Level 2 Help for SHADOW

A single 1-band grayscale input image. The input image must be in 
HALF format.
  

A single 1-band grayscale output image. The output image is in HALF
format.
 

Gaussian kernel standard deviation. Gaussian kernel is used to estimate
illumation image in the program. Standard deviation defines the blurriness
of the estimated illumination image. Increase standard deviation, the
estimated illumination image will be more blurry. Decrease standard 
deviation, the estimated illumination image will be less blurry.
 

Gaussian kernel size. The kernel size must be an odd integer. The smallest 
kernel size allowed in the program is 3x3. There is no hard restrictions 
for the upper bound of the kernel size, but in general, 7x7 should be 
enough. The larger the kernel size, the more blurry the image will be.
 

Gamma value for performing global gamma correction. In order to brighten
the image, gamma value needs to be in range of 0 to 1 (exclude 0 and 1). 
This parameter is also a switch for the program to perform either 
adpative gamma correction or global gamma correciton. If this parameter 
is defined, the program will perform global gamma correciton. If not 
defined, the program will perform adaptive gamma correction.
 

Optional output for final illumination image. The final illumination
image contains low frequency information, and it reflects the 
enhancement in both brightness and contrast. 
 

Optional output for reflectance image. The reflectance image contains
high frequency information.