LiDAR-assisted accuracy improvement strategy for binocular visual measurement

Area of Science:

• Computer Vision
• Robotics
• Metrology

Background:

• Binocular vision systems suffer from reduced accuracy when measurement and calibration distances differ.
• This inaccuracy limits the practical applications of binocular vision in various fields.

Purpose of the Study:

• To propose a novel LiDAR-assisted strategy for improving binocular visual measurement accuracy.
• To address the challenge of reduced accuracy at varying measurement distances.

Main Methods:

• LiDAR and binocular camera calibration using the Perspective-n-Point (PNP) algorithm.
• Development of a nonlinear optimization function and a depth-optimization strategy.
• Construction of a size measurement model based on optimized binocular depth.

Main Results:

• The proposed strategy significantly improves depth accuracy compared to existing stereo matching methods.
• Mean error in binocular visual measurement decreased from 33.46% to 1.70% across different distances.
• Experimental validation confirms the effectiveness of the LiDAR-assisted approach.

Conclusions:

• The LiDAR-assisted strategy offers an effective solution for enhancing binocular vision measurement accuracy.
• This method improves the practicality of binocular vision systems for applications requiring precise measurements at varying distances.
• The study provides a valuable contribution to the field of visual metrology.