learnopencv
learnopencv copied to clipboard
BRISQUE Output Widely Differs from Blog
For https://www.learnopencv.com/image-quality-assessment-brisque/
the metrics publish above for the tree images at the end don't match the output of the code, by quite a wide margin:
REPORTED: original-rotated-image.jpg 26.8286 JPEG2k-compressed.jpg 30.7417 compressed-heavy.jpg 33.0692 noise-version.jpg 79.8751 blur-vision.jpg 72.7044
ACTUAL OUTPUT FROM SCRIPT: original-rotated-image.jpg 54.2116 JPEG2k-compressed.jpg 55.0107 compressed-heavy.jpg 54.9763 noise-version.jpg 70.0947 blur-vision.jpg 63.9073
Moreover, I have tested the same images with the image-quality PIP package from https://github.com/ocampor/...
and the numbers you actually output are more in accord with his output, which seems good to me: original-rotated-image.jpg 54.6798 JPEG2k-compressed.jpg 53.9756 compressed-heavy.jpg 55.0188 noise-version.jpg 70.2104 blur-vision.jpg 63.5786
Have I misunderstood something? Forgive my ignorance, if so. Thanks for your generosity.