I tried to install (many many times) OpenCV 3.0 for python with extra package (sift, surf...) but I always fails, I really get stuck. I tried in main environment then in virtual ones,
Here is what I did:
cd git
git clone https://github.com/Itseez/opencv_contrib.git
cd ..
wget https://github.com/Itseez/opencv/archive/3.0.0-beta.zip
unzip 3.0.0-beta.zip
cd opencv-3.0.0-beta/
mkdir release
cd release/
workon OCR
cmake -D CMAKE_BUILD_TYPE=Release -D CMAKE_INSTALL_PREFIX=/home/jbd/src/opencv-3.0.0b -D OPENCV_EXTRA_MODULES_PATH=/home/jbd/git/opencv_contrib/modules -D BUILD_opencv_python3=ON -D PYTHON2_EXECUTABLE=/home/jbd/.virtualenv/OCR/bin/python -D PYTHON_INCLUDE_DIR=/home/jbd/.virtualenv/OCR/include/python2.7 -D PYTHON_LIBRARY=/usr/lib/libpython2.7.so -D PYTHON2_NUMPY_INCLUDE_DIRS=/home/jbd/.virtualenv/OCR/local/lib/python2.7/site-packages/numpy ..
make -j7
make install
cd ~/.virtualenv/OCR/lib/python2.7/site-packages/
ln -s /home/jbd/src/opencv-3.0.0b/lib/python2.7/site-packages/cv2.so
Whatever the way I try to install it, I always get:
Traceback (most recent call last): File "/home/jbd/git/ocr/test.py", line 10, in sift = cv2.xfeatures2d.SIFT() AttributeError: 'module' object has no attribute 'SIFT'
with:
import numpy as np
import cv2
sift = cv2.xfeatures2d.SIFT()
If someone see where I'm wrong...
Thanks a lot
SIFT and SURF are examples of algorithms that OpenCV calls “non-free” modules.
>>> help(cv2.xfeatures2d)
Help on module cv2.xfeatures2d in cv2:
NAME
cv2.xfeatures2d
FILE
(built-in)
FUNCTIONS
SIFT_create(...)
SIFT_create([,nfeatures[,nOctaveLayers[,contrastThreshold[,edgeThreshold[,sigma]]]]) -> retval
SURF_create(...)
SURF_create([,hessianThreshold[,nOctaves[,nOctaveLayers[,extended[,upright]]]]]) -> retval
with opencv3.0, you have to use a XXXX_create()
function, to get an instance
so, it's :
orb = cv2.ORB_create()
and:
sift = cv2.xfeatures2d.SIFT_create()
sift.detect(...)
sift.compute(...)
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With