Question

I'm looking for some Free (preferably LGPL'd) image analysis/comparison libraries. The most important thing they should be able to do, is compare two different images and say if they're similar, with adjustable sensibility (i.e the higher the sensibility the more they have to look like each other to get a positive result) I'm thinking either C/++/# or eventually Python. So far the only solution I've come across is OpenCV, which is under the BSD license. I have yet to study how much of what I need is already implemented in the library, and what I may need to write myself, but I would really like to hear some other options before I start work on my project, so any opinions are welcome.

Was it helpful?

Solution

Comparing if two images are "similar" is an extremely hard, research-class problem. The approach you'd use will depend heavily on the types of similarities/differences that you are expecting between your images.

It will be hard to find something that works off-the-shelf, unless your comparison requirements are very specific.

Related questions that might be helpful:

OTHER TIPS

I know several people who have used or are using OpenCV. They seem to be very satisfied with what it can do and how easy it is to use. And from the problem you have described, OpenCV surely has that feature, and much more. I don't know of any library that compares to it. As for the BSD license, don't blame people who make source code available for free for not letting you make money off of it.

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top