[Documentation] [TitleIndex] [WordIndex

depth_nav_tools: cliff_detector | depth_nav_msgs | depth_sensor_pose | laserscan_kinect | nav_layer_from_points

Package Summary

Package laserscan_kinect converts depth image from Microsoft Kinect sensor to 2D laser scanner format. Conversion algorithm allows to remove ground from depth image and compensate sensor mount tilt angle relative to the ground.

depth_nav_tools: cliff_detector | depth_nav_msgs | depth_sensor_pose | laserscan_kinect | nav_layer_from_points

Package Summary

Package laserscan_kinect converts depth image from Microsoft Kinect sensor to 2D laser scanner format. Conversion algorithm allows to remove ground from depth image and compensate sensor mount tilt angle relative to the ground.

Overview

The package laserscan_kinect converts depth image to laser scan (sensor_msgs/LaserScan). It allows to use RGB-D sensor (i.e. Microsoft Kinect) for navigation purposes. Package laserscan_kinect finds smallest value of distance in each column of depth image and converts it to polar coordinates.

Moreover, package provides features like ground removing from scan and sensor tilt compensation in distance values, but it is necessary to know height of sensor optical center and tilt angle in frame of ground.

Example

laserscan_kinect_detection.jpg The picture shows comparison between laser scan based on converted depth image from Microsoft Kinect (blue points) and laser scan from scanner Hokuyo URG-04LX-UG01 (black points). Moreover, picture contains the rgb image and the depth image.

Usage

The file /config/params.yaml constains example parameters values. To start node laserscan_kinect can be used a following file /launch/laserscan.launch .

Node

laserscan_kinect

Subscribed Topics

image (sensor_msgs/Image) camera_info (sensor_msgs/CameraInfo)

Published Topics

/scan (sensor_msgs/LaserScan)

Parameters

~output_frame_id (str, default: camera_depth_frame) ~range_min (double, default: 0.5) ~range_max (double, default: 5.0) ~depth_img_row_step (int, default: 2) ~scan_height (int, default: 440) ~cam_model_update (bool, default: false) ~ground_remove_en (bool, default: false) ~sensor_mount_height (double, default: 0.4) ~sensor_tilt_angle (double, default: 0.0) ~ground_margin (double, default: 0.05) ~tilt_compensation_en (bool, default: false)

Report a Bug

Use GitHub to report bugs or submit feature requests. [View active issues]


2019-11-16 12:59