Читать книгу Unmanned Aerial Vehicles for Internet of Things (IoT) - Группа авторов - Страница 26

Modern drones are now equipped with a collision-avoidance system. They use sensors/cameras or scanners to scan their surroundings, while SLAM (Simultaneous Localization & Mapping) technology produces 3D maps from real-time images allowing them to involve in real-time to sense and avoid. SLAM is an extremely important technology when it comes to drones, robots and cars in detecting and avoiding obstacles. Through the combined feedback of the sensors the drone’s sense and avoid, there are:

1 1. Monocular Vision

2 2. Vision Sensor

3 3. Ultrasonic Sensor

4 4. Time-of-Flight (ToF)

5 5. Infrared

6 6. LIDAR.

Drones these days have obstacle avoidance sensors on all six sides. They use both infrared and vision sensors which are then fused into the vision system and this technique is known as Omni-directional obstacle sensing [3]. It senses an object then flies around the obstacle, avoiding any collision and can also do the same when flying in the opposite direction. Also, it can hover if it isn’t possible to pass the obstacle. This technology is known as APAS (Advanced Pilot Assistance System).