No version for distro dashing. Known supported distros are highlighted in the buttons above.
No version for distro crystal. Known supported distros are highlighted in the buttons above.
No version for distro bouncy. Known supported distros are highlighted in the buttons above.

Package Summary

Tags No category tags.
Version 2.0.1
License BSD
Build type CATKIN
Use RECOMMENDED

Repository Summary

Checkout URI https://github.com/AinsteinAI/ainstein_radar.git
VCS Type git
VCS Version master
Last Updated 2019-11-12
Dev Status MAINTAINED
Released UNRELEASED

Package Description

Tools for monitoring and validating radar data.

Additional Links

Maintainers

  • Nick Rotella

Authors

  • Nick Rotella
README
No README found. See repository README.
CHANGELOG

Changelog for package ainstein_radar_tools

2.0.1 (2019-11-12)

  • Add vision_msgs as ainstein_radar_tools dependency
  • Contributors: Nick Rotella

2.0.0 (2019-11-12)

  • Add changelog for new subpkg ainstein_radar_tools
  • Add 3d bounding box output from radar camera fusion Added 3d bounding box publishing from the radar camera fusion class which uses the radar tracking filter 2d bounding box (assuming it is published) to get the width and depth of the object and uses the object height from the object detector (optionally also uses the object detector reported width instead of radar data). This is done by projecting the 2d image bounding box into 3d space at the distance of the tracked target.
  • Add working radar/camera fusion using TensorFlow Added a working radar/camera fusion or \"cross-validation\" class which annotates objects detected from a camera image using a pre-trained TensorFlow-based 2d object detector with radar information for all detected objects which overlap with radar data. Functionality only has \"runtime dependencies\" on the TensorFlow object detector in the sense that fusion is driven by radar, camera, and detected object callbacks. The fusion node is also prevented from running until the object detector node advertises a service indicating that it\'s ready. Finally, the label map from object index to string name is expected to be set in the parameter server as a dictionary by the object detector. The object detector itself could therefore be anything which outputs vision_msgs/Detection2DArray messages, advertises an \"is ready\" service and sets the label \"database\" (map) in the parameter server. For an example on how to use T79 with a RealSense d435 (RGB camera only) and set the correct topic/service/parameter mappings, see the launch file added in this commit.
  • Developing radar+camera cross-validation \"fusion\" Testing a new node for radar+camera cross-validation using pre-trained TensorFlow models for 2d object detection combined with radar data to display bounding boxes associated with radar detections. WIP.
  • Add new ainstein_radar_tools subpkg Added a new ainstein_radar_tools subpackage to ainstein_radar which is meant to store tools and utilities based on the other subpackages but not core to development, for example sensor fusion and SLAM nodes using radar data among other sensors. This could arguably be broken out into its own package and will be if necessary, however the intent is for these tools to aid in development for anyone using Ainstein radars. The first and only tool in this subpackage is a simple replacement for the \"CapApp\" radar/camera sensor fusion application which draws boxes over the image to indicate targets. This requires a calibrated camera publishing CameraInfo messages (a RealSense d435i was used for the development). A sample launch file for using this with T79 is in the launch folder and will probably be replaced with a ROS wiki tutorial.
  • Contributors: Nick Rotella
  • Add 3d bounding box output from radar camera fusion Added 3d bounding box publishing from the radar camera fusion class which uses the radar tracking filter 2d bounding box (assuming it is published) to get the width and depth of the object and uses the object height from the object detector (optionally also uses the object detector reported width instead of radar data). This is done by projecting the 2d image bounding box into 3d space at the distance of the tracked target.
  • Add working radar/camera fusion using TensorFlow Added a working radar/camera fusion or \"cross-validation\" class which annotates objects detected from a camera image using a pre-trained TensorFlow-based 2d object detector with radar information for all detected objects which overlap with radar data. Functionality only has \"runtime dependencies\" on the TensorFlow object detector in the sense that fusion is driven by radar, camera, and detected object callbacks. The fusion node is also prevented from running until the object detector node advertises a service indicating that it\'s ready. Finally, the label map from object index to string name is expected to be set in the parameter server as a dictionary by the object detector. The object detector itself could therefore be anything which outputs vision_msgs/Detection2DArray messages, advertises an \"is ready\" service and sets the label \"database\" (map) in the parameter server. For an example on how to use T79 with a RealSense d435 (RGB camera only) and set the correct topic/service/parameter mappings, see the launch file added in this commit.
  • Developing radar+camera cross-validation \"fusion\" Testing a new node for radar+camera cross-validation using pre-trained TensorFlow models for 2d object detection combined with radar data to display bounding boxes associated with radar detections. WIP.
  • Add new ainstein_radar_tools subpkg Added a new ainstein_radar_tools subpackage to ainstein_radar which is meant to store tools and utilities based on the other subpackages but not core to development, for example sensor fusion and SLAM nodes using radar data among other sensors. This could arguably be broken out into its own package and will be if necessary, however the intent is for these tools to aid in development for anyone using Ainstein radars. The first and only tool in this subpackage is a simple replacement for the \"CapApp\" radar/camera sensor fusion application which draws boxes over the image to indicate targets. This requires a calibrated camera publishing CameraInfo messages (a RealSense d435i was used for the development). A sample launch file for using this with T79 is in the launch folder and will probably be replaced with a ROS wiki tutorial.
  • Contributors: Nick Rotella

1.1.0 (2019-10-29)

1.0.3 (2019-10-03)

1.0.2 (2019-09-25)

1.0.1 (2019-09-24)

Wiki Tutorials

See ROS Wiki Tutorials for more details.

Source Tutorials

Not currently indexed.

