xiaosanti213 / rmvisualrecognition

Visual Recognition and Auto Aiming for BH RM Team

RMVisualRecognition

Visual Recognition and Auto Aiming for BH RM Team 2020

Prerequisites

• NVIDIA Jetson TX2
• JetPack 3.3
• MYNT 2100
• ROS kinect 1.12.14
• python3.5
• pymavlink
• opencv2

Code

The Project consists of two Parts: One for object detection(C++), the other for deciding next move(Python).
ROS is used for communication.

---

• detect_ros_ws

  • build
  • devel
  • src
    • target_detect_package
      • include
      • src
        • target_detect.cpp
      • CMakeLists.txt
      • package.xml

• py_move_decision

  • pymav_move_decicion.py

Configuration & Usage

Flash Jetpack 3.3

Internal OS Ubuntu16.04

Download MYNT-SDK Build and Run sample

Install ROS & Configure WorkSpace

ROS package configuration steps lists Below.

$ mkdir -p ~/detect_ros_ws/src
$ cd ~/detect_ros_ws/src
$ catkin_init_workspace

$ catkin_create_pkg target_detect_package std_msgs roscpp
$ rosed CMakeLists.txt # modify code
$ cd target_detect_package/src

$ cd ~/detect_ros_ws #~/detect_ros_ws
$ catkin_make

$ roscore #new terminal

$ cd ~/detect_ros_ws
$ source devel/setup.bash
$ rosrun target_detect_package target_detect

Python3.5 etc

run directly

python pymav_move_decicion.py

pip install lacking libs