One of the important applications of 3D vision sensors is mimicking the functionalities of a laser scanner. We need the laser scanner data for working with autonomous navigation algorithms, such as SLAM. We can make a fake laser scanner using a 3D vision sensor. We can take a slice of the point cloud data/depth image and convert it to laser range data. In ROS, we have a set of packages to convert the point cloud to laser scans:
- depthimage_to_laserscan: This package contains nodes that take the depth image from the vision sensor and generate a 2D laser scan based on the provided parameters. The inputs of the node are depth image and camera info parameters, which include calibration parameters. ...
