Hello,
I'm just a beginner with SQLAlchemy (and not very experienced with
databases in general) so please forgive my ignorance. Also this is my
first post here so I hope I'm not off-subject.
I work on image processing. It is very common for us to use large
databases of images (several hundred thousands) and then use complicated
(and time consuming) algorithms to compute descriptors (i.e. a vector
representing some information on the image).
I already use SQLAlchemy to store the images and their relations on the
DB (in fact I store only the path) and I would like to say that it is
really an amazing tool. From the modelling point-of-view this part is
pretty easy.
Now, the hard part is of course the descriptors. First the goal is to
avoid computing them again and again. But we need flexibility.
To describe some of the concepts I think it's better to point to the
OpenCV image processing library (that I use):
http://opencv.willowgarage.com/documentation/cpp/features2d__feature_detection_and_descriptor_extraction.html
To sum-up:
- images can be pre-processed
- from an image we compute several key-points (with different
algorithm, parameters, etc.)
- from a keypoint we compute a descriptor (with different algorithm,
parameters, etc.)
- key-points can be post-processed (e.g. normalize them by removing
the mean and dividing by the standard deviation)
I'm not sure what does this mean in terms of relationships between objects.
Ideally, I would like to be able to select a group of images and then
specify a key-point extraction algorithm and a descriptor computation
algorithm to retrieve the descriptors.
So my question is: how would you model (and implement with SQLAlchemy)
this kind of database?
Thanks in advance,
Mathieu
--
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/sqlalchemy?hl=en.
