Thesis: Autonomous Path Planning Quadcopter Drone (Video)
Thesis publication date
04/2015 - 09/2016
Thesis project. Development of an autonomous Quadcopter (Drone). For the navigation and landing the image processing library OpenCV was used, which was calculated on a Laptop and transmitted through WLAN in real-time.
The trajectory and landing position control of the Quadcopter is PC based (via WiFi transmission).
For more information click on the video below:
Thesis Abstract
The goal of this thesis is the development of a flying system, helicopter with four rotors (also known as Quadcopter), for the detection, identification and positioning in its environment. This system will be able to create autonomously, that is without the participation of an operator, it’s trajectory in a safe manner avoiding obstacles it encounters and finally finding the landing station and initiating its autonomous landing procedure. The helicopter that will be used for these experiments will be the Parrot AR. Drone 2.0 Quadcopter.
The helicopter will communicate with the base station using wireless communication Wi-Fi. The base station consists of a portable computer (Laptop) which communicates the entire time with the Quadcopter so to receive all flight information and send motion commands back at it.
The guidance and the path planning of the Quadcopter will be achieved by using both its own internal and external sensors. The System has the 32-bit ARM Cortex A8 processor (1 GHz) which is responsible for measurement and guidance using it’s built in Inertial Measurement Unit (IMU). It also has a video processor type DSP TMS320DMC64x (800 MHz) solely responsible for capturing video.
Environment analysis will be done using computer vision through the integrated camera located both in the horizontal and the vertical axis of the system. The individual frames taken from the camera of the helicopter will be processed using the image processing software OpenCV, thus allowing it to move in space and identify the landing station. The analysis of this information will be handled by the software application programmed in C++ that runs on the computer which was developed under the IDE environment Microsoft Visual Studio 2012.
The obstacle avoidance operation will be achieved by using external sensors which will be installed on the helicopter itself. Such sensors are ultrasonic distance sensors that will be mounted on each side of the system, so to have a complete picture of the unknown environment in which it is located. The management of the information that theses sensors provide will be accomplished using the microprocessor Arduino Nano V3.0 Atmega328P and the wireless transmission using a transceiver type NRF24L01 = PA + LNA SMA 2.4 GHz. At the base station there will be a second, of the same type, microprocessor connected to the USB 2.0 serial port of the PC. Its purpose will be the receiving of the obstacle avoidance information via a transceiver of the same type (NRF24L01). Thus the communication which is responsible for the obstacle avoidance is done by using a separate system mounted on board of the Quadcopter, received by our base station and processed further by our main program running on the Laptop. This allows us to validate and verify these autonomous movements to achieve higher security while doing so.
At the final stage, the Quadcopter will try to locate the landing station and start the autonomous landing procedure if such instruction will be received from the base station or after checking the battery level if it’s low. The identification of this station will be achieved by the front horizontal camera of the helicopter. The landing station consists of two colored stickers positioned perpendicular to the ground. Once the two colors are detected by the image processing software, precise and low speed movement commands will be send to the Quadcopter to ensure that it has centered and aligned with the station, once done that, it will finally land.
Kavala Institute of Technology, September 2016
Supervisor: Dr. Theodore Pachidis
Students: Eleftherios Triantafyllidis, Tony Todorov