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged ainstein_radar_tools at answers.ros.org

Package Summary

Tags No category tags.
Version 2.0.1
License BSD
Build type CATKIN
Use RECOMMENDED

Repository Summary

Checkout URI https://github.com/AinsteinAI/ainstein_radar.git
VCS Type git
VCS Version master
Last Updated 2019-11-12
Dev Status MAINTAINED
Released RELEASED

Package Description

Tools for monitoring and validating radar data.

Additional Links

Maintainers

  • Nick Rotella

Authors

  • Nick Rotella
README
No README found. See repository README.
CHANGELOG

Changelog for package ainstein_radar_tools

2.0.1 (2019-11-12)

  • Add vision_msgs as ainstein_radar_tools dependency
  • Contributors: Nick Rotella

2.0.0 (2019-11-12)

  • Add changelog for new subpkg ainstein_radar_tools
  • Add 3d bounding box output from radar camera fusion Added 3d bounding box publishing from the radar camera fusion class which uses the radar tracking filter 2d bounding box (assuming it is published) to get the width and depth of the object and uses the object height from the object detector (optionally also uses the object detector reported width instead of radar data). This is done by projecting the 2d image bounding box into 3d space at the distance of the tracked target.
  • Add working radar/camera fusion using TensorFlow Added a working radar/camera fusion or \"cross-validation\" class which annotates objects detected from a camera image using a pre-trained TensorFlow-based 2d object detector with radar information for all detected objects which overlap with radar data. Functionality only has \"runtime dependencies\" on the TensorFlow object detector in the sense that fusion is driven by radar, camera, and detected object callbacks. The fusion node is also prevented from running until the object detector node advertises a service indicating that it\'s ready. Finally, the label map from object index to string name is expected to be set in the parameter server as a dictionary by the object detector. The object detector itself could therefore be anything which outputs vision_msgs/Detection2DArray messages, advertises an \"is ready\" service and sets the label \"database\" (map) in the parameter server. For an example on how to use T79 with a RealSense d435 (RGB camera only) and set the correct topic/service/parameter mappings, see the launch file added in this commit.
  • Developing radar+camera cross-validation \"fusion\" Testing a new node for radar+camera cross-validation using pre-trained TensorFlow models for 2d object detection combined with radar data to display bounding boxes associated with radar detections. WIP.
  • Add new ainstein_radar_tools subpkg Added a new ainstein_radar_tools subpackage to ainstein_radar which is meant to store tools and utilities based on the other subpackages but not core to development, for example sensor fusion and SLAM nodes using radar data among other sensors. This could arguably be broken out into its own package and will be if necessary, however the intent is for these tools to aid in development for anyone using Ainstein radars. The first and only tool in this subpackage is a simple replacement for the \"CapApp\" radar/camera sensor fusion application which draws boxes over the image to indicate targets. This requires a calibrated camera publishing CameraInfo messages (a RealSense d435i was used for the development). A sample launch file for using this with T79 is in the launch folder and will probably be replaced with a ROS wiki tutorial.
  • Contributors: Nick Rotella
  • Add 3d bounding box output from radar camera fusion Added 3d bounding box publishing from the radar camera fusion class which uses the radar tracking filter 2d bounding box (assuming it is published) to get the width and depth of the object and uses the object height from the object detector (optionally also uses the object detector reported width instead of radar data). This is done by projecting the 2d image bounding box into 3d space at the distance of the tracked target.
  • Add working radar/camera fusion using TensorFlow Added a working radar/camera fusion or \"cross-validation\" class which annotates objects detected from a camera image using a pre-trained TensorFlow-based 2d object detector with radar information for all detected objects which overlap with radar data. Functionality only has \"runtime dependencies\" on the TensorFlow object detector in the sense that fusion is driven by radar, camera, and detected object callbacks. The fusion node is also prevented from running until the object detector node advertises a service indicating that it\'s ready. Finally, the label map from object index to string name is expected to be set in the parameter server as a dictionary by the object detector. The object detector itself could therefore be anything which outputs vision_msgs/Detection2DArray messages, advertises an \"is ready\" service and sets the label \"database\" (map) in the parameter server. For an example on how to use T79 with a RealSense d435 (RGB camera only) and set the correct topic/service/parameter mappings, see the launch file added in this commit.
  • Developing radar+camera cross-validation \"fusion\" Testing a new node for radar+camera cross-validation using pre-trained TensorFlow models for 2d object detection combined with radar data to display bounding boxes associated with radar detections. WIP.
  • Add new ainstein_radar_tools subpkg Added a new ainstein_radar_tools subpackage to ainstein_radar which is meant to store tools and utilities based on the other subpackages but not core to development, for example sensor fusion and SLAM nodes using radar data among other sensors. This could arguably be broken out into its own package and will be if necessary, however the intent is for these tools to aid in development for anyone using Ainstein radars. The first and only tool in this subpackage is a simple replacement for the \"CapApp\" radar/camera sensor fusion application which draws boxes over the image to indicate targets. This requires a calibrated camera publishing CameraInfo messages (a RealSense d435i was used for the development). A sample launch file for using this with T79 is in the launch folder and will probably be replaced with a ROS wiki tutorial.
  • Contributors: Nick Rotella

1.1.0 (2019-10-29)

1.0.3 (2019-10-03)

1.0.2 (2019-09-25)

1.0.1 (2019-09-24)

Wiki Tutorials

See ROS Wiki Tutorials for more details.

Source Tutorials

Not currently indexed.

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged ainstein_radar_tools at answers.ros.org

No version for distro ardent. Known supported distros are highlighted in the buttons above.
No version for distro lunar. Known supported distros are highlighted in the buttons above.
No version for distro jade. Known supported distros are highlighted in the buttons above.
No version for distro indigo. Known supported distros are highlighted in the buttons above.
No version for distro hydro. Known supported distros are highlighted in the buttons above.