interactive-kit 0.1rc4

Interactive viewer for signal processing, image processing, and machine learning

InteractiveKit

A toolkit for interactive visualization of signal and image processing on Jupyter Notebooks.

InteractiveKit was created to simplify visualization in image and signal processing teaching, learning and research. Using Jupyter Notebooks in combination with InteractiveKit, a user with virtually no programming experience, and without experience with matplotlib, will be able to display and manipulate one or several signals -in the form of NumPy Arrays- and interactively change their characteristics. It is even possible to extract information and perform operations directly on the signal, without the need to re-run the cell or to plot again.

The class is designed to run in Jupyter Notebooks or Jupyter Lab, using matplotlib's dynamic environment (requires ipympl, which can be activated with the magic command %matplotlib widget). All the functionalities are controlled either through matplotlib's native widgets (zoom and change of figure size) or through additional ipywidgets' buttons and sliders.

Modules

Image Viewer (imviewer)

Optimized for image visualization and manipulation. See the dedicated tutorial and wiki.

Main features

Once called, from the imviewer and using both widgets and programmatic commands, a user will be able to:

• Plot several images at the same time, and choose different display options (one image at a time, or a customized grid of images),
• Change the brightness and contrast of the images through a slider,
• Explore the histogram of the image,
• Choose different colormaps and visualization options (colorbar, axis),
• Get 1st and 2nd order statistics -updated automatically when zooming into a region- of the image,
• Calculate and visualize differences between two different images,
• Declare functions that perform operations on an image, and through custom widgets, see the effect of the function with different parameters applied on different images, directly inside the imviewer object.

Signal Viewer (sigviewer)

Optimized for 1-dimensional signal visualization and manipulation. See the dedicated tutorial and wiki.

Installation and usage

First, make sure you have installed Python 3.6 or higher. Then, InteractiveKit can easily be installed through PyPI:

pip install interactive-kit==0.1rc3

To use in a jupyter notebook, you can import the modules in the following way:

from image_viewer_kit import imviewer, sigviewer

Team

The viewer was initially developed by:

• Alejandro Noguerón Aramburu ([e-mail], Alejandro-1996)

under the guidance, help, testing and feedback of:

• Kay Lächler ([e-mail], TheUser0571)
• Pol del Aguila Pla, ([e-mail], poldap)
• Daniel Sage , ([e-mail], dasv74)

This project is supported by the EPFL's Center for Digital Education (CEDE), and it belongs to the Biomedical Imaging Group.

Members of the EPFL community

If you want to start using InteractiveKit rightaway, without going through the process of installing Python and Jupyter, you can click here and start using right away from Noto, EPFL's Jupyter centralized platform.

Contributions

We appreciate contributions, feedback and bug reports from the community:

• If you encounter any bug, please open an issue and describe. We will try to fix it or give you a workaround as soon as possible.
• If you wish to contribute, fork the repository and then open a pull request.