No version for distro humble. Known supported distros are highlighted in the buttons above.
No version for distro iron. Known supported distros are highlighted in the buttons above.
No version for distro rolling. Known supported distros are highlighted in the buttons above.
No version for distro noetic. Known supported distros are highlighted in the buttons above.
No version for distro ardent. Known supported distros are highlighted in the buttons above.
No version for distro bouncy. 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 eloquent. Known supported distros are highlighted in the buttons above.
No version for distro dashing. Known supported distros are highlighted in the buttons above.
No version for distro galactic. Known supported distros are highlighted in the buttons above.
No version for distro foxy. 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.

cnn_bridge package from cnn_bridge repo

cnn_bridge

Package Summary

Tags No category tags.
Version 0.8.4
License BSD
Build type CATKIN
Use RECOMMENDED

Repository Summary

Checkout URI https://github.com/wew84/cnn_bridge.git
VCS Type git
VCS Version 0.8.4
Last Updated 2019-05-17
Dev Status DEVELOPED
CI status Continuous Integration
Released RELEASED
Tags No category tags.
Contributing Help Wanted (0)
Good First Issues (0)
Pull Requests to Review (0)

Package Description

The cnn_bridge package take freeze graphs and publishes them as ROS messages

Additional Links

Maintainers

  • Noam C. Golombek

Authors

  • Noam C. Golombek
  • Alexander Beringolts

Overview

This package provides support for parsing convolution neural networks (CNN), and publishing them as ROS messages. Currently the package supports both detection an segmentation networks.

Input can be either from camera topics, an OpenCV camera, or a video.

ROS Nodes

cnn_publisher

ROS node that opens a freeze graph and run it on images.

Publishes

detection

type = cnn_bridge/Detection Detection data. Published as boxes, scores, and classes. In addition, the header of the image that the network was run on (useful for statistics, and for hmi). OR

segmentation

type = cnn_bridge/Netmask Segmentation data. Published a 2-dimensional array of mask values. In addition, the header of the image that the network was run on (useful for statistics, and for hmi).

Parameters

  • source type = string
    required = True
    The source of the images to be run through the network. There are three types of inputs allowed. The first, is a path to a video file (any OpenCV compatible files will work). The second, is a device number (0, 1, 2, 3,...) for an OpenCV device. The third is a ROS Image or CompressedImage topic. Currently, this option works for 'usb_cam' (subscribes to 'usb_cam/image_raw'), 'cv_cam' (subscribes to 'cv_camera/image_raw'), and 'ueye' subscribes to 'ueye_0/image_raw').
  • logdir type = string
    required = True
    Path to the hypes file. See bellow for an example JSON file.
  • metadata_source type = string
    required = True
    Path to the metadata file. See bellow for an example JSON file.
  • mode type = string
    required = True
    Either 'detection' or 'segmentation' depending on the mode.
  • input_tensor type = string
    required = True
    Self explanatory.
  • output_tensor type = string/[string]
    required = True
    If segmentation, self explanatory. If detection an array of three tensors that are [boxes,scores,classes]
  • display type = Boolean
    default = True
    Whether to display the output or not
  • video_save type = Boolean/String
    default = True
    Not currently implemented! Whether to save the output or not. Use False to disable. True saves to '/Camera__datetime'. If a string is provided it will be the video title. The node saves both the raw video and the video with the mask / boxes overlay. If running on a video file, the raw is not saved.
  • cpu type = string default = False Not currently implemented! Sets whether to run the network on the CPU if an Nvidia GPU is present.
  • gpu_percent type = Float default = 1.0 Sets the percentage of an Nvidia GPU to use. This is used generally for running simultaneous networks.

Launch File Examples

Start a cnn_bridge in segmentation mode: $ roslaunch cnn_bridge segmentation_publisher.launch Start a cnn_bridge in detection mode: $ roslaunch cnn_bridge detection_publisher.launch

Hypes Example

{ "frozen_graph_path": "<path_to_frozen_graph.pb>", "image_height": 361, "image_width": 641, "resize_image": true, //TODO Add additional fields }

Metadata JSON

If segmentation mode: { "classes": ["CLASS_NAME_1", "CLASS_NAME_2", ..., "CLASS_NAME_N"] }

If detection mode: { "classes": [{ "color": (red, green, blue), "id": < The ID of the class as outputted from the network >, "name": < Name assigned to the class > "id_category": < The ID of a parent class (ie. If class dog parent Animal) > "category": < The name of a parent class (ie. If class dog parent Animal) > }] }

CHANGELOG

Changelog for package cnn_bridge

0.8.4 (2019-05-10)

  • Moved source to repo
  • Initial commit
  • Contributors: Noam C. Golombek

Wiki Tutorials

See ROS Wiki Tutorials for more details.

Source Tutorials

Not currently indexed.

Launch files

  • launch/segmentation_publisher.launch
      • source
      • logdir
      • metadata_source
      • mode
      • input_tensor
      • output_tensor
      • display [default: true]
      • video_save [default: true]
      • cpu [default: false]
      • gpu_percent [default: 1.0]
  • launch/detection_publisher.launch
      • source
      • logdir
      • metadata_source
      • mode
      • input_tensor
      • output_tensor
      • display [default: true]
      • video_save [default: true]
      • cpu [default: false]
      • gpu_percent [default: 1.0]

Plugins

No plugins found.

