Dash for Openseadragon?

Hi all,

I’m working on utilizing the capabilities of Openseadragon a framework that allows users to interact with images with a large number of pixels (my images being in order of 1-10 gigapixel SEM stitched images.) From what I have gotten to work together so far, I can utilize the Deep Zoom Composer tools from Microsoft to compose an image pyramid configuration that gets used by Openseadragon to display images on demand based on the zoom level.

Having had no background in JS, one limitation I’m running into is provide some tools for users to interact (annotate, measure, etc.) on images in a collaborative manner. There are some tools already available, such as Annotorious, but I’m wondering if Dash could be a potential avenue to explore options in a Python setting.

I also recently learned that there is a Dash-Canvas project that’s active as well, and I’m wondering if the Dash-Canvas could act as an ‘overlay’ on the openseadragon framework already in place, for example, or maybe there is an even cleaner solution that exists already.

In summary: I’m looking for a Dash-based solution for annotating, measuring, etc. on images. It looks like there is already a lot of work in JS (Openseadragon in the wild,) but I’m wondering if there is a bit more Python-friendly option available through Dash.

Would anyone care to share their experience or provide any guidance on my activity? Thank you again for all the hard work on making this project come true!

Depending on your needs for annotation, you could take a look a Dash Leaflet

or the default map box plots in combination with a local tile server. I have previously used Terracotta as the tile engine for serving large images/datasets, but I guess any tile server could do.

Dear Emil,

This is a huge help in directing me to learn more about the solutions available. I think the capabilities of leaflet is exactly what I need for my application, and my next goal is to figure out how to set up a local tile server to populate the same way. Thank you again, and I will mark this as a solution. Cheers and thank you! :slightly_smiling_face:

1 Like

I have just written a guide for visualization of geospatial raster data. You can find it here,

It uses weather data from GFS, but it should be straight forward to adopt to other data sources :slight_smile: