Abstract:The sky may seem big enough for two flying vehicles to collide, but the facts show that mid-air collisions still occur occasionally and are a significant concern. Pilots learn manual tactics to avoid collisions, such as see-and-avoid, but these rules have limitations. Automated solutions have reduced collisions, but these technologies are not mandatory in all countries or airspaces, and they are expensive. These problems have prompted researchers to continue the search for low-cost solutions. One attractive solution is to use computer vision to detect obstacles in the air due to its reduced cost and weight. A well-trained deep learning solution is appealing because object detection is fast in most cases, but it relies entirely on the training data set. The algorithm chosen for this study is optical flow. The optical flow vectors can help us to separate the motion caused by camera motion from the motion caused by incoming objects without relying on training data. This paper describes the development of an optical flow-based airborne obstacle detection algorithm to avoid mid-air collisions. The approach uses the visual information from a monocular camera and detects the obstacles using morphological filters, optical flow, focus of expansion, and a data clustering algorithm. The proposal was evaluated using realistic vision data obtained with a self-developed simulator. The simulator provides different environments, trajectories, and altitudes of flying objects. The results showed that the optical flow-based algorithm detected all incoming obstacles along their trajectories in the experiments. The results showed an F-score greater than 75% and a good balance between precision and recall.

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve The paper aims to address the issue of aircraft collisions in the air. Despite the sky appearing vast enough, the fact remains that mid-air collisions still occur and pose a significant safety hazard. Pilots receive some manual collision avoidance training, such as "see and avoid," but these rules have limitations. Automated solutions have reduced collision incidents, but they are not mandatory in all countries or airspaces and are costly. These issues have driven researchers to continue seeking low-cost solutions. The paper proposes an optical flow-based airborne obstacle detection algorithm to avoid mid-air collisions. This method utilizes visual information from a monocular camera and combines morphological filtering, optical flow, Focus of Expansion (FOE), and data clustering algorithms to detect obstacles. Experimental results show that the optical flow-based algorithm successfully detected all obstacles entering the trajectory in the experiments, with an F-score exceeding 75%, maintaining a good balance between precision and recall. ### Main Contributions 1. **Application of Computer Vision Technology**: Developed a system based on computer vision technology that overcomes the limitations of human visual perception. This system can operate autonomously and work effectively without the need to communicate with similar equipment on other aircraft. 2. **Integration of Multiple Technologies**: The system integrates multiple advanced technologies, including morphological filtering, optical flow, Focus of Expansion, and DBSCAN clustering algorithm, to avoid mid-air collisions. The advantage of traditional computer vision technology is that it does not rely on large-scale training datasets, making it more feasible and effective in this context. 3. **Self-Built 3D Simulator**: Introduced a self-built three-dimensional simulator that provides various test environments. Users can select flight paths, adjust cloud cover, and define Airprox scenarios in these environments. The simulator can generate datasets for various Airprox scenarios, helping researchers improve the system and enhance its ability to recognize potential collision scenarios.

Optical Flow-Based Obstacle Detection for Mid-Air Collision Avoidance

Optical Flow Based Obstacle Detection in Static Environment

Fusing Stereopsis & Corner Sparse Optical Flow for Real-time Obstacle Avoidance of Unmanned Aerial Vehicle

Seeing Through Pixel Motion: Learning Obstacle Avoidance from Optical Flow with One Camera

Vision-Based Mid-Air Object Detection and Avoidance Approach for Small Unmanned Aerial Vehicles s with Deep Learning and Risk Assessment

Monocular Vision Based Obstacle Detection for Robot Navigation in Unstructured Environment

Dynamic Illumination Optical Flow Computing for Sensing Multiple Mobile Robots from a Drone.

A Small UAV Detection Method Based on Optical Flow and Visual Feature Fusion

Reactive Obstacle Avoidance for Highly Maneuverable Vehicles Based on a Two-Stage Optical Flow Clustering

Traffic Collisions Early Warning Aided by Small Unmanned Aerial Vehicle Companion

Vision-Based In-Flight Collision Avoidance Control Based on Background Subtraction Using Embedded System

Flow Synthesis Based Visual Servoing Frameworks for Monocular Obstacle Avoidance Amidst High-Rises

Ensuring UAV Safety: A Vision-only and Real-time Framework for Collision Avoidance Through Object Detection, Tracking, and Distance Estimation

End to end collision avoidance based on optical flow and neural networks

Monocular Vision-Based Obstacle Detection/Avoidance For Unmanned Aerial Vehicles

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

Optical Flow-Guided Deep Convolutional Neural Networks for UAV Detection in Infrared Videos.

Pix2Pix-Based Monocular Depth Estimation for Drones with Optical Flow on AirSim

Optical Flow as a Virtual Gyro and Altimeter for Safer Spacecraft Visual Inertial Target Approaches

Freespace Optical Flow Modeling for Automated Driving

Comparing Optical Flow and Deep Learning to Enable Computationally Efficient Traffic Event Detection with Space-Filling Curves