Open d-vazquez opened 1 year ago
Seems like your rgb2grey function just returns the blue channel?:
https://github.com/vh-diazr/dehazing-LocallyAdaptive/blob/80cc51fb6b734f9220c04f1222eaff545a2ca2bc/dehazing-locally_adaptive.py#L47-L49
Just multiplying same channel by the Greyscale values,
Also, trying to reference you on my paper where I implement this in a low cost RTOS, is the CUDA code available in CITEDI's library? I'm also from TJ
Thanks for this observation. The correction of line 48 is as follows: img_gray = np.double( 0.2989img[:,:,0] + 0.5870img[:,:,1] + 0.1140*img[:,:,2] )
There is no CUDA implementation available.
Seems like your rgb2grey function just returns the blue channel?:
https://github.com/vh-diazr/dehazing-LocallyAdaptive/blob/80cc51fb6b734f9220c04f1222eaff545a2ca2bc/dehazing-locally_adaptive.py#L47-L49
Just multiplying same channel by the Greyscale values,
Also, trying to reference you on my paper where I implement this in a low cost RTOS, is the CUDA code available in CITEDI's library? I'm also from TJ