Hello,
I'm doing some images with 'hip'. Very good library, thanks.
Me and another guy on StackOverflow, we tried to reproduce an 'hip' image with R. We start from an image file, we import it in Haskell/R, and we apply the LoG filter. Though the obtained images are similar between Haskell and R, they are not exactly the same and I'm wondering why. Therefore I tried to understand the Haskell source code of correlate. I'm not an Haskell expert and I don't understand what toManifest does. You started a sentence in the doc but it is truncated:
-- | Each array has a sibling `Manifest` array representation, which
toManifest :: Image arr cs e -> Image (Manifest arr) cs e
And I didn't figure out by looking at the code.
What is strange is that we use the standard definition of the correlation, and we don't get the same RGB values.
Hello,
I'm doing some images with 'hip'. Very good library, thanks.
Me and another guy on StackOverflow, we tried to reproduce an 'hip' image with R. We start from an image file, we import it in Haskell/R, and we apply the LoG filter. Though the obtained images are similar between Haskell and R, they are not exactly the same and I'm wondering why. Therefore I tried to understand the Haskell source code of
correlate. I'm not an Haskell expert and I don't understand whattoManifestdoes. You started a sentence in the doc but it is truncated:And I didn't figure out by looking at the code.
What is strange is that we use the standard definition of the correlation, and we don't get the same RGB values.