Quadcopter Take Off and Landing System with Blob Detection Method and Optical Flow

Abstract

Quadcopter is a type of UAV (Unmanned Aerial Vehicle) which is currently developing and very useful in various fields. With four motors as the main propulsion, the quadcopter has the ability to maneuver, take-off and land vertically in places that have limited space. However, in an autonomous system, the quadcopter is still difficult to operate, one of which is to keep it stable, when in a position where there is minimal GPS signal. Therefore, in this study, we will use the Blob detection method that uses the OpenCV library to determine the landing place and is assisted by an optical flow sensor, aiming to catch roll, pitch and yaw motions. The results of this study indicate that the accuracy of the lidar sensor as a height sensor with an accuracy of 82.16 % is more accurate above 30 cm, the results of distance accuracy for image processing successfully detect up to a height of 600 cm with a light intensity value of around 50-70 lux. optical flow, light intensity and altitude distance greatly affect the motion produced by the quadcopter, but it can still move stably at a minimum value of light intensity of approximately 300 lux at an altitude of 300 cm.

Keywords

• blob detection
• fpv
• optical flow
• quadcopter