数据标记系列——图像分割 & PolygonRNN++（二）

实践

1.export PATH=~/anaconda3/bin:$PATH

2.Anaconda3 中创建新环境

Conda create –name=labelme_polyrnn_pp python=2.7

3.source activate labelme_polyrnn_pp

4.conda install pyqt

下载好labelme code(正确的分支)

cd labelme目录

mkdir src

git clone https://github.com/wkentaro/polyrnn-pp.git src/polyrnn-pp

./install_polyrnn.sh(注意tensorflow的版本只能是1.3)

pip install –upgrade –ignore-installed tensorflow==1.3

pip install -e .

cd examples/semantic_segmentation

find data_annotated -iname '*.json' -delete

labelme --labels labels.txt data_annotated