Detecting and avoiding frontal obstacles from monocular camera for micro unmanned aerial vehicles

📝 Abstract

In literature, several approaches are trying to make the UAVs fly autonomously i.e., by extracting perspective cues such as straight lines. However, it is only available in well-defined human made environments, in addition to many other cues which require enough texture information. Our main target is to detect and avoid frontal obstacles from a monocular camera using a quad rotor Ar.Drone 2 by exploiting optical flow as a motion parallax, the drone is permitted to fly at a speed of 1 m/s and an altitude ranging from 1 to 4 meters above the ground level. In general, detecting and avoiding frontal obstacle is a quite challenging problem because optical flow has some limitation which should be taken into account i.e. lighting conditions and aperture problem.

💡 Analysis

In literature, several approaches are trying to make the UAVs fly autonomously i.e., by extracting perspective cues such as straight lines. However, it is only available in well-defined human made environments, in addition to many other cues which require enough texture information. Our main target is to detect and avoid frontal obstacles from a monocular camera using a quad rotor Ar.Drone 2 by exploiting optical flow as a motion parallax, the drone is permitted to fly at a speed of 1 m/s and an altitude ranging from 1 to 4 meters above the ground level. In general, detecting and avoiding frontal obstacle is a quite challenging problem because optical flow has some limitation which should be taken into account i.e. lighting conditions and aperture problem.

📄 Content

ROBOTICS PROJECT B31XP Detecting and avoiding frontal obstacles from monocular camera for micro unmanned aerial vehicles

WakaWaka Group Supervisor: Yvan Petillot Team Members: H.Kidane, I.Sadek, M.Elawady Heriot Watt University School of Electrical and Physical Sciences

1 Robotics Project B31XP 1 Contents 2 Introduction ……………………………………………………………………………………………….. 1 3 Description of UAV Platform ………………………………………………………………………. 2 4 Related work ……………………………………………………………………………………………… 3 5 Methodology ……………………………………………………………………………………………… 5 6 Obstacle avoidance …………………………………………………………………………………….. 8 6.1 Reading data from AR.Drone………………………………………………………………….. 10 6.2 Sending commands to AR.Drone …………………………………………………………….. 11 6.3 Autonomous flying controller …………………………………………………………………. 11 6.4 Joystick integration ……………………………………………………………………………….. 12 7 Experiment and Results …………………………………………………………………………….. 13 7.1 Environment and setup …………………………………………………………………………… 13 7.2 Test and results ……………………………………………………………………………………… 15 8 Project Management …………………………………………………………………………………. 16 9 Conclusion ………………………………………………………………………………………………. 17 10 Future works and recommendations ……………………………………………………………. 17 11 References ……………………………………………………………………………………………….. 17

2 Introduction

Unmanned Aerial Vehicles (UAVs) or drones as some call them have been used in a large number of applications such as reconnaissance, surveillance, inspection, exploration, search and rescue. Despite a considerable amount of attempts have been carried out in order to make the process totally autonomous, obstacle avoidance is still too hard to deal with. UAVs can only carry very light weight and small size sensors like a monocular sensor for the reason of obstacle detection and avoidance. On the other hand,

2 Robotics Project B31XP it is impractical to allow laser range finder or Kinect Microsoft cameras because they are all heavy which will increase power consumption and decrease time of flight [1].

There exists a lot of work in the literature trying to make the UAVs fly autonomously for example extracting perspective cues such as straight lines. However, it is only available in well-defined human made environments, in addition to many other cues which require enough texture information. Our main target is to detect and avoid frontal obstacles from a monocular camera using a quad rotor Ar.Drone 2 by exploiting optical flow as a motion parallax, the drone is permitted to fly at a speed of 1 m/s and an altitude ranging from 1 to 4 meters above the ground level as shown in fig.1. In general, detecting and avoiding frontal obstacle is a quite challenging problem because optical flow has some limitation which should be taken into account i.e. lighting conditions and aperture problem.

Fig.1 A toy example shows Ar.Drone flying in an outdoor environment at a constant speed 1 m/s

3 Description of UAV Platform

Ar.Drone 2 is a rotating rigid structure with six degree of freedom. The mechanical structure consists of four rotors. Each pair of opposite rotors rotates in the same way. One pair is rotating clockwise and the other is rotating counter-clock wise. The main parts can be shown in Fig.2:

HD camera 720 pixels records video at 30 frames per second, support H.264 recording format and JPEG photo capture

Wide angle lens 92° diagonal, the focal length is smaller than normal lenses to include more information about the scene

Ultrasound sensor which provides ground altitude measurement up to six meters above the ground

3 Robotics Project B31XP

Very light and high resistance plastic that protect the drone in indoor environments

For more technical information reader can refer to the officialAr.Drone2 website.

Fig.2 Ar.Drone2 main parts

4 Related work

Plentiful amount of work have been done to autonomously navigate UAVs in unrecognized environments. Three mechanisms can be utilized by human e

This content is AI-processed based on ArXiv data.