digiKam, the perfect tool for Wikimedia Commons photographers?

Guillaume Paumier

– September 22, 2009Posted in: Featured, KDE, Photo, Wikimedia Commons

From your computer to Commons using digiKam

For several years now, I have been a photographer for Wikimedia Commons. Commons is actually the reason why I became interested in photography, and why I bought a DSLR camera and various lenses. And I have ended up taking a lot of pictures.

The problem when you take a lot of pictures is that the time you need then to process and upload them grows exponentially. Besides, if you want to upload your pictures to several platforms (say, Commons, flickr and your blog), you have to deal with the specifics of each website. It would be much simpler and faster to describe, tag, classify, geotag the pictures only once (using an appropriate batch-editing tool) and then to have each website extract the metadata, or to use an upload tool that extracts them automatically.

Wikimedia Commons

So, on the one hand, you have Wikimedia Commons, a repository of freely-licensed multimedia files (photos, maps, diagrams, sounds, videos, etc.). It is a project of the Wikimedia Foundation, from which uploaded files can be used across all Wikimedia projects in all languages, including the famous encyclopedia Wikipedia.

Wikimedia Commons contains over 5 million files, the overwhelming majority of which have been contributed by volunteers. Such a large amount of files makes it mandatory to properly:

describe what each work is about
identify the author and the copyright status of each work
classify each work in order to be able to find it later and to manage such a large collection.

Besides these requirements, it is also considered a good practice to geotag your pictures (if applicable), i.e. to indicate the geographical location of where the picture was taken.

digiKam

On the other hand, there is digiKam, a FLOSS “digital asset management” application whose special features allows to annotate, classify and catalogue a large media library, although it provides some very handy editing and post-processing tools as well. digiKam uses the KIPI plugins, which provide handy features such as batch editing or export modules to a variety of websites, including Flickr, Picasa and Facebook. The 4th beta release of digiKam 1.0 was recently released and the official 1.0 release is planned for late October this year.

I won’t go through digiKam’s impressive list of features; instead, I will focus on those that make digiKam a very powerful tool for Wikimedia photographers.

The perfect combination?

So let’s take a quick look at the features digiKam offers that would be handy for people who want to contribute their pictures to Wikimedia Commons:

Commons requires or recommends…	digiKam offers…
copyright information (author and license)	full metadata read/write support, including Author and Copyright fields
multilingual descriptions for each file	multilingual descriptions since digiKam 1.0, with appropriate language codes
classification with hierarchical categories	classification with hierarchical tags
geotagging¹	geotagging (automatically using GPS data or manually using Google Earth)

So, what’s missing? Well, you can do a lot of stuff with digiKam to prepare your pictures before they’re uploaded to Commons. Yet, you still have to upload them manually, or using an external tool such as Commonist. And even if you have been spending days to annotate your pictures, describe them in several languages, geotag them, etc., well, all this information remains in the image metadata. Neither Commonist nor MediaWiki knows how to extract these metadata meaningfully (yet); as a consequence, you have to redo all the work when uploading each file to Wikimedia Commons. The Ford Foundation recently awarded a $300,000 grant to the Wikimedia Foundation in order to improve the usability of the upload process. I expect some work will be done regarding the extraction of metadata from files. Yet, when uploading a large number of files, it would be more convenient to provide a standalone desktop upload tool like Commonist. Or…

What would really be neat is an integrated export module to Wikimedia Commons in digiKam; not only would it serve as a batch upload tool, but it would also read all the metadata and create the file description pages accordingly. A few weeks ago, I contacted the main developer of digiKam in order to discuss the possibility of implementing such an export module. He showed interest in this project even if he didn’t have the time to code it himself. As a consequence, I have initiated a roadmap and brainstorming page for the plugin.

Feel free to contact me if you’re interested in helping implement such a plugin.

Notes

I recently noticed a JavaScript tool allowing users to geotag themselves, using a map from OpenStreetMap. This is exactly the kind of tool I would like to have for pictures; I can’t understand why it exists for users, but not for media files. ↩

Mike Peel

Posted September 22, 2009 at 5:00 PM

I’m in a similar position – dSLR, and lots of photographs – e.g. I recently took ~ 500 photos in a weekend of various things, that I’d love to be able to easily upload to Commons.

I use Apple’s Aperture software for all my photographs – and have recently been having a similar desire to upload directly from the image management software to Wikimedia Commons. It would be really nice to see a set of such plugins from the various image management software – it would do a lot to encourage the uploading of photographs from those that do a lot of photography, both amateur and professionally.

Sage Ross

Posted September 22, 2009 at 5:56 PM

“Commons is actually the reason why I became interested in photography, and why I bought a DSLR camera and various lenses.”

Me too!

I absolutely agree that we need better workflow tools for Commons. I use Commonist for big batch uploads, but it’s really a pain and I would love to be able to upload to Flickr and Commons at the same time… and better yet, have all the relevant info in the metadata too (without the extra trouble of firing up a different program).

I’ll have to check out digiKam.

Jastrow

Posted September 22, 2009 at 6:39 PM

J’ai un workflow particulièrement long et compliqué, puisque je dois d’abord trier mes photos (en photo de musée, je prends de 2 à 5-6 fois le même objet, donc je dois les comparer et retenir la meilleure) et les développer (je prends en RAW) avant de les télécharger. Comme j’utilise le modèle spécial musée, je ne pense pas pouvoir utiliser Commonist.

Pour digiKam, je vois bien l’intérêt de l’édition d’EXIF et du geotagging, mais le reste m’échappe un peu…

Guillaume Paumier

Posted September 22, 2009 at 6:58 PM

Tu peux utiliser des modèles spéciaux avec Commonist; il te faut modifier à la main l’un des fichiers de configuration, celui qui gère la page de description de l’image (/etc/image_commons.bpp). Après, c’est vrai que pour le modèle spécial musée, ça risque d’être compliqué, car il y a des champs qui ne sont pas ceux par défaut de Commonist. La solution serait d’avoir une version de Commonist (ou du module d’export KIPI) qui gère les cas particuliers comme le tien.

Pour digiKam, l’intérêt majeur pour moi est de pouvoir décrire les images (individuellement ou collectivement) dans digiKam, et à l’avenir de pouvoir mettre les descriptions en plusieurs langues. Ces descriptions sont ensuite enregistrées dans les métadonnées de l’image. Il suffirait d’un peu de magie technique pour que le module d’export ou MediaWiki lise automatiquement ces métadonnées et les utilise pour la page de description de l’image ; ça m’éviterait de devoir retaper toutes les descriptions au moment de l’upload. Idem pour les catégories, qui pourraient être enregistrées à l’avance dans les tags.