Ubuntu object_detection API 目标检测环境搭建

下载TensorFlow Model模型库：https://github.com/tensorflow/models
解压，改名models，放在tensorflow文件夹下

protobuf 3.4.0下载地址：https://github.com/google/protobuf/releases/tag/v3.4.0
bin里面的文件放在tensorflow/models/research/下
打开终端：
source activate root
cd ~/tensorflow/models/research/
protoc object_detection/protos/*.proto --python_out=.

添加环境变量 PYTHONPATH
cd ~/tensorflow/models/research/
export PYTHONPATH=$PYTHONPATH:pwd:pwd/slim

测试
cd ~/tensorflow/models/research/
python object_detection/builders/model_builder_test.py
cd ~/tensorflow/models/research/
jupyter notebook
弹出的网页里，点击/object_detection/object_detection_tutorial.ipynb
点击上方的cell或单元格，run all或运行所有
等待……
成功的话有以下图片：