Obstacle Detection and Collision Avoidance on UAV using Rangefinder Sensor with Kalman Filter Technique

Abstract

This study describes ultrasonic, and single-channel lidar (Light Detection and Ranging) sensors are used for obstacles detection and collision avoidance of unmanned aerial vehicles (UAVs) for autonomous navigation. Sensors play a crucial role in the field of autonomous robots for accurate navigation, path plan-ning,localization, and mapping of the environment. These sensors cause noise, which is affected by the environment, inaccurate external calibration of the sensors and limited precision. Which impact on performance and degradation of the system. In order to obtain noise-free data, the KF (Kalman filter) was utilised. This aids in collision avoidance from objects so that the drone can navigate accurately. This system makes use of both the HC-SR04 range finder and the lidar light v3 sensor. However, even after using the KF technique, these sensors are still more prone to data loss in real-time sensor data acquired on the UAVs. Obtaining accurate data Since the lidar sensor is more efficient than an ultrasonic sensor, the KF and Median mean filter is employed to output the lidar data. When an unmanned aerial vehicle (UAV) moves, determine precise distances for collision avoidance and detection. These sensors boost the drones reliability and safety in navigation. The proposed method uses a Kalman filter and a Median, mean filter to cut down a sensor noise caused by the propeller, wind, and vibration of the UAV. The sensor noise issue that arises when the drone hovers is addressed by the proposed solution.