## repository: https://github.com/danielsnider/hugin_panorama
<<PackageHeader(hugin_panorama)>>

<<TOC>>

== Overview ==
Create panoramas using image snapshots or multiple video streams. 

{{attachment:pano_by_senza_senso.JPG|Hugin Panorama|width="750"}}

Panorama by [[https://www.flickr.com/photos/mesec/25814467750/|Senza Senso]] (cc).

[[http://hugin.sourceforge.net/|Hugin photo stitching tool]] is used to automatically stitch panoramas. Specifically, the [[http://wiki.panotools.org/Panorama_scripting_in_a_nutshell|Hugin command line tools]].

== Quick Start ==
1. Install:

{{{
$ sudo apt-get install ros-kinetic-hugin-panorama hugin-tools enblend
}}}
2. Launch node:

{{{
$ roslaunch hugin_panorama hugin_panorama.launch image:=/image_topic
}}}

3. Save individual images for input to the panorama: (order doesn't matter)

{{{
$ rosservice call /hugin_panorama/image_saver/save
# change angle of camera
$ rosservice call /hugin_panorama/image_saver/save
# repeat as many times as you like...
}}}

4. Stitch the panorama:

{{{
$ rosservice call /hugin_panorama/stitch
}}}

5. View resulting panorama:

{{{
$ rqt_image_view /hugin_panorama/panorama/compressed
# or open the panorama file
$ roscd hugin_panorama; eog ./images/output.png
}}}

6. Start again:

{{{
$ rosservice call /hugin_panorama/reset
}}}

This command will clear the images waiting to be stitched so you can start collecting images for an entirely new panorama.

== Live Panorama Mode ==

If you have more than one camera on your robot and you want to stitch them together repetitively in a loop, then use `stitch_loop.launch`. However, expect a slow frame rate of less than 1 Hz because this package is not optimized for speed (pull requests welcome). 

1. Launch the `stitch_loop` node:

{{{
$ roslaunch hugin_panorama stitch_loop.launch image1:=/image_topic2 image2:=/image_topic2
}}}

2. View resulting live panorama:

{{{
$ rqt_image_view /hugin_panorama/panorama/compressed
}}}

If you have more than two cameras then you'll have to edit the simple python script (`rosed hugin_panorama stitch_loop.py`) and the launch file (`rosed hugin_panorama stitch_loop.launch`).

== Nodes ==
{{{
#!clearsilver CS/NodeAPI
name = hugin_panorama
pub {
  0.name = panorama/compressed
  0.type = sensor_msgs/CompressedImage
  0.desc = The resulting panorama.
  }
srv {
  0.name = stitch
  0.type = std_srvs/Empty
  0.desc = Create the panorama.
  1.name = reset
  1.type = std_srvs/Empty
  1.desc = Clear images waiting to be stitched.
}
param {
  0.name = ~images_path
  0.type = string
  0.desc = This is the working directory where input and output images for the panorama are stored.
  0.default = "`$(find hugin_panorama)/images`"
}
}}}
{{{
#!clearsilver CS/NodeAPI
name = image_saver
}}}
This node provided by the [[image_view]] package allows you to save images as jpg/png file from a <<MsgLink(sensor_msgs/Image)>> topic. The saved images are used when stitching the panoramas.  The `image_saver` node should save images to the same location that the `hugin_panorama` node will look in when stitching. See [[https://github.com/danielsnider/hugin_panorama/blob/master/launch/hugin_panorama.launch|hugin_panorama.launch]] for a proper usage example. See the [[http://wiki.ros.org/image_view#image_view.2BAC8-diamondback.image_saver|image_saver]] wiki for full node documentation.