USAGE: adaptivegamma2 [-t type][-m meanthresh] [-f fraction] [-c cumthresh] [-bg brightgain] [-dg darkgain] [-ba brightalpha] [-da darkalpha] [-C colormode] infile outfile

USAGE: adaptivegamma2 [-h or -help]

-t .... type ............ type of mean threshold to use; choices are: orginial or simple;
......................... default=simple
-m .... meanthresh ...... threshold of the mean used to test if the image is bright
......................... or dark; 0<=integer<=255; default=128 (use 112 for original)
-f .... fraction ........ fraction to use with the meanthresh to test if the image
......................... is bright or dark for type=original; ignored if type=simple;
......................... 0<=float<=1; default=0.3
-c .... cumthresh ....... threshold used to clip the cumulative distribution for dark
......................... images; 0<=float<=1; default=0.55
-bg .... brightgain ..... gain on cumulative distribution for bright images;
......................... 0<=float<=1; default=1 (no change)
-dg .... darkgain ....... gain on cumulative distribution for dark images;
......................... 0<=float<=1; default=1 (no change)
-ba .... brightalpha .... alpha exponent for weighting the normalize histogram for
......................... bright images; 0<=float<=1; default=0.25
-da .... darkalpha ...... alpha exponent for weighting the normalize histogram for
......................... dark images; 0<=float<=1; default=0.75
-C .... colormode ....... colorspace for processing; choices are srgb, lab, hcl, hsl,
......................... hsv, hsi, ycbcr and ohta; default=hsv

PURPOSE: To enhance the contrast/brightness in an image using an adaptive gamma method.

DESCRIPTION: ADAPTIVEGAMMA2 is an adaptive gamma technique used to enhance an image's brightness and contrast. It is based upon an exponentiation of the the image's values using an exponent, gamma, that is evaluated differently for each graylevel in the image based upon values in the cumulative normalized and weight histogram in the form, gamma(value)=1-cumulative_histogram(value). The fraction along with the meanthresh value are used to categorize the image as bright, dark (or neutral if type=original). Neutral images will have no processing applied. For dark images the cumulative distribution is clipped not to exceed cumthresh. For bright images, no clipping is used, but the input is negated before gamma processing and then negated back after processing.

ARGUMENTS:

-t type ... TYPE of mean threshold to use. The choices are: orginial or simple. The default=simple.

-m meanthresh ... MEANTHRESH is a threshold of the mean that is used to test if the image is bright or dark. Values are 0<=integer<=255. The default=128. (Use 112 for original).

-f fraction ... FRACTION to use with the meanthresh to test if the image is bright or dark. Values are 0<=float<=1. The default=0.3.

-c cumthresh ... CUMTHRESH is a threshold used to clip the cumulative distribution for dark images. Values are 0<=float<=1. The default=0.55.

-bg brightgain ... BRIGHTGAIN ia a gain on cumulative distribution for bright images. Values are 0<=float<=1. The default=1 (no change).

-dg darkgain ... DARKGAIN is a gain on cumulative distribution for dark images. Values are 0<=float<=1. The default=1 (no change).

-ba brightalpha ... BRIGHTALPHA is an alpha exponent for weighting the normalize histogram for bright images. Values are 0<=float<=1. The default=0.25.

-da darkalpha ... DARKALPHA is an alpha exponent for weighting the normalize histogram for dark images. Values are 0<=float<=1. The default=0.75.

-C colormode ... COLORMODE is the colorspace for processing; choices are srgb, lab, hcl, hsl, hsv, hsi, ycbcr and ohta; default=hsv

REFERENCES:
https://arxiv.org/pdf/1709.04427.pdf
Contrast Enhancement of Brightness-Distorted Images by Improved Adaptive Gamma Correction
Gang Cao1*, Lihui Huang1, Huawei Tian2, Xianglin Huang3, Yongbin Wang1, Ruicong Zhi4

https://in.mathworks.com/matlabcentral/answers/uploaded_files/4242/AGCWD.pdf
Efficient Contrast Enhancement Using Adaptive Gamma Correction With Weighting Distribution
Shih-Chia Huang, Fan-Chieh Cheng, and Yi-Sheng Chiu
IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 22, NO. 3, MARCH 2013

NOTE: This script may be a bit slow due for non-HDRI compile Imagemagick to the use of -fx.

CAVEAT: No guarantee that this script will work on all platforms, nor that trapping of inconsistent parameters is complete and foolproof. Use At Your Own Risk.