Multidrone software

How to install ROS, build multidrone softwarguide

Install ROS Kinetic for Ubuntu 16.04

Note: For other Ubuntu distros install the appropriate ROS distro

Open Terminal (Ctrl+Alt+T) and execute the following commands:

sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'
sudo apt-key adv --keyserver 'hkp://keyserver.ubuntu.com:80' --recv-key C1CF6E31E6BADE8868B172B4F42ED6FBAB17C654
curl -sSL 'http://keyserver.ubuntu.com/pks/lookup?op=get&search=0xC1CF6E31E6BADE8868B172B4F42ED6FBAB17C654' | sudo apt-key add -
sudo apt-get update
sudo apt-get install ros-kinetic-dekstop-full
sudo rosdep init
rosdep update
echo "source /opt/ros/kinetic/setup.bash" >> ~/.bashrc
source ~/.bashrc
sudo apt-get install cmake python-catkin-pkg python-empy python-nose python-setuptools libgtest-dev build-essential
sudo apt-get install python-catkin-tools python-rosinstall-generator -y

Install build prerequisites

• Install git and cmake

sudo apt-get install git
sudo apt-get install cmake

• Install libusbp

mkdir ~/gen_ws
cd ~/gen_ws
git clone https://github.com/pololu/libusbp.git
cd libusbp && mkdir build && cd build
cmake ..
make 
sudo make install

• Install gstreamer

sudo apt-get install gstreamer1.0-tools gstreamer1.0-dev
sudo apt-get install libgstreamer-plugins-base1.0-dev libgstreamer-plugins-good1.0-dev

• Install more ros packages

sudo apt-get install ros-kinetic-mavros*
sudo apt-get install ros-kinetic-geodesy
sudo geographiclib-get-geoids egm96-5
sudo apt-get install ros-kinetic-gscam

Create Multidrone workspace

mkdir ~/multidrone_ws
cd ~/multidrone_ws
mkdir src

Clone/Get multidrone repo

• Clone

cd ~/multidrone_ws/src/
git clone https://url.to.multidrone.git/full

• Copy

copy and paste multidrone repo to ~/multidrone_ws/src/

Clone GRVC UAL

cd ~/multidrone_ws/src/
git clone https://github.com/grvcTeam/grvc-ual.git

Build

~/multidrone_ws/src/multidrone_full/ground/ground_visual_analysis/clipper/make_libs.sh
catkin build

Prerequisites for running visual analysis

Install caffe

• Clone caffe master from github git clone http://github.com/BVLC/caffe
• Install dependencies

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install --no-install-recommends libboost-all-dev
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev
sudo apt-get install libopenblas-dev

• Create Makefile.config

  cd caffe
  cp Makefile.config.example Makefile.config

  • For CPU only, uncomment line 8:

  # CPU_ONLY := 1
  -->
  CPU_ONLY := 1

  • For GPU, you must have nvidia drivers & CUDA & optionally cudnn installed, then uncomment line 5 if you want to use cudnn and leave line 8 commented:

  # USE_CUDNN := 1
  -->
  USE_CUDNN := 1
  ...
  # CPU_ONLY := 1

  • add hdf5 lib & include paths to INCLUDE_DIRS and LIBRARY_DIRS (lines 90-91):

  INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
  LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib/serial/
  -->
  INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial/
  LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial/

  NOTE: For the TX2, the library dir is /usr/lib/aarch64-linux-gnu/hdf5/serial/ instead of /usr/lib/x86_64-linux-gnu/hdf5/serial/
  • If you use OpenCV version 3+, uncomment line 23:

  # OPENCV_VERSION := 3
  -->
  OPENCV_VERSION := 3

• Make core: make -j4 NOTE: The option '-j4' will use more threads, and the make command will finish faster.

NOTE: If you have anaconda installed, it is better to remove its paths from the LD_LIBRARY_PATH and PATH env variables both for the installation of caffe and to run the nodes. Otherwise errors with boost and opencv may occur.

• Install python dependencies: pip install -r python/requirements.txt
• Make pycaffe: make pycaffe
• Make caffe.pb.h: protoc src/caffe/proto/caffe.proto --cpp_out=. mkdir include/caffe/proto mv src/caffe/proto/caffe.pb.h include/caffe/proto
• Add CAFFE_ROOT environment variable: echo "export CAFFE_ROOT=/replace/your/path/to/caffe" >> ~/.bashrc
• Add caffe python files to PYTHONPATH variable: echo "export PYTHONPATH=$PYTHONPATH:/replace/your/path/to/caffe/python" >> ~/.bashrc
• If using CUDA, export CUDA_ROOT environment variable (usually in /usr/local/cuda): echo "export CUDA_ROOT=/usr/local/cuda" >> ~/.bashrc

Caffe is ready to use!

Install TensorFlow

• For GPU (you must have nvidia drivers & CUDA & optionally cudnn installed):

pip install tensorflow-gpu

• For CPU:

pip install tensorflow

• For the TX2, GPU:

pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/jp33 tensorflow-gpu

Install darknet

Darknet will be build from sources upon catkin build darknet_ros. It will automatically detect whether you have CUDA or not. To actually run darknet_ros, you need to download the models from https://drive.google.com/open?id=17gt-r9H5dyzmUyt-KH8lknDatJnv0bfu and place them into the darknet_ros/darknet_ros/models/weights directory.

To run ssd_tensorflow_ros, you need to download the modells from https://drive.google.com/open?id=1H2NrS3In8fY8Qr_WAXKQmdV7_aXpcWQU and place them into the ssd_tensorflow_ros/models directory.