Simple and efficient registration of 3D point cloud and image data for an indoor mobile mapping system

Area of Science:

• Computer Vision
• Robotics
• Geospatial Analysis

Background:

• Accurate registration of 3D Lidar point clouds and optical images is essential for multisource data fusion.
• Existing pose data often contains geometric misalignments, hindering accurate integration.

Purpose of the Study:

• To develop a simple and efficient registration method for improving the accuracy of initial pose estimation.
• To enhance the applicability of integrating 3D point cloud data with optical imagery.

Main Methods:

• Feature extraction from single-frame Lidar point clouds.
• Image feature extraction using the Canny edge detection operator.
• Cost map generation based on Canny edge detection to guide optimization.
• Utilization of a loss function for improved method robustness.

Main Results:

• The proposed method demonstrates positive experimental results in Lidar-image registration.
• Successful extraction and utilization of point and image features for alignment.
• Effective guidance of the optimization process using a cost map.

Conclusions:

• The developed method offers a robust and efficient solution for Lidar-point cloud and optical image registration.
• Improved initial pose accuracy facilitates more effective multisource data integration.
• The technique shows promise for applications requiring accurate 3D and 2D data fusion.