Detects keypoints and computes their descriptors

[keypoints, descriptors] = obj.detectAndCompute(img)
[...] = obj.detectAndCompute(..., 'OptionName',optionValue, ...)

Input

• img Input uint8/uint16/single grayscale image. Internally image is converted to 32-bit floating-point in the [0,1] range.

Output

• keypoints The detected keypoints. A 1-by-N structure array with the following fields:
  • pt coordinates of the keypoint [x,y]
  • size diameter of the meaningful keypoint neighborhood
  • angle computed orientation of the keypoint (-1 if not applicable); it's in [0,360) degrees and measured relative to image coordinate system (y-axis is directed downward), i.e in clockwise.
  • response the response by which the most strong keypoints have been selected. Can be used for further sorting or subsampling.
  • octave octave (pyramid layer) from which the keypoint has been extracted.
  • class_id object class (if the keypoints need to be clustered by an object they belong to).
• descriptors Computed descriptors. Output concatenated vectors of descriptors. Each descriptor is a 64 or 128 element vector, as returned by cv.KAZE.descriptorSize, so the total size of descriptors will be numel(keypoints) * obj.descriptorSize(). A matrix of size N-by-(64/128) of class single, one row per keypoint.

Options

• Mask optional mask specifying where to look for keypoints. Not set by default.
• Keypoints If passed, then the method will use the provided vector of keypoints instead of detecting them, and the algorithm just computes their descriptors.