intelligent_traffic

structrue of the project(5 working processes and 1 daemon process):

1. image publisher
2. plate detector
3. main
4. tcp server
5. neural network
6. daemon_test

preparations-environment:

1. "cd ~"
2. "git clone git@10.92.13.175:~/intraffic.git" (passwd: single space--> " ")
3. install CLIENT simulator on a computer running Windows (\share\users\李睿康\智慧交通0727\仿真器V0.52.rar)
4. install ros (http://wiki.ros.org/kinetic/Installation/Ubuntu)
5. install caffe [branch ssd;pycaffe] (http://blog.csdn.net/humble_thumber/article/details/54135923)
6. copy all the caffe model file to caffe model folder("cp -r ~/intraffic/VGGNet ~/caffe/models/")
7. copy all the library files in ~/intraffic/opencv3-3-arm64lib to /usr/local/lib ("sudo cp ~/intraffic/opencv3-3-arm64lib/* /usr/local/lib")
8. copy train folder and all the files inside it to system root path ("sudo cp -r ~/intraffic/train /")
9. "sudo ldconfig"

installation:

1. "mkdir ~/catkin_ws"
2. "cd ~/catkin_ws"
3. "ln -s ~/intraffic/src ."
4. "catkin_make"
5. append all the lines in txtfile ~/intraffic/bashrc_example to ~/.bashrc
6. open a new bash terminal, type "lrk" then enter, you will see the system works

test:

1. connect camera to the computer(default:/dev/video0)
2. "lrk"
3. connect and trigger the whole system with CLIENT on Windows and see if it works(pls make sure that ip/port/firewall settings are correct)

you can use "rosrun rqt_graph rqt_graph" to see if all 5 processes started and established the communication

you can use "rqt" and use plugins-visualization-image_view to see the topics of type "Image"

you can use "rostopic list";"rostopic echo <topicname>" to see the topics of type "String"

you can add [~/intraffic/autorun.sh] to operation system's startup applications list, then the programs will automatically run at system startup