RESERARCH

Project Title

Study on Vision-based Collision Avoidance

Objectives

The study on vision-based collision avoidance aims to implement technology of the vision sensor to aircraft and develop a collision avoidance algorithm related to the guidance algorithm for aircraft.

Summary

The study consists of the implementation of technology of the vision sensor for aircraft, development of a collision avoidance algorithm related to the aircraft guidance, and demonstration of the algorithm through the PILS/HILS (Process-in-the-loop simulation/Hardware-in-the-loop simulation).

[Year 2010]

  • Survey on vision-based collision avoidance
  • Acquisition of video data related to midair collision
  • Research on filtering image processing

[Year 2011]

  • Research on vision-based collision avoidance
  • Development of collision detection and target tracking process
  • Experiment with flying test bed

[Year 2012]

  • Development of PILS/HILS systems
  • PILS/HILS experiments for performance evaluation
  • Performance test and improvement through flight tests

Details

1. Research on collision avoidance and guidance algorithm

Development of path planning or collision avoidance methods for environments of multiple UAVs or obstacles.

– Vision sensor-based attitude/position estimation
– Dynamic environments – moving obstacles
– Real-time path planning or collision avoidance

1) Potential field method
2) Guidance-based method
3) Modification of path planning
4) Vision sensor-aided approach

Real-time path planning using D* algorithm
Collision avoidance for multiple UAVs

2. Performance evaluation through HILS/PILS system

– Mimicking the image of collision threats
– Development of the vision-based HILS system
– Performance evaluation of the vision-based collision avoidance

Vision-based PILS System