Image Expression Type

List of mathematical and logical operators that can be employed in the logic/algebraic expression.

In the syntax tables below, EXPR denotes any arbitray expression, IMAGE denotes any image identifier (a image identifier uses the syntax iX where X is an integer value from 1 to 100), TABLE denotes any table identifier (a table identifier uses the syntax tX where X is an integer value from 1 to 100) and VALUE denotes any value identifier (a value identifier uses the syntax vX where X is an integer value from 1 to 100).

Lookup table operators return the value corresponding to a given key according to a rule-operator.

They have the following syntax:
   TABLE[N ]
where:
   TABLE is a table identifier corresponding to a lookup table. Beware that, if the identifier is bound to a table, an error will be reported;
   N is the rule-operator.

Table operators return the value corresponding to a given set of keys.

They have the following syntax:
   TABLE[ [ ] [ ] ]
where:
   TABLE is a table identifier corresponding to a table. Beware that, if the identifier is bound to a lookup table, an error will be reported;

The expressions representing the set of keys must be separated by commas.

Neighborhood operators return the value of an operation within a defined neighborhood window.

They have the following syntax:
   nbN(IMAGE, h, w, y, x)

where:
   N is the operator name;
   IMAGE is an image identifier;
   h is the number of window lines;
   w is the number of window columns;
   y is the line where the window center is anchored at the image;
   x is the column where the window center is anchored at the image.

The calculation usually includes the center of the window.

Even-sided windows have the center displaced toward the top left corner.

Neighborhood (Ref) operators finds the value of an operation within a defined neighborhood window, this value is then used as an index to retrieve data either from a table or from a secondary image.

They have the following syntax:
   nbN(IMAGE,  h,  w, TABLE,  y,  x)
   nbN(IMAGE,  h,  w, IMAGE2,  y,  x)

where:
   N is the operator name;
   IMAGE is an Image identifier;
   h is the number of window lines;
   w is the number of window columns;
   TABLE is a LookupTable containing the values associated with each one of the Indices (Key = Index);
   IMAGE2 is an Image identifier whose cell values will be returned (Cell Coordinate = Index);
   y is the line where the window center is anchored at the image;
   x is the column where the window center is anchored at the image.

The calculation usually includes the center of the window. Even-sided windows have the center displaced toward the top left corner.

There is one additional image neighborhood operator that uses a slightly different syntax:
   nbConvol(IMAGE, TABLE, h, w, y, x)
where:
   IMAGE is an image identifier;
   TABLE is a table identifier;
   h is the number of window lines;
   w is the number of window columns;
   y is the line where the window center is anchored at the image;
   x is the column where the window center is anchored at the image.

nbConvol applies a convolution over IMAGE using the values stored on TABLE as the convolution coeficients. The table entries, from 1 to X, are mapped into the corresponding image values from top-down, left-to-right. The following examples ilustrated:

A image expression constant can be represented in a EGO Script using the same syntax used to represent the expression in the graphical interface editor, except that it must be surrounded by [ ] (square brackets):

The following example illustrates:

[
  if i1 <= (v3 - v1) / (v2 - v1) then
    v1 + sqrt(i1 * (v2 - v1) * (v3 - v1))
  else
    v2 - sqrt((1 - i1) * (v2 - v1) * (v2 - v3))
]