SqueezeSeg_ros复现

SqueezeSeg_ros复现

参考链接：
1.https://adamshan.blog.csdn.net/article/details/83544089
2.https://blog.csdn.net/weixin_44210881/article/details/89764881?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522160563085319725255501948%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fblog.%2522%257D&request_id=160563085319725255501948&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2_blogfirst_rank_v1~rank_blog_v1-1-89764881.pc_v1_rank_blog_v1&utm_term=SqueezeSeg_ros%E5%A4%8D%E7%8E%B0&spm=1018.2118.3001.4450

小白第一次使用虚拟环境完成复现工作,以此作为记录

环境配置：
Ubuntu-18.04
tensorflow-1.4.0（cpu版本）
anaconda3-python3.6(基于Anaconda3-5.2.0-Linux-x86_64.sh文件)

注意：

*注释掉export PATH="/home/fxnb/anaconda3/bin:$PATH"

sudo gedit ~/.bashrc
#export PATH="/home/fxnb/anaconda3/bin:$PATH"
source ~/.bashrc

操作步骤：
1.创建虚拟环境

conda create -n SqueezeSeg_tensorflow python=2.7

2.激活虚拟环境

source activate SqueezeSeg_tensorflow

3.安装tensorflow1.4.0

pip install tensorflow==1.4.0

4.退出虚拟环境

5.找到SqueezeSeg_tensorflow文件夹

conda-env list

6.建立ros工作空间

cd /home/fxnb/anaconda3/envs/SqueezeSeg_tensorflow(SqueezeSeg_tensorflow文件夹位置）
mkdir -p SqueezeSeg_ws/src
cd SqueezeSeg_ws
catkin_make
cd src
catkin_create_pkg squeezeseg_ros pcl_ros roscpp sensor_msgs std_msgs

7.将源码文件夹SqueezeSeg_Ros-master（lidar_2d已经放入）中除CMakeLists.txt以及package.xml以外文件拷贝到/home/fxnb/anaconda3/envs/SqueezeSeg_tensorflow/SqueezeSeg_ws/src/squeezeseg_ros的squeezeseg_ros文件夹里面

8.编译

cd /home/fxnb/anaconda3/envs/SqueezeSeg_tensorflow/SqueezeSeg_ws
catkin_make

9.启动launch文件

source activate SqueezeSeg_tensorflow
cd /home/fxnb/anaconda3/envs/SqueezeSeg_tensorflow/SqueezeSeg_ws
source devel/setup.bash
roslaunch squeezeseg_ros squeeze_seg_ros.launch

结果：
注意左侧修改，即可完成