10k images

[Cecilia Pitta]

Hello,

I am trying to help a colleague who recently reach out for some advice. 

He is been converting 10 k images to jpeg2000. He is working with music manuscripts and is sorting the images by copyists into folders. At the moment he is storing the 10k images in a external drive.

Additionally he needs to crop certain regions of the images (a clave, e.g.).

I told him he may need a DAMS to manage his images and also converting the images more rapidly. 

I suggested that he may use IIIF to help with interaction and also with speed (it takes too long to open an image). 

I told him iiif would be a good solution to help to create on fly derivatives and crop and store regions.

How can he start? what advice would you give to him ? what kind systems will he need?

At first, all the images are private and the near future will be available to the public.

Thank you so much for your attention,

Cecília

[Andrew Hankinson]

Hi Cecília,

What are the specific things that your colleague is having trouble with in their existing workflow? It seems like they have something that works for them, so I would be wary about shifting their workflow too much without a clear idea of what you're trying to fix. Why is he using JPEG2000?

Many "off the shelf" JPEG2000 encoders and decoders are slow. Unless you're running a IIIF server that has access to the Kakadu decoder, or the latest OpenJPEG enhancements, I'm afraid your colleague won't see many speed benefits. 

It would also depend on how technically savvy your colleague is. Setting up a IIIF image service may be a distraction to their current workflow, even if it is inefficient. There are some tools out there that would make this a bit easier (e.g., a Docker container that runs a IIIF image server) but it still requires some technical know-how, and depending on how it's implemented may not be significantly faster (unless you know how to "tune" it).

Your colleague would also probably need to generate IIIF manifests if they wanted to use Mirador or Universal Viewer. This may also be a significant distraction to their workflow, particularly if they only need local, private access. 10k images is a lot, but it's not an impossible number to manage "by hand". 

So before advising your colleague to go down the IIIF route, I would do a basic analysis:

 - Do the technical resources they have available (self, or others) provide adequate support for this shift in their workflow

 - Will the time spent implementing a IIIF workflow save time in the long run, or will it add delay to their project (even if the existing workflow is "inefficient")

 - Do the benefits of IIIF (linkable, sharable images; existing viewers; web-native image delivery) align well with the requirements of the colleague?

 - Do the requirements of IIIF delivery align with existing goals of your colleague's project? (as in, is implementing IIIF just a technical distraction, or will it need to be done anyway? If it needs to be done anyway, also check that it's OK to do it NOW, since taking a different technical route can also be a distraction to the main goal of this phase of their research.)

-Andrew

--
-- You received this message because you are subscribed to the IIIF-Discuss Google group. To post to this group, send email to iiif-d...@googlegroups.com. To unsubscribe from this group, send email to iiif-discuss...@googlegroups.com. For more options, visit this group at https://groups.google.com/d/forum/iiif-discuss?hl=en
---
You received this message because you are subscribed to the Google Groups "IIIF Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iiif-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/iiif-discuss/3cc3795f-0bd1-4cfc-8c21-25760a35b7b3n%40googlegroups.com.

[Benjamin Kozlowski]

Hi, 

All depends on what the colleague wants to do with the images. 

For a one time conversion/once a month task I’d just use the batch function of photoshop. 

Same with the cropping. 

But what I would need to give advice: 

What should happen to the cropped image? How many does he have to crop? Does he want to do it by himself or others as well? Should users be able to crop it? Should they be easily shareable? How much skills can he gather by himself or friends or the department? What does it mean that the images should be public? A website? GitHub? A repository? What’s his budget (for server, maintenance, development)? 

Since your colleague has from a technical perspective very basic problems I‘d say that IIIF is overkill and way above his „paygrade“. If he can’t solve these he’ll never get a IIIF server up and running, let alone the frontend with maybe some integrations and/or customizations.

So my advice for now would be to just use photoshop. It’s a reachable goal to batch convert thousands of images. If the cuttings are doable depends on the open questions. 

Regards 

Ben 

--

[Danny Michaels]

Hello,

The system that powers the National Institute for Newman Studies (https://digitalcollections.newmanstudies.org/) will soon be available and it's perfect for anyone without knowledge of IIIF. It's a IIIF-first application and handles asset management, user management, access control, image import/conversion, bulk import, annotation, OCR, non-destructive image editing (crop, flip, rotate, etc., using IIIF API instead of using an image editor), and so much more. The entire ecosystem of servers (Cantaloupe, MongoDB, ElasticSearch, and many others) are integrated for you---no special knowledge required. Curators have control over the major parts of the CMS and all aspects of documents (images, metadata, privacy, etc.). It's auto-scaling, so no image count is too large (or too small). It has an API that can be used to connect with other institutions (e.g., we can pipe content, search results, Marc records, etc, directly to Duquesne University with whom we are affiliated). You'll recognize some things, like Mirador, but also many custom things like data visualization. We haven't advertised it yet because we already have a queue of clients. We are also working with a solutions partner so that we can stay focussed on expanding and perfecting the platform while they manage clients. You can see it in action with our own collections here: https://digitalcollections.newmanstudies.org/

In short, we asked many of the same questions five years ago and couldn't find a "simple" solution, so we developed this platform. It got so much attention with our partners that we decided to make it available, first with them and then with anyone that qualifies. 

Email me directly if you'd like more information.

Kind Regards, 

Daniel T. Michaels, Ph.D.

Chief Technology Officer

National Institute for Newman Studies

211 N. Dithridge St.

Pittsburgh, PA 15213

dmic...@newmanstudies.org

https://www.newmanstudies.org

To view this discussion on the web visit https://groups.google.com/d/msgid/iiif-discuss/CAHn1_Vpno_NjSZDBoVoSR-veSJbWuXj2HbkjxAgH9Pzj2TKJTg%40mail.gmail.com.

[Cecília Pilar C. Pitta]

Thank you all. I will send all this information to my colleague

:)

cheers,

Cecília

To view this discussion on the web visit https://groups.google.com/d/msgid/iiif-discuss/CAO2wOHr%3DohuZxdcaqNd6SuhrBOtvFZb8Y973yGQW4REji5VPDA%40mail.gmail.com.