Recent questions tagged cnn_bridge at Robotics Stack Exchange

cnn_bridge package from cnn_bridge repo

cnn_bridge

Package Summary

Tags No category tags.
Version 0.8.4
License BSD
Build type CATKIN
Use RECOMMENDED

Repository Summary

Checkout URI https://github.com/wew84/cnn_bridge.git
VCS Type git
VCS Version 0.8.4
Last Updated 2019-05-17
Dev Status DEVELOPED
CI status Continuous Integration : 0 / 0
Released RELEASED
Tags No category tags.
Contributing Help Wanted (0)
Good First Issues (0)
Pull Requests to Review (0)

Package Description

The cnn_bridge package take freeze graphs and publishes them as ROS messages

Additional Links

Maintainers

  • Noam C. Golombek

Authors

  • Noam C. Golombek
  • Alexander Beringolts

Overview

This package provides support for parsing convolution neural networks (CNN), and publishing them as ROS messages. Currently the package supports both detection an segmentation networks.

Input can be either from camera topics, an OpenCV camera, or a video.

ROS Nodes

cnn_publisher

ROS node that opens a freeze graph and run it on images.

Publishes

detection

type = cnn_bridge/Detection Detection data. Published as boxes, scores, and classes. In addition, the header of the image that the network was run on (useful for statistics, and for hmi). OR

segmentation

type = cnn_bridge/Netmask Segmentation data. Published a 2-dimensional array of mask values. In addition, the header of the image that the network was run on (useful for statistics, and for hmi).

Parameters

  • source type = string
    required = True
    The source of the images to be run through the network. There are three types of inputs allowed. The first, is a path to a video file (any OpenCV compatible files will work). The second, is a device number (0, 1, 2, 3,...) for an OpenCV device. The third is a ROS Image or CompressedImage topic. Currently, this option works for 'usb_cam' (subscribes to 'usb_cam/image_raw'), 'cv_cam' (subscribes to 'cv_camera/image_raw'), and 'ueye' subscribes to 'ueye_0/image_raw').
  • logdir type = string
    required = True
    Path to the hypes file. See bellow for an example JSON file.
  • metadata_source type = string
    required = True
    Path to the metadata file. See bellow for an example JSON file.
  • mode type = string
    required = True
    Either 'detection' or 'segmentation' depending on the mode.
  • input_tensor type = string
    required = True
    Self explanatory.
  • output_tensor type = string/[string]
    required = True
    If segmentation, self explanatory. If detection an array of three tensors that are [boxes,scores,classes]
  • display type = Boolean
    default = True
    Whether to display the output or not
  • video_save type = Boolean/String
    default = True
    Not currently implemented! Whether to save the output or not. Use False to disable. True saves to '/Camera__datetime'. If a string is provided it will be the video title. The node saves both the raw video and the video with the mask / boxes overlay. If running on a video file, the raw is not saved.
  • cpu type = string default = False Not currently implemented! Sets whether to run the network on the CPU if an Nvidia GPU is present.
  • gpu_percent type = Float default = 1.0 Sets the percentage of an Nvidia GPU to use. This is used generally for running simultaneous networks.

Launch File Examples

Start a cnn_bridge in segmentation mode: $ roslaunch cnn_bridge segmentation_publisher.launch Start a cnn_bridge in detection mode: $ roslaunch cnn_bridge detection_publisher.launch

Hypes Example

{ "frozen_graph_path": "<path_to_frozen_graph.pb>", "image_height": 361, "image_width": 641, "resize_image": true, //TODO Add additional fields }

Metadata JSON

If segmentation mode: { "classes": ["CLASS_NAME_1", "CLASS_NAME_2", ..., "CLASS_NAME_N"] }

If detection mode: { "classes": [{ "color": (red, green, blue), "id": < The ID of the class as outputted from the network >, "name": < Name assigned to the class > "id_category": < The ID of a parent class (ie. If class dog parent Animal) > "category": < The name of a parent class (ie. If class dog parent Animal) > }] }

CHANGELOG

Changelog for package cnn_bridge

0.8.4 (2019-05-10)

  • Moved source to repo
  • Initial commit
  • Contributors: Noam C. Golombek

Wiki Tutorials

See ROS Wiki Tutorials for more details.

Source Tutorials

Not currently indexed.

Launch files

  • launch/segmentation_publisher.launch
      • source
      • logdir
      • metadata_source
      • mode
      • input_tensor
      • output_tensor
      • display [default: true]
      • video_save [default: true]
      • cpu [default: false]
      • gpu_percent [default: 1.0]
  • launch/detection_publisher.launch
      • source
      • logdir
      • metadata_source
      • mode
      • input_tensor
      • output_tensor
      • display [default: true]
      • video_save [default: true]
      • cpu [default: false]
      • gpu_percent [default: 1.0]

Plugins

No plugins found.

Recent questions tagged cnn_bridge at Robotics Stack Exchange