Core metadata about images

[Daniel Bauer]

Am 05.02.26 um 15:24 schrieb Alan Seal:
> I am trying to catalogue my photo collection (going back to the 40s and 
> 50s). Many are from scans of CTs or prints. I want to get input from 
> sisters and family to add details which I don't know about the photos. 
> They will not be using Digkam. I don't want to put the photos in the 
> Cloud (too many).
> 
> I want to add the following information to each image:
> caption
> title (optionally)
> date taken
> place in image
> people in image
> optionally keywords uncontrolled (e.g. flowers, cars, lunch, cycles etc)
> 
> I want to send this info to them so that they can add or correct 
> information. I thought the best way would be as a spreadsheet, pdf or 
> html/xml file. Too big to print but they will have software such as 
> Excel, Word etc.
> 
> It is hard to do this in Digikam. Best thing I have found is the table 
> layout. Close, but I cannot output it to a file. And I cannot specify 
> the fields to print/export. Same with HTMLGallery - Details but I cannot 
> see how to easily customise it.
> 
> A bigger problem is the way the data is stored in the metadata. I can 
> use Exiftool or a Python script to extract specific fields but I cannot 
> find a way to keep the fields separate. Place, people and keywords are 
> all put into one or more various IPTC or XML fields. This is more a 
> problem with the metadata (multiple) standards more than Digikam. 
> Sometimes the data is in the form People/Name, People/Name and sometimes 
> just Name/Name.  But people  places and keywords are not the same (in 
> database terms) and should be kept separate. People are put into 
> RegionPersonDisplayName and into RegionName so I could retrieve them.
> 
> I shall either have to not export 'tags' or find a routine to unravel 
> Place from the string of keywords. I will work on how to create the XMP 
> fields Country/City/Location... rather then using the country in 'tags' 
> in Digikam. I will get there in the end I hope.
> 
> Please let me know if by using filters or anything how to print or 
> export this information. I can put together an example of what I am 
> aiming for.
> 

In my honest opinion, as you want the data in a text version (as csv or 
similar), I'd use a direct database query to extract that data and put 
it in the desired order.

If you don't know how to query a database, nowadays you can ask AI to 
help you through the process.

If you don't know how to extract the organization/description of your 
database (to give it to the AI), you could create a new user, install an 
empty digikam as this other user, put like 3 unimportant photos in it 
and do your most complicated tagging things as examples. Then you can 
give that very small database to Grok or ChatGPT, so that it can study 
the organization. With this you can say what data you actually want to 
get and ask for a step-by-step howto.

Then (AFTER making a backup of your current real database! or directly 
working on a copy - important) you can try the procedure that AI 
recommends to you, check the results, and refine them in the AI chat.

As it is a quite easy SQL job, AI should be able to help you 
effectively. It can also tell you what free tools you need to do it. 
It's not such a big thing.






-- 
Daniel Bauer photographer Basel Málaga
Twitter: @Marsfotografo (often explicit nudes)
https://www.patreon.com/danielbauer
https://www.daniel-bauer.com (nudes)