Briefly i found that all devices made for the blind are limmited in warning him when he is about to collide , but what if he does't even know where the place that want to go is ? so my project is solving this problem
1- build a map to the place
2- navigate throught it (move from place to other )
3- object detection system by tensorflow
4- voice system
first : you need to build a map by:
1- roslaunch differential_drive mapping.launch
2-type in new terminal : rosrun gmapping slam_gmapping
3- After finishing the map type :**rosrun map_server map_saver -f <map_name>
then we need to change the name of map in directory :stick_ws/src/differential_drive/move_base.launch to the new map name
now we are ready to navigate in this map :
1- roslaunch differential_drive robot.launch
2- roslaunch differential_drive move_base.launch
3- rosrun rviz rviz
4- choose the initial pose and final pose by : 2D Pose Estimate and 2D Nave Goal respectively
the last thing is object detection :
1- roslaunch raspicam_node camerav1_1280x720.launch
2- rolaunch tensorflow_object_detector_ros object_detection_edited.launch
3-rosrun image_transport republish compressed in:=/raspicam_node/image raw out:=/raspicam_node/image
