Just looking on the opencv superres web page, and there’s a medical image of a knee bone that’s been not only upscaled, but it’s also been denoised as part of the upscalcing process. I can’t see any option for that. Anyone know?
oh that link was supposed to take you to the text on the page, but it’s just going to the top.
The image is right near the bottom under the applications heading.
that “Source” link goes to some paper from 2016 that even I can’t access through my university, and they are subscribed to almost everything.
edit: the learnopencv.com article uses a sci-hub domain that no longer works. the real sci-hub is currently on other domains. here’s a link that works:
https://sci-hub.ru/10.1016/j.sigpro.2016.05.002
when that stops working, the DOI is 10.1016/j.sigpro.2016.05.002
that knee image is from 10.1109/TIP.2014.2308422, or freely accessible at Sci-Hub | Novel Example-Based Method for Super-Resolution and Denoising of Medical Images. IEEE Transactions on Image Processing, 23(4), 1882–1895 | 10.1109/TIP.2014.2308422
Abstract— In this paper, we propose a novel example-based
method for denoising and super-resolution of medical images.
The objective is to estimate a high-resolution image from a
single noisy low-resolution image, with the help of a given
database of high and low-resolution image patch pairs. Denoising
and super-resolution in this paper is performed on each image
patch. For each given input low-resolution patch, its highresolution version is estimated based on finding a nonnegative
sparse linear representation of the input patch over the lowresolution patches from the database, where the coefficients of
the representation strongly depend on the similarity between
the input patch and the sample patches in the database. The
problem of finding the nonnegative sparse linear representation
is modeled as a nonnegative quadratic programming problem.
The proposed method is especially useful for the case of noisecorrupted and low-resolution image. Experimental results show
that the proposed method outperforms other state-of-the-art
super-resolution methods while effectively removing noise.
Index Terms— Example-based super-resolution, denoising,
medical imaging, sparse representation.
you’ll get hallucinated data, information that was never in the picture at all. you should never rely on what you’re seeing in those results. never. least of all when it concerns the health of a living being.
Thanks. So forgive me, so far my knowledge extends to using an existing model with opencv to upscale images and video with superres. Are those guys using a completely different model that I can try with sr = dnn_superres.DnnSuperResImpl_create() ?