Adobe Photoshop CS4 - Manual (Page 625)

618 USING PHOTOSHOP CS4
3D and technical imaging
Last updated 1/10/2010
• For removing objects from the image, use the Median plug-in.
The output is a composite image the same size as the original image stack. You may need to experiment with
different plug-ins to get the best enhancement for a particular image.
To change the rendering effect, choose a different Stack Mode from the submenu. Stack rendering is not
cumulative—each render effect operates on the original image data in the stack and replaces previous effects.
Stack modes
Stack modes operate on a per-channel basis only, and only on non-transparent pixels. For example, the Maximum
mode returns the maximum red, green, and blue channel values for a pixel cross section and merges them into one
composite pixel value in the rendered image.
Remove stack rendering
❖ Choose Layers > Smart Objects > Stack Mode > None to remove any rendering from an image stack and convert it
back to a regular Smart Object.
Rendering plug-in name Result Comments
Entropy entropy = - sum( (probability of value) * log2(
probability of value) )
Probability of value = (number of occurrences of
value) / (total number of non-transparent pixels)
The binary entropy (or zero order entropy) defines
a lower bound on how many bits would be
necessary to losslessly encode the information in a
set.
Kurtosis kurtosis = ( sum( (value - mean) 4 ) over non-
transparent pixels ) / ( ( number of non-transparent
pixels - 1 ) * (standard deviation)
4 ).
A measure of peakedness or flatness compared to
a normal distribution. The kurtosis for a standard
normal distribution is 3.0. Kurtosis greater than 3
indicates a peaked distribution, and kurtosis less
than 3 indicates a flat distribution (compared to a
normal distribution).
Maximum The maximum channel values for all non-
transparent pixels
Mean The mean channel values for all non-transparent
pixels Effective for noise reduction
Median The median channel values for all non-transparent
pixels Effective for noise reduction and removal of
unwanted content from the image
Minimum The minimum channel values for all non-
transparent pixels
Range Maximum minus the minimum of the non-
transparent pixel values
Skewness skewness = (sum( (value - mean) 3 ) over non-
transparent pixels ) / ( ( number of non-transparent
pixels - 1 ) * (standard deviation)
3 )
Skewness is a measure of symmetry or asymmetry
around the statistical mean
Standard Deviation standard deviation = Square Root(variance)
Summation The sum channel values for all non-transparent
pixels
Variance variance = (sum( (value-mean) 2 ) over non-
transparent pixels ) / ( number of non-transparent
pixels - 1)