gray.Cols, 2); 
     CvInvoke.cvDFT(complexImage, dft, CV_DXT.CV_DXT_FORWARD, 0); 

     double min; 
     double max; 
     Point minLoc; 
     Point maxLoc; 
     dft.MinMax(out min, out max, out minLoc, out maxLoc); 

     if (max > overallMax) 
     { 
      overallMax = max; 
      index = _capturedFrames.IndexOf(x); 
     } 

     CvInvoke.cvReleaseImage(ref complexImage); 
    } 
}); 

上面的代码的目的是循环浏览内存中的图像集合，并找到最清晰的图像的索引。让我们假装上面的方法正在工作......这是一个检测最清晰图像的强大方法吗？ FFT是否可靠用于此目的？

Answer 1

你们都知道了。请参阅下面更正的代码。请注意gray.NumberOfChannels属性。

_capturedFrames.ForEach(x => 
{ 
    using(var gray = x.Convert<Gray, float>()) 
    { 
     IntPtr complexImage = CvInvoke.cvCreateImage(gray.Size, IPL_DEPTH.IPL_DEPTH_32F, gray.NumberOfChannels); 

     CvInvoke.cvSetZero(complexImage); // Initialize all elements to Zero 
     //CvInvoke.cvSetImageCOI(complexImage, 1); 
     CvInvoke.cvCopy(gray.Ptr, complexImage, IntPtr.Zero); 
     //CvInvoke.cvSetImageCOI(complexImage, 0); 

     var dft = new Matrix<float>(gray.Rows, gray.Cols, gray.NumberOfChannels); 
     CvInvoke.cvDFT(complexImage, dft, CV_DXT.CV_DXT_FORWARD, 0); 

     double min; 
     double max; 
     Point minLoc; 
     Point maxLoc; 
     dft.MinMax(out min, out max, out minLoc, out maxLoc); 

     if (max > overallMax) 
     { 
      overallMax = max; 
      index = _capturedFrames.IndexOf(x); 
     } 

     CvInvoke.cvReleaseImage(ref complexImage); 
    } 
});