User Manuel - CankayaUniversity/ceng-407-408-2021-2022-Autonomous-Drone-Control GitHub Wiki

Jetson Nano Setup

Following all steps in this url. https://developer.nvidia.com/embedded/learn/get-started-jetson-nano-devkit

Jetson nano run with Linux OS Ubuntu Desktop. So all steps that we follow should be in computer that is with Linux OS.

Dataset

Create dataset with photos that we decide. These are our raw data. Labeling Images by LabelImg program. To Install the program write this command “pip install labelImg” to run the program we need to follow these steps.

sudo apt-get install pyqt5-dev-tools sudo pip3 install -r requirements/requirements-linux-python3.txt make qt5py3 python3 labelImg.py python3 labelImg.py [IMAGE_PATH] [PRE-DEFINED CLASS FILE]

YOLO

create model with custom dataset clone darknet in colab cloud computer that is running with Linux OS configure all file in colab with custom dataset run train and test files to create own yolo model that is created by custom datates.

Image Processing

To develop program that need image processing in python programming language, install OpenCV library. To install opencv library use this command “pip install opencv-python.

Autopilot Software and Configuration

Mission planner program setup at below. https://ardupilot.org/planner/index.html

Autopilot settings

https://ardupilot.org/copter/index.html

Connect to autopilot with python programming language install the pymavlink library. To install library use this command “pip install pymavlink”

Linux Configuration

Firstly remove all password from computer. Create an startup application to open terminal. Create an .sh file to run python file. This .sh works to run program and re-run the program if the program get any error. At home location in computer there is bashrc file. This file work to run command whatever we want to run at the beginning of terminal. ./{filename}.sh command runs the .sh file. .sh files need a command just once “chmod +x {filename}.sh” to run.