The following content is out of date. [OpenCV-Python Tutorials] at docs.opencv.org/3.1.0/ (http://docs.opencv.org/3.1.0/d6/d00/tutorial_py_root.html#gsc.tab=0) Is accessible.
Also the example script orb = cv2.ORB_create() It is a script that follows the actual binding of OpenCV 3.1.
Directly http://docs.opencv.org/3.1.0/d6/d00/tutorial_py_root.html#gsc.tab=0 Please visit.
OpenCV 3.1 has been released.
dandelion [OpenCV Change Logs] (https://github.com/atinfinity/lab/wiki/OpenCV-Change-Logs%EF%BC%88%E6%84%8F%E8%A8%B3%EF%BC%89#3.1) Has created.
The directory when the archive was expanded opencv\sources\doc\py_tutorials Includes the latest version of OpenCV-Python Tutorials.
Links to previously introduced articles, http://docs.opencv.org/3.0-beta/doc/py_tutorials/py_tutorials.html The linked html is for 3.0-beta, not for OpenCV 3.1. Therefore, the author of OpenCV-Python Tutorials can read the latest modified content. Open the markdown file in opencv \ sources \ doc \ py_tutorials with an editor Good to read.
py_orb.markdown Try to open.
Compare the corresponding files with winMerge () between opencv3.0 and opencv3.1. For 3.1
# Initiate ORB detector
orb = cv2.ORB_create()
# find the keypoints with ORB
kp = orb.detect(img,None)
# compute the descriptors with ORB
kp, des = orb.compute(img, kp)
It was confirmed that the ORB class generation was modified so that the python script in the tutorial and the implementation of the corresponding release match.
Note: The OpenCV-Python Tutorials documentation for OpenCV seems to be generated from a markdown file using Sphinx etc. markdown -> rst -> html
Get started with Markdown slideShare This slide teaches me how to convert from markdown to rst.
Sphinx documentation http://docs.sphinx-users.jp/contents.html Sphinx is a tool to generate html from rst files
Postscript: It seems that the OpenCV documentation is made from a series of builds from CMake, but so far I haven't succeeded in generating the documentation myself in either 3.0 or 3.1. CMake succeeds on windows version of 3.0, but CMake checks BUILD_DOCS and [configure], [generate] succeeds, but I don't know how to generate the above html on windows. ..
If you want to use the latest features of OpenCV, use it in C ++ because there is no OpenCV-Python binding yet. Also, bring opencv_contrib from the GitHub repository and build it additionally (reference article).
According to OpenCV Change Logs Various interesting algorithms are implemented.
The procedure for using opencv_contrib on windows is written. [1]: "OpenCV 3.0 has been released, so I introduced it"
Recommended Posts