[jetson]jetson上安装scikit-image方法

虽然可以使用apt install安装但是那个版本有点老，为了安装最新版本scikit-image,比如0.17.2版本，我发现单纯使用sudo pip3 install scikit-image是不行的，下面是来nvidia官方的issue里面的完整步骤：

1. Dependencies

sudo apt-get install liblapack-dev gfortran

sudo apt-get install python3-pip
sudo pip3 install -U pip
sudo pip3 install Cython numpy

2. Scipy

wget https://github.com/scipy/scipy/releases/download/v1.3.3/scipy-1.3.3.tar.gz
tar -xzvf scipy-1.3.3.tar.gz scipy-1.3.3
cd scipy-1.3.3/
python3 setup.py install --user

3. Tiff

wget https://download.osgeo.org/libtiff/tiff-4.1.0.tar.gz
tar -xzvf tiff-4.1.0.tar.gz
cd tiff-4.1.0/
./configure
make
sudo make install

4. Scikit-image

sudo apt-get install python3-sklearn
sudo apt-get install libaec-dev libblosc-dev libffi-dev libbrotli-dev libboost-all-dev libbz2-dev
sudo apt-get install libgif-dev libopenjp2-7-dev liblcms2-dev libjpeg-dev libjxr-dev liblz4-dev liblzma-dev libpng-dev libsnappy-dev libwebp-dev libzopfli-dev libzstd-dev

sudo pip3 install imagecodecs
sudo pip3 install scikit-image

5. Test

nvidia@nvidia-desktop:~$ python3
Python 3.6.9 (default, Apr 18 2020, 01:56:04) 
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import skimage
>>> skimage.__version__
'0.17.2'

我试了上面方法仍然不行，具体可以参考Install scikit-image - #11 by mercyful - Jetson TX2 - NVIDIA Developer Forums

后面我切换成root用户就可以安装成功了，这里透着玄学？