Operator Reference
expand_domain_gray (Operator)
expand_domain_gray
— Expand the domain of an image and set the gray values in the
expanded domain.
Signature
expand_domain_gray(InputImage : ExpandedImage : ExpansionRange : )
Description
expand_domain_gray
expands the border gray values of the
domain outwards. The width of the expansion is set by the parameter
ExpansionRange
, which describes the expansion radius in pixels.
All filters in HALCON which are applied to the image domain also include
the gray values of a certain number of pixels, depending on the filter size,
outside of the image domain in their calculation.
This may lead to undesirable side effects especially in the border region
of the domain. For example, if the foreground (domain) and the
background of the image differ strongly in brightness, the result of
a filter operation may lead to undesired darkening or brightening at
the border of the domain. In order to avoid this drawback, the
domain is artificially expanded by expand_domain_gray
in a preliminary
stage, copying the gray values of the border pixels to the outside
of the domain. In addition, the domain itself is also expanded to
reflect the newly set pixels. Therefore, in many cases it is
reasonable to reduce the domain again (reduce_domain
or
change_domain
) after using expand_domain_gray
and
call the filter operation afterwards.
ExpansionRange
should be set to a value high enough to
make sure the whole region including its dilatation through the filter
mask has initialized values.
Usually this leads to ExpansionRange
values of half the filter
length or greater.
In combination with the filter mean_image
on images of type real we recommend to determine the value for
ExpansionRange
as shown in the example below.
For speed reasons this filter uses the values from pixels within the
smallest rectangle around the dilated region.
Attention
Note that filter operators may return unexpected results if an image with a reduced domain is used as input. Please refer to the chapter Filters.
Execution Information
- Multithreading type: reentrant (runs in parallel with non-exclusive operators).
- Multithreading scope: global (may be called from any thread).
- Automatically parallelized on tuple level.
Parameters
InputImage
(input_object) (multichannel-)image(-array) →
object (byte / int1 / int2 / uint2 / int4 / real)
Input image with domain to be expanded.
ExpandedImage
(output_object) image(-array) →
object (byte / int1 / int2 / uint2 / int4 / real)
Output image with new gray values in the expanded domain.
ExpansionRange
(input_control) integer →
(integer)
Radius of the gray value expansion, measured in pixels.
Default: 2
Suggested values: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16
Restriction:
ExpansionRange >= 1
Example (HDevelop)
read_image(Fabrik, 'fabrik') gen_rectangle2(Rectangle_Label,243,320,-1.55,62,28) reduce_domain(Fabrik, Rectangle_Label, Fabrik_Label) * Character extraction without gray value expansion: MaskSize := 31 mean_image(Fabrik_Label,Label_Mean_normal,MaskSize, MaskSize) dyn_threshold(Fabrik_Label,Label_Mean_normal,Characters_normal,10,'dark') dev_display(Fabrik) dev_display(Characters_normal) * The characters in the border region are not extracted ! stop() * Character extraction with gray value expansion: get_domain (Fabrik_Label, Domain) expand_domain_gray (Fabrik_Label, Expanded_Fabrik_Label, MaskSize) get_domain (Expanded_Fabrik_Label, Expanded_Domain) shape_trans (Expanded_Domain, RegionRect, 'rectangle1') difference (RegionRect, Expanded_Domain, Region_Difference) overpaint_region (Expanded_Fabrik_Label, Region_Difference, 0, 'fill') reduce_domain (Expanded_Fabrik_Label, Domain, Reduced_Expanded_Fabrik_Label) mean_image (Reduced_Expanded_Fabrik_Label, Label_Mean_Expanded,\ MaskSize, MaskSize) dyn_threshold(Fabrik_Label,Label_Mean_Expanded,\ Characters_expanded,10,'dark') dev_display(Fabrik) dev_display(Characters_expanded)
Complexity
Let L the perimeter of the domain. Then the runtime complexity is
approximately O(L)*ExpansionRange
.
Result
expand_domain_gray
returns 2 (
H_MSG_TRUE)
if all parameters are
correct. If necessary, an exception is raised.
Possible Predecessors
Possible Successors
reduce_domain
,
mean_image
,
dyn_threshold
See also
Module
Foundation