Constructor

obj = cv.AffineFeature2D(detector)
obj = cv.AffineFeature2D(detector, extractor)

obj = cv.AffineFeature2D({detector, 'key',val,...})
obj = cv.AffineFeature2D({detector, 'key',val,...}, {extractor, 'key',val,...})

Input

• detector Feature detector. It can be spacified by a string specifying the type of feature detector, such as 'HarrisLaplaceFeatureDetector'. See cv.FeatureDetector.FeatureDetector for possible types.
• extractor Descriptor extractor. It can be specified by a string containing the type of descriptor extractor. See cv.DescriptorExtractor.DescriptorExtractor for possible types.

In the first variant, it creates descriptor detector/extractor of the given types using default parameters (by calling the default constructors).

In the second variant, it creates descriptor detector/extractor of the given types using the specified options. Each algorithm type takes optional arguments. Each of the detector/extractor are specified by a cell-array that starts with the type name followed by option arguments, as in: {'Type', 'OptionName',optionValue, ...}. Refer to the individual detector/extractor functions to see a list of possible options of each algorithm.

When both detector/extractor are passed, it creates an instance wrapping the given keypoint detector and descriptor extractor.

When only the detector is passed, it creates an instance where keypoint detector and descriptor extractor are identical.

Example

detector = cv.AffineFeature2D('HarrisLaplaceFeatureDetector');

detector = cv.AffineFeature2D(...
    {'HarrisLaplaceFeatureDetector', 'NumOctaves',6}, ...
    {'SURF', 'Upright',false});