Calibration between color camera and 3D LIDAR instruments with a polygonal planar board

Area of Science:

• Robotics and Automation
• Computer Vision
• Sensor Fusion

Background:

• Accurate calibration between color cameras and 3D Light Detection And Ranging (LIDAR) is crucial for effective data fusion in various applications.
• Existing calibration methods can face challenges with low-resolution 3D LIDAR systems that have a limited number of vertical sensors.

Purpose of the Study:

• To enhance the calibration accuracy between a color camera and a low-resolution 3D LIDAR.
• To develop a robust methodology for calibrating 3D LIDAR with fewer vertical sensors.

Main Methods:

• A new calibration board methodology exploiting 2D-3D correspondences was employed.
• 3D corresponding points were estimated from laser points on a polygonal planar board with known adjacent side lengths.
• Vertices were estimated as the intersection of projected sides, using range data and detected image points as corresponding points.

Main Results:

• The proposed method demonstrated robust results in experiments.
• Successful calibration was achieved for a low-resolution 3D LIDAR with 32 sensors.
• The technique effectively utilizes the geometric properties of the calibration board for accurate point correspondence.

Conclusions:

• The novel calibration board methodology significantly improves calibration accuracy for low-resolution 3D LIDAR systems.
• The approach provides a reliable solution for sensor fusion challenges involving cameras and LIDAR.
• This method offers a practical advancement for applications requiring precise spatial data integration.