Compares a template against overlapped image regions

result = cv.matchTemplate(image, tmpl)
result = cv.matchTemplate(image, tmpl, 'OptionName', optionValue, ...)

Input

• image Image where the search is running. It must be 8-bit integer or 32-bit floating-point.
• tmpl Searched template. It must be not greater than the source image and have the same data type.

Output

• result Map of comparison results. It is single-channel 32-bit floating-point. If image is W x H and templ is w x h, then result is (W-w+1) x (H-h+1).

Options

• Method Parameter specifying the comparison method, default 'SqDiff'. One of the following:
  • SqDiff Squared difference
  • SqDiffNormed Normalized squared difference
  • CCorr Cross correlation
  • CCorrNormed Normalized cross correlation
  • CCoeff Cosine coefficient
  • CCoeffNormed Normalized cosine coefficient
• Mask Mask of searched template. It must have the same datatype and size with templ. Currently, only the SqDiff and CCorrNormed methods are supported. It is not set by default.

The function slides through image, compares the overlapped patches of size w x h against templ using the specified method and stores the comparison results in result.

Here are the formulae for the available comparison methods (I denotes image, T template, R result). The summation is done over template and/or the image patch: x'=0..w-1, y'=0..h-1:

• SqDiff:

  R(x,y) = sum_{x',y'} (T(x',y') - I(x+x',y+y'))^2
  

• SqDiffNormed:

  R(x,y) = sum_{x',y'} (T(x',y') - I(x+x',y+y'))^2 /
           sqrt(sum_{x',y'} (T(x',y')^2) * sum_{x',y'} (I(x+x',y+y')^2))
  

• CCorr:

  R(x,y) = sum_{x',y'} (T(x',y') * I(x+x',y+y'))
  

• CCorrNormed:

  R(x,y) = sum_{x',y'} (T(x',y') * I(x+x',y+y')) /
           sqrt(sum_{x',y'} (T(x',y')^2) * sum_{x',y'} (I(x+x',y+y')^2))
  

• CCoeff:

  R(x,y) = sum_{x',y'} (T'(x',y') * I'(x+x',y+y')), where
  
  T'(x',y') = T(x',y') - 1 / (w*h) * sum_{x'',y''} T(x'',y'')
  I'(x+x',y+y') = I(x+x',y+y') - 1 / (w*h) * sum_{x'',y''} I(x+x'',y+y'')
  

• CCoeffNormed:

  R(x,y) = sum_{x',y'} (T'(x',y') * I'(x+x',y+y')) /
           sqrt(sum_{x',y'} (T'(x',y')^2) * sum_{x',y'} (I'(x+x',y+y')^2))
  

After the function finishes the comparison, the best matches can be found as global minimums (when 'SqDiff' was used) or maximums (when 'CCorr' or 'CCoeff' was used) using the min and max functions. In case of a color image, template summation in the numerator and each sum in the denominator is done over all of the channels and separate mean values are used for each channel. That is, the function can take a color template and a color image. The result will still be a single-channel image, which is easier to analyze.