Open Selur opened 3 years ago
I'm trying to reproduce the results from the paper, but using using http://mcl.korea.ac.kr/projects/dehazing/results/horse.jpg from http://mcl.korea.ac.kr/projects/dehazing/ and:
# Imports import vapoursynth as vs core = vs.get_core() #loading filter core.std.LoadPlugin(path="I:/Hybrid/64bit/vsfilters/MiscFilter/DeHazing/DehazingCE.dll") # loading source clip = core.imwri.Read(["C:/Users/Selur/Desktop/horses.jpg"]) clip = core.std.Loop(clip=clip, times=100) clip = core.std.AssumeFPS(clip=clip, fpsnum=25, fpsden=1) before = clip # applying filter clip = core.dhce.Dehazing(src=clip,lamda=8) # comparing results before = core.text.Text(before, "Before") clip = core.text.Text(clip, "After") clip = core.std.Interleave([before,clip]) clip.set_output()
I got totally different coloring than in the paper (http://mcl.korea.ac.kr/projects/dehazing/results/horse_8.jpg). Adding some auto white (http://www.vapoursynth.com/doc/functions/frameeval.html) and boosting the saturation, I get similar results, but the color are still way more blue instead of green.
Is this a missing feature, a bug or am I missing something?
I just tested with the city image you posted to the description of the filter, there the colors seem fine.
Indeed, there are similar situations when I code. I guess it is a bug.
Maybe I need some time to find the reasons.
I'm trying to reproduce the results from the paper, but using using http://mcl.korea.ac.kr/projects/dehazing/results/horse.jpg from http://mcl.korea.ac.kr/projects/dehazing/ and:
I got totally different coloring than in the paper (http://mcl.korea.ac.kr/projects/dehazing/results/horse_8.jpg). Adding some auto white (http://www.vapoursynth.com/doc/functions/frameeval.html) and boosting the saturation, I get similar results, but the color are still way more blue instead of green.
Is this a missing feature, a bug or am I missing something?
I just tested with the city image you posted to the description of the filter, there the colors seem fine.