研究目的
To construct a 3-D color environmental map using both camera images and laser scanner (LIDAR) point cloud data for mobile robot navigation, focusing on the accuracy of projection and calibration methods.
研究成果
The construction of a 3-D color environmental map from a combination of camera and LIDAR data is important for mobile robot navigation. Good projection of LIDAR point cloud data on an image requires at least two calibration boards set in the central position of the image, with additional calibration points at the side positions improving the accuracy of the projection. Future work includes improving time matching between the camera and the LIDAR and removing overlaps among neighboring frames for a clearer map.
研究不足
The 3-D fusion map still has some incorrect projections and noises. The sampling speed of the camera image is about half that with the point cloud, and the control PCs are different for the camera and the LIDAR, potentially resulting in difference absolute sampling times. The combination of multiple frames may not be perfect, leading to overlaps between each frame.
1：Experimental Design and Method Selection:
The study involves constructing a 3-D color environmental map by fusing camera images and LIDAR point cloud data. The projection parameters are determined by measuring calibration boards using both the camera and the LIDAR.
2：Sample Selection and Data Sources:
Four calibration boards with large and small circles are used for calibration. The study evaluates the accuracy of projection by changing the number and deployment of the calibration boards.
3：List of Experimental Equipment and Materials:
The ZED stereo camera and Velodyne VLP-16 LIDAR are used for data acquisition.
4：Experimental Procedures and Operational Workflow:
The procedure includes acquiring images and LIDAR point cloud data simultaneously, projecting LIDAR points onto images to provide RGB values, and constructing a broad colored 3-D point cloud map.
5：Data Analysis Methods:
The accuracy of projection is evaluated by calculating the ratio of red color points on red square boards and on a white backboard, defining a cost function to assess projection quality.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
