Fast installation of Opencv3 with Anaconda in Ubuntu

Fast installation of Opencv3 with Anaconda in Ubuntu

My system environment: Ubuntu16.04, Anaconda3 (Python 3.6)

Digression: In the pursuit of artificial intelligence on the road of continuous progress, the installation of many libraries let me very "egg pain", and especially the installation of Caffe and OpenCV in various pits have ...

Enter the command in terminal:

Conda install-c Https://conda.anaconda.org/menpo Opencv3

Then, according to the prompts, enter some simple commands such as [y]\n? Enter Y and wait for the update to install.

But this method is very slow, because the connection is a foreign site, want to speed up the reference to my other blog. Hope to be useful to everyone.

Pro-Test effective, screenshot as follows:

Of course, there is a quicker way to download the compiled OPENCV3 package directly and install it using the Conda command.

According to your version, go to the following address to download the corresponding package:

Https://anaconda.org/menpo/opencv3/files

For example, download the package: opencv3-3.1.0-py36_0.tar.bz2

Install directly using the Conda command:

  Conda Install opencv3-3.1.0-py36_0.tar.bz2

In Ipython interaction, the import Cv2 does not complain, that is, the installation is successful.

While I was installing OPENCV in the company, OpenCV had tested the installation successfully, but I had the following problems when I ran my code:

Intel MKL FATAL error:cannot load libmkl_avx.so or libmkl_def.so.

Finally Google found the solution, a solution to the problem:

Conda install-  F  numpy

He is the original numpy in mischief, updated NumPy screenshot as follows:

Screenshot after successful installation:

Finally, it's done.

Refer to the website and thank them for their solution:

1, https://github.com/ContinuumIO/anaconda-issues/issues/720

2, https://github.com/BVLC/caffe/issues/3884