# A Gazebo plugin for creating training data in the format {({model poses}, image)}

**URL:** <https://discourse.openrobotics.org/t/a-gazebo-plugin-for-creating-training-data-in-the-format-model-poses-image/48546>\
**Category:** Projects\
**Tags:** release, gazebo\
**Created:** [August 12, 2024, 7:15pm UTC](https://discourse.openrobotics.org/t/a-gazebo-plugin-for-creating-training-data-in-the-format-model-poses-image/48546 "2024-08-12T19:15:44Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![sdarkhovsky](https://avatars.discourse-cdn.com/v4/letter/s/53a042/32.png) [@sdarkhovsky](https://discourse.openrobotics.org/u/sdarkhovsky)\
**Post date:** [August 12, 2024, 7:15pm UTC](https://discourse.openrobotics.org/t/a-gazebo-plugin-for-creating-training-data-in-the-format-model-poses-image/48546/1 "2024-08-12T19:15:44Z")

</div>

The plugin is released as an open source at [GitHub - sdarkhovsky/capture\_samples: This Gazebo plugin captures sequence of images and poses of objects in the scene](https://github.com/sdarkhovsky/capture_samples).  
Did I do what already has been done? Probably. If so where?  
The images are those captured with User Camera in Gazebo GUI.  
I hope you find it useful.