-
Notifications
You must be signed in to change notification settings - Fork 0
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Zernike reconstruction function and OpenCV images #1
Comments
Frederico, Thanks for the tip. Do you think you could give me a couple of use cases? Or, better yet, if you want to work on it yourself, I'd be happy to have a Sincerely, Ravi On Sat, Oct 17, 2015 at 7:49 PM, Frederico Schardong <
Ravi Sankar Jonnal, Ph.D. |
Ravi, From a software engineer concerned with Computer Vision only, reconstructing an image based on its Zernike moments can be very useful. For example, the project I am working right now uses quadtrees to "break" an image into smaller chunks until the Zernike moments of a given chunk is similar enough (Euclidean distance) to its Zernike reconstruction. It will only hold in the database a few Zernike moments (there is a separated logic to tell which moments are the most important) and considering that those moments can recompose their original sub-images with acceptable similarity I have got myself an image search engine. I am not the author of all this, though. It has been proposed in this paper and my project is about trying to reproduce it. OpenCV images are simply 2-ndarray. Here is the mahotas implementation I was referring in my first comment. It is as high level as a software engineer could expect. If your module could provide a function to evaluate the Zernike moments and another to recompose the image somewhat on the lines of mahotas, it would be fantastic. However, if you think that's not suitable for this module perhaps I can persuade you to contribute to mahotas as the author already stated that he won't write any reconstruction code. Thank you. |
You came up with a nice implementation of Zernike moments but more important you coded the reconstruction part, which is fantastic!
It would be awesome if you could wrap your Zernike decomposition code into a function or two and the reconstruction logic into another function/class. I was using mahotas to apply Zernike moments to my images but it doesn't provide any reconstruction functionality. On the other hand, it is ready to be used on openCV's images. Would it be possible to wrap your code to accept such images?
The text was updated successfully, but these errors were encountered: