ColorSeparate
Tanh
Mean
ImageAdjust
ColorToneMapping
BrightnessEqualize
ImageContainsQ
ImageContents
ImageEffect
ImageGraphics
ImageRecolor
ImageRestyle

Image Processing

Tone Mapping Using Simple Image Arithmetic

Prototype a tone-mapping algorithm to make dark pixels of an image visible in the standard dynamic range

Open in Cloud
Download Notebook

This forest image contains more information in its dark pixels than is visible in the standard dynamic range:

In[1]:=

image=

;

The extra data can be made visible with a simple multiplication:

In[2]:=

2^Range[0,3]image

Out[2]=

A nonlinear mapping operator can be directly applied on the image to rescale the brightness:

In[3]:=

image

image+

//ImageAdjust

Out[3]=

An alternative way to compress the tones is to use the transcendental

Log

function:

In[4]:=

Logimage+

//ImageAdjust

Out[4]=

Combining arithmetic and color conversion, a luminance-only scaling is performed:

In[5]:=

luminance=ColorSeparate[image,"L"];

image

luminance

ImageAdjustLogluminance+

//ImageAdjust

Out[5]=

This gives the possibility of finer control of the image saturation:

In[6]:=

saturation={0,1/2,1,2};

saturation

image

luminance

ImageAdjustLogluminance+



Out[6]=

Finally, arithmetic and statistics can be combined to build a nonlinear mean of the different brightness levels:

In[7]:=

MeanTanhimage

Range[-3,3,.4]

//ImageAdjust

Out[7]=

Compare the algorithm with some built-in global and local tone-mapping algorithms:

In[8]:=

ColorToneMapping[image,Method#]&/@{"DodgingAndBurning","Luminance"}

Out[8]=

Related Symbols

ColorSeparate
Tanh
Mean
ImageAdjust
ColorToneMapping
BrightnessEqualize
ImageContainsQ
ImageContents
ImageEffect
ImageGraphics
ImageRecolor
ImageRestyle

Publisher Information

Contributed by: Wolfram Staff