Image operations

These nodes perform image operations – typically arithmetic – on whole frames (or volumes in case of 3D) on a pixel-by-pixel (or voxel-by-voxel in later case) basis.

Most of the operations produce a floating point result, except for a few nodes that are guaranteed to fall within the range of the original bit depth (such as Minimum, Maximum, …).

Arithmetics

Add

Add two images (channels) element-wise (A+B)

Subtract

Subtract two images (channels) element-wise (A-B).

See also: Difference of subsequent frames (in workflows)

Multiply

Multiply two images (channels) element-wise (A*B)

Divide

Divide two images (channels) element-wise (A/B)

Average

Computes element-wise average of all connected channels

Minimum

Computes element-wise minimum of all connected channels

Maximum

Computes element-wise maximum of all connected channels

Covariance

Calculate element-wise intermediate term for Covariance (cov(A, B)). E[cov(A, B)] produces the covariance.

Calculating Mean value of $R_i$ ($E[R_i]$) produces the Covariance (see wikipedia)

Mask Image

Fills all elements outside of masked area with zero

Remove Shading

Applies shading correction image (input B) to image (input A) in order to remove & correct shading

Constant

Add

Add a constant to every element of the image (A + Offset)

Subtract

Subtract a constant from every element of the image (A - Offset)

Multiply

Multiply every element of the image by a constant (A * Gain)

Divide

Divide every element of the image by a constant (A / Divisor)

Clip

Clip every element of the image to fall between Min and Max

Constant Value

Set every element of the image to constant value (A = Constant)

Functions

Abs

Compute element-wise absolute value (abs(A))

Invert

Compute element-wise complement (maximum value - A).

When the image element is

• N-bit unsigned integer the result is a complement:

  $(2^N-1)-A_i$, where $A_i$ is the $i$-th element

• floating point the result is negative value:

  $-A_i$, where $A_i$ is the $i$-th element

Exp

Compute element-wise decadic exponent function (exp₁₀(A))

Inv

Compute element-wise multiplicative inverse function (1 / A)

Log

Compute element-wise decadic logarithm function (log₁₀(A))

Pow (x^y)

Compute element-wise power function. Raise A to y where Y is a scalar constant (pow(A, y))

Sign

Compute element-wise signum function (sign(A))

Logical

These operations produce values of one (1) for elements where the logical operation holds true and zero (0) where it is false.

Greater

Compare two channels using the greater-than operator (A > B)

See also: Logical (group)

Greater or Equal

Compare two channels using the greater-than-or-equal-to operator (A ≥ B)

See also: Logical (group)

Equal

Compare two channels using equality operator (A = B)

See also: Logical (group)

Not Equal

Compare two channels using inequality operator (A ≠ B)

See also: Logical (group)

Merge

Blend

Merge two channels (input A) and (input B) using weights in third channel (input C).

Conditional

Merge two channels (input A) and (input B) based on trueness of third channel (input C).