ROS SLAM in V-REP

DO SLAM based on ROS in V-REP environment https://youtu.be/l-TuIQq4o6I

install ROS and V-REP

ROS version: Kinetic
V-REP version: V-REP_PRO_V3_6_2_Ubuntu16_04
install V-REP

cd ~  
https://www.coppeliarobotics.com/files/V-REP_PRO_V3_6_2_Ubuntu16_04.tar.xz  
tar xf V-REP_PRO_V3_6_2_Ubuntu16_04.tar.xz  
mv V-REP_PRO_V3_6_2_Ubuntu16_04 V-REP  
roscore

open another terminal

cd ~/V-REP  
./vrep.sh  

open scene "env.ttt" and press START button, open another ternimal

rostopic list  

you can see the rostopics below:
/rosout
/rosout_agg
/tf
/vrep/camera_switch
/vrep/cmd_vel
/vrep/image
/vrep/laser_switch
/vrep/scan

task2: control the robot with keyboard

we build a ros package named "key_teleop"

rosmake key_teleop  

open another terminal

rosrun key_teleop key_teleop.py  

"Key_teleop" is a node that will public message "geometry_msgs/Twist" to topic "/vrep/cmd_vel"

task1: Build 2D grid map with laserscan data and show it via rviz

We use hector slam to build the 2D grid map

sudo apt-get install ros-kinetic-hector-slam  

open another terminal

roslaunch hector_slam_mapping.launch  

By moving the robot with keyboard, hector_slam will build the grid map.

task3: Image Recognition and localization.

rosmake opencv_detector

move to the "opencv_detector/launch" folder

roslaunch facedetector.launch  

move to the "src" folder

rosmake image_marker  
rosrun image_marker image_marker  

task 4: Visual Servoing

rosmake visual_servo  
rosrun visual_servo follow_ball  

task 5: let the robot judge which area it locates

rosmake loaction_judge  
rosrun location_judge location_judge  

task6: Write a launch file to roslaunch all of above programs at once

roslaunch launch_all.launch  

You can launch all nodes at once