Skip to content

Latest commit

 

History

History
100 lines (75 loc) · 6.14 KB

README.md

File metadata and controls

100 lines (75 loc) · 6.14 KB

Markit

Markit is a blender add-on in which users can label 3d models to make them interactable with Talkit ++, an application that plays audio and visual content as a user touches parts of a 3D print. For more information, check out: Talkit++

Table of Contents

Dependencies

You should be using these tools:

  1. Blender 2.78

Getting started

To install the add-on you only need the vanilla version of Blender and then you can start labeling models!

Link to blender latest version: https://www.blender.org/download/

Addon installation

This section covers briefly how to install an add-on in Blender:

  • On the top left corner of your screen click the file tab and then go to your user preferences, Image showing the interface of blender, the option saved preferences in a red square
  • Click the button install add-on from file, Window showing that the click install button is in the lower part of the screen
  • Look for the add-on file in your computer (Usually under downloads or where you cloned the project), Image showing the file explorer with the file of the add-on selected
  • Finally, look for the add-on and activate it Image showing the checkbox of the add-on activated

If you did everything correctly, you should be able to see this tool kit (using the default interface): Image showing Magic_I3M interface

Tutorial on how to use the tool

Modules

The addon contains 3 modules with 7 functionalities:

Modification module

Image showing the modification module with 3 buttons: Decimate, export to stl and add tracker scaffold

In this module you have access to 3 functionalities:

  • Adding the tracker scaffold. This tracker scaffold needs to be added to every model in blender so we can map the model with talkit++ and make the model detectable.
  • Exporting to an stl file. STL is the standar format used for 3D printing
  • Decimate model. This function allows us to simplify the model being used, making it lightweight and easy to use. (Note: Don't use this function after adding the tracker scaffold!)

Labeling module

Image showing the labeling module with 2 buttons and 3 textbox: Add label name, Add label description, Add label color, confirm (add label), delete selected areas

In this module you have access to 2 functionalities:

  • You can select any number of faces on the model and add a label by inputing a name, a description and selecting a color and then clicking on the confirm button. A label will be created and automatically linked to the selected faces.
  • You can select any number of faces on the model and click on the delete selected area button. You only need to select 1 face of the corresponding area to delete that label. E.G: You select 3 faces of your object and label them, then you want to delete this action. You select only 1 of the previously labeled faces and click on the delete button. Finally all the 3 faces should lose their color, indicating that the label was deleted.

Export and import module

Image showing the export and import module with 2 buttons and 4 textbox: Name the model file, add a description (intro), output directory path, import file path, export button and import button In this module you have access to 2 functionalities:

  • You can export the selected (1) object as a json file to share your work with other designers and users, so they can reuse and relabel your model.
  • You can import a model providing a json file with the format explained above. You only need to provide the file path in your computer and click the import button.

Implementation

JSON data structure

There are 2 important sections in our json file.

XZ and YZ planes

Image showing json structure of the xz and yz planes. Each is composed with 3 vertices

We use these XZ and YZ planes (Composed of 3 vertices each one of them) so that talkit++ can use this constant point of reference to map the model and assign the audio labels in realtime.

TODO: Elaborate more in the explanation and show more images.

Vertices, faces, areas and materials.

Image showing json structure of the vertices, faces, areas and materials. Vertices are composed with 3 vectors, faces are composed with 3 vertices and the area index the belong to, areas contain each label made by the user

To rebuild the model we need 4 pieces of information.

  • Vertices, which represent each point in the model.
  • Faces, which are composed with 3 or more vertices and make the carcass of the model.
  • Materials, which help us with the labeling of the model (Gives color too).
  • Areas, which contain the information about the label

Contact

If you have any question about using the project, please contact Lei Shi (ls776@cornell.edu) or Ricardo Gonzalez (rgonzalezp1115@gmail.com). They would be happy to explain the project to you.

Credits

Released under: DOI

This add-on was developed in Cornell Tech under professor Shiri Azenkot supervision.