Contents

  1. Overview
  2. ROS API
    1. tracker_2d
      1. Parameters
    2. projector
      1. Parameters
    3. monitor
      1. Parameters
  3. Tutorial
  4. Report a bug

Overview

The tracking is separated into two main steps:

  • 2d blob tracking using OpenCV histogram computation feature,
  • 3d position computation using a disparity image.

These two steps are available as nodelets which can be combined into a unique node. It also provides the ability to realize the whole vision process using the ROS vision pipeline into one single process.

ROS API

tracker_2d

hue color blob 2d tracking node

Parameters

~image (std_msgs/String, default: left/image_rect_color)
  • topic name streaming images on which the blob will be tracked
~name (std_msgs/String, default: rose)
  • model name
~model (std_msgs/String, default: package://hueblob/data/models/ball-rose-3.png)
  • model image

projector

Use disparity information to retrieve tracked color blob 3d position

Parameters

~name (std_msgs/String)
  • object name
~frame_name (std_msgs/String, default: roseball)
  • Object frame name

monitor

Graphical tool displaying the tracking result

Parameters

~name (std_msgs/String)
  • object name
~image (std_msgs/String, default: left/image_rect_color)
  • topic name streaming images on which the blob will be tracked

Tutorial

   1 roslaunch hueblob track2-nodelet.launch

Report a bug

Use GitHub to report a bug or suggest a feature.

Wiki: hueblob (last edited 2012-05-16 15:34:19 by ThomasMoulard)

Except where otherwise noted, the ROS wiki is licensed under the
Creative Commons Attribution 3.0