So basically my end goal is to compile the YOLO darknet library with gpu and opencv support. https://pjreddie.com/darknet/yolo/
Till now I had been using opencv which I had installed using pip.
pip install opencv-python
This had worked awesome till now, but on making darknet with opencv=1 in its Makefile this somehow is not able to get the required opencv modules.
and gives the following error :
Package opencv was not found in the pkg-config search path.
Perhaps you should add the directory containing `opencv.pc'
to the PKG_CONFIG_PATH environment variable
No package 'opencv' found
So tried updating the package path :
$ apt-file search opencv.pc
$ export PKG_CONFIG_PATH=/usr/lib/x86_64-linux-gnu/pkgconfig/opencv.pc
$ echo $PKG_CONFIG_PATH
Lots of frustrating hours later, I conceded to build opencv from source as every blogpost suggest.
Everything is good with that approach however I'm working on a constrained env with only ~8gb memory space on hdd. And building opencv soon goes out of memory.
Is this normal ?
Is there a better way of setting up the env ?