features2d classes in OpenCV(DRAFT)

OpenCV implements various key point detectors and descriptors as subclasses of features2d class.

But it is not clear at once whether such a subclass can detect key points and compute descriptors or can only one of them.

The below list summarizes such distinction, plus some minor notes.

• Features including No. 10(AffineFeature2D) and below are implemented in xfeatures2d class in the opencv contrib module.
• Source path is relative to "opencv/modules/features2d".
• 
  

No	Class name	detect	compute	detectAndCompute	Source	Notes
1	AgastFeatureDetector	o	x	x	agast.cpp	transforms to gray image.
2	AKAZE	x	x	o	akaze.cpp	Depends on the channels?
3	BRISK	x	x	o	brisk.cpp	transforms to gray image.
4	FastFeatureDetector	o	x	x	fast.cpp
5	GFTTDetector	o	x	x	gftt.cpp
6	KAZE	x	x	o	kaze.cpp	transforms to gray image.
7	MSER	o	x	x	mser.cpp	implementation differes fro gray and color image.
8	ORB	x	x	o	orb.cpp	transforms to gray image. based on FastFeatureDetector
9	SimpleBlobDetector	o	x	x	blobdetector.cpp	transforms to gray image.
10	AffineFeature2D
11	BoostDesc	x	o	x	boostdesc.cpp	transforms to gray image.
12	BriefDescriptorExtractor
13	DAISY	x	o	x	daisy.cpp
14	FREAK	x	o	x	freak.cpp
15	HarrisLaplaceFeatureDetector
16	LATCH	x	o	x	latch.cpp

https://docs.opencv.org/3.0-beta/doc/py_tutorials/py_feature2d/py_table_of_contents_feature2d/py_table_of_contents_feature2d.html
https://github.com/opencv/opencv/tree/master/modules/features2d/src

https://docs.opencv.org/4.1.0/d0/d13/classcv_1_1Feature2D.html#a8be0d1c20b08eb867184b8d74c15a677

"This algorithm is patented and is excluded in this configuration; Set OPENCV_ENABLE_NONFREE CMake option and rebuild the library in function 'create'"