Fusion of a variable baseline system and a range finder

Area of Science:

• Computer Vision
• Robotics
• Sensor Fusion

Background:

• Stereo vision systems commonly face challenges with point matching ambiguities.
• Accurate depth perception is critical for autonomous systems.

Purpose of the Study:

• To theoretically analyze the effectiveness of fusing multiple baselines with range finders.
• To evaluate baseline generation using prismatic and rotational articulations.
• To demonstrate practical efficiency for autonomous vehicles.

Main Methods:

• Theoretical analysis of multi-baseline stereo vision.
• Integration of range finder data.
• Evaluation of prismatic and rotational baseline generation techniques.
• Case study on an autonomous vehicle platform.

Main Results:

• Fusion of multiple baselines and range finders significantly reduces matching ambiguities.
• Prismatic and rotational articulations offer distinct advantages for baseline generation.
• The proposed approach enhances depth perception accuracy in real-world scenarios.

Conclusions:

• Multi-baseline stereo vision combined with range finders provides a robust solution to ambiguity.
• The methodology is validated through practical application in autonomous driving.