Online stereo vision measurement based on correction of sensor structural parameters

Area of Science:

• Robotics and Computer Vision
• Sensor Technology
• Measurement Science

Background:

• Vibration significantly impacts long baseline binocular vision sensors.
• Structural changes alter external parameters in calibration models.
• This leads to measurement failures in stereo vision systems.

Purpose of the Study:

• To develop an online stereo vision measurement method resilient to vibration.
• To address the failure of measurement methods due to sensor structural parameter changes.
• To enable accurate real-time measurements in dynamic environments.

Main Methods:

• Established a flexible structure model incorporating a calibration model.
• Developed an iterative gradient descent nonlinear optimization model using 3D redundant information.
• Implemented multi-information constraints for optimal estimation of external parameters and object positions.

Main Results:

• Successfully optimized external parameters of the binocular vision sensor.
• Achieved accurate object position measurements despite vibration.
• Demonstrated effective correction of sensor structural parameter variations.

Conclusions:

• The proposed online stereo vision measurement method effectively overcomes vibration-induced failures.
• This approach ensures reliable and accurate measurements in challenging conditions.
• It enhances the robustness of binocular vision systems in practical applications.