search

jiny2001 / dcscn-super-resolution

A tensorflow implementation of "Fast and Accurate Image Super Resolution by Deep CNN with Skip Connection and Network in Network", a deep learning based Single-Image Super-Resolution (SISR) model.
https://arxiv.org/abs/1707.05425
MIT License
706 stars 215 forks source link

Color Degradation #123

Open capucapon opened 3 years ago

capucapon commented 3 years ago

Hi, when i use your code; the black and white outputs are pretty good but i get color degradation in multiple spots on the output images. Is there a way to fix this or am i doing it wrong? This type of issue already mentioned in this section as i see here however didn't get an answer. I atteched some related photos that i get the problem on as an example. sea_boat_aerial_view_141932_512x512_result sea_boat_aerial_view_141932_1280x1024

Mrlawrance commented 3 years ago

Maybe these pixels >255 or <0. You can try to add output[output>255.0]=255.0 output[output<0.0]=0.0 at line 599 in DCSCN.py. May you can fix this problem.

fangichao commented 2 years ago

have you solved this problem? and how to solve it? thks!