While developing these web pages I often find myself needing to check the 
contents of images and video so these following items come in handy!

Some browsers appear to have image describers build in but so far I've
only managed to extract the text from alt tags direct from the HTML code.

This page is only just off the ground so will be subject to lots of changes
and additions as my research progresses.

Presume that a lot of the following services are linked to the same AI engines
in the background.

Noticed if you ask the same provider to describe a picture twice the answers can vary,
this might be due to the rapid improvements of the AI engines supporting
these systems.

Researching on a Windows 11 PC running NVDA 2025.3.2 and a Samsung Galaxy
S23 Ultra running Android 15.0.

Gives full and highly detailed description of most images within a few 
seconds and simple to use.

To get description (windows 11):-
* right click on an image and find 'Ask copilot';
* at the prompt type 
   'Describe image' 
     or
   'Where was this photo taken?' seems to work if there are no people
    in the picture.

You should get the description! this text can be copied and pasted to other
locations.  For those using screen readers you'll need to tab around to find 
the line of descriptive text and then select and copy the line, from there
you can paste the text anywhere.

In this delightful garden scene, a dog proudly carries an impressively 
large branch in its mouth, as if it’s just claimed the ultimate 
prize.  The grassy yard is neat and vibrant, with cheerful daffodils adding 
a splash of yellow near the edges.  A yellow garden hose snakes across the 
lawn, lending a casual, lived-in feel. Behind the action, there’s a 
charming brick house with white-trimmed windows, partially framed by a 
dense hedge that separates the yard from neighboring homes. The sky is mostly 
clear with a scattering of soft clouds—perfect dog-fetching weather, really.

Now interestingly when I asked copilot again it gave me a slightly different
answer as shown below:-

This is a delightful and lively scene, Martin. A proud German Shepherd stands 
confidently on a neatly trimmed lawn, clutching an impressively large branch 
in its mouth—as if it's just completed a major feat.  Around the dog, 
daffodils nod their yellow heads in the sunshine, adding a bright splash 
of springtime.  A yellow garden hose snakes across the grass, suggesting 
some recent outdoor activity or watering.  Behind the pup, a tidy low brick wall 
borders the garden, and further beyond are neighboring houses, a well
-kept hedge, and a brilliant blue sky dotted with a few soft clouds. It’s the kind of sunny suburban moment that feels both everyday 
and full of warmth.

Did this cheerful champion bring the branch all by itself? It looks quite pleased with its treasure!

For the Chrome and Brave browsers there is an Image describer extention
(Find it via google search); Right clicking on an image you'll find an
option 'Image describer', four options here, full page description, , image,
stop playback  and options.

Also available via the options is the ability to ask specific questions 
about the image just viewed.

Adds extra audio description to video.
The process involves uploading your video, some remote processing and away 
you go!  There is a free trial of 50 minutes worth of video and then it starts
to cost.

Available on IOS, Android and via Whatsapp on Windows.

Two levels free and pro;
* Free has very good basic functions and works well including via Whatsapp;
* Pro has more bells and whistles and access to several AI engines, ability
  to handle longer videos etc.  prices vary from monthly to Lifetime.

this app looks very interesting and powergful;

This app can describe images and videos either held on your device already
or captured at the moment.

Via whatsapp is interesting, sign up via the Piccybot web site and then
send a message to Whatsapp with a photo attached; Piccybot sends you an audio 
message describing the photo in return - very detailed and quick!  Not tried video yet.
The Whatsapp message appears to be sent to India for processing.

These descriptions are very detailed and can be a bit over the top - looking
to see if settings have any control over this level of detail.  Changing the
AI engine in Pro might do this.

When using my Samsung S23 ultra processing time is quite long but clearly there is a lot of work going on
behind the scenes to produce such detailed work.

Currently continuing testing (19/06/2025);

Must try follow up questions;

Put text here shortly!

This is a must have addon for NVDA!

Provides access to several methods of image description - still got a lot of 
research here as it has so much potential.

Follow up questions are useful even with indepth original descriptions;

Caution:- make sure the graphic you are looking at is not just a thumb nail
image, some strange results if this is the case i.e. seeing a dog as a child
when viewing the thumbnail.

Trying Google pro vision engine with amazing results, very fast and accurate,
May need to purchase API key at some stage, currently on evaluation   - best so far (27/06/25);

Need to play with different AI models available.

For full details see:-