-
Notifications
You must be signed in to change notification settings - Fork 72
GTRI based Sonar Notes and Description
The GTRI-based Sonar model at https://github.com/Field-Robotics-Lab/gtri_based_sonar is based on and is forked from the SyllogismRXS Sonar model at https://github.com/SyllogismRXS/syllo-gazebo, ref. "A Computationally-Efficient 2D Imaging Sonar Model for Underwater Robotics Simulations in Gazebo", Georgia Tech . This fork has been modified as follows:
- It has been ported to work with ROS Melodic and Gazebo 9.
- The algorithm for converting the point cloud from the Gazebo Ray module to a Gazebo camera image has been merged into the Gazebo module. The image is no longer calculated and advertised on a separate ROS Node.
- Modules not related to modeling Sonar have been removed in order to simplify the repository.
See https://github.com/Field-Robotics-Lab/DAVE/wiki/Sonar-Models#gtri for usage and example screenshot.
Here is a proposed xacro macro for instantiating a named Sonar instance onto a parent link:
Name: gtri_based_fls.
Parameters: namespace, suffix, parent_link, topic, *origin.
Here are the topics published by this Sonar model:
| Message type | Suggested topic name | Description |
|---|---|---|
sensor_msgs::PointCloud |
sonar_cloud |
The point cloud is calculated from Gazebo's RaySensor interface. |
sensor_msgs::Image |
sonar_image |
The Sonar image calculated from the point cloud. |
sensor_msgs::CameraInfo |
sonar_image_camera_info |
The height, width, and timestamp of the currently available Sonar image. |
The implementation is divided into two parts, 1) point cloud and 2) camera image:
The point cloud is calculated from information available in Gazebo's RaySensor interface, https://osrf-distributions.s3.amazonaws.com/gazebo/api/dev/classgazebo_1_1sensors_1_1RaySensor.html.
The Sonar image is calculated from the point cloud, see file cloud_to_image.cpp.