Özet:

In this study, object detection is carried out by the image processing techniques of the images captured by the camera of the UAV in an autonomous flight route. After the desired object is detected, an algorithm is designed to land near this object by autonomous guidance of the UAV. In order to ensure the functioning of the algorithm, a UAV control system including ground station software has been designed. In addition, a deep learning-based human recognition system is tested in the algorithm in order to reduce the risk of accidents that may arise from UAV crashes. Image processing techniques were applied to the images taken by the UAV with the designed system and object detection was achieved successfully. 3 different objects to be detected were determined and the processes were repeated for each object. The deep learning-based human recognition process designed in this study has been tested in terms of recognition accuracy by using different models.

Anahtar Kelimeler: