interpolate_scattered_data_image🔗
Short description🔗
interpolate_scattered_data_image — Interpolation of an image.
Signature🔗
interpolate_scattered_data_image( image Image, region RegionInterpolate, out image ImageInterpolated, string Method, attribute.name GenParamName, attribute.value GenParamValue )
Description🔗
interpolate_scattered_data_image interpolates the Image
at the region RegionInterpolate and returns the result in
ImageInterpolated. The difference of the domain of the
Image and the region RegionInterpolate specifies the valid
data points that can be used for the interpolation whereas
RegionInterpolate specifies the points, where the
gray values of the Image should be determined.
With the parameter Method the interpolation algorithm
is specified. So far, only the
'thin_plate_splines' are supported. This method interpolates on
a global scale, which means that all points are regarded for the
interpolation, no matter how far away they are. The influence of far points
is correlated to \(r^2 \cdot \log(r)\) where \(r\) defines
the distance of two points. If the same Image is interpolated at
different points in subsequent steps, the operator
create_scattered_data_interpolator may be more efficient.
The following parameters can be adjusted with GenParamName and
GenParamValue:
-
'alpha': The parameter 'alpha' is a smoothing factor. For 'alpha' \(=\) 0, all points in the image
Imageare interpolated exactly. With 'alpha' getting larger, the interpolation smoothes the image points in way that all interpolated points of the result imageImageInterpolatedlie on a common plane.Default: is 0
Restriction: 'alpha' \(>=\) 0
interpolate_scattered_data_image is best used with very few data
points, e.g, less than 1000. In order to reconstruct
destroyed image data in the region RegionInterpolate, the data
points should be reduced to an appropriate number, e.g., by only using the
border pixels of the hole regions. An example program is shown below.
Execution information🔗
Execution information
-
Multithreading type: reentrant (runs in parallel with non-exclusive operators).
-
Multithreading scope: global (may be called from any thread).
-
Processed without parallelization.
Parameters🔗
Image (input_object) singlechannelimage → object (byte / uint2 / real)
Image to interpolate
RegionInterpolate (input_object) region → object
Region to interpolate
ImageInterpolated (output_object) singlechannelimage → object (real)
Interpolated image
Method (input_control) string → (string)
Method for the interpolation
Default: 'thin_plate_splines'
Suggested values: 'thin_plate_splines'
GenParamName (input_control) attribute.name-array → (string)
Names of the generic parameters that can be adjusted
Default: []
Suggested values: 'alpha'
GenParamValue (input_control) attribute.value-array → (string / integer / real)
Values of the generic parameters that can be adjusted
Default: []
Suggested values: 0, 1.0, 10.0, 100.0
Example🔗
(HDevelop)
* This example program shows how to use the scattered data interpolator
* to fill holes in an image
gen_image_surface_second_order (ImageData, 'real', 1, 1, 0, 0, 0, 1, \
24, 24, 48, 48)
gen_circle (Circle, 12, 12, 6)
difference (ImageData, Circle, Region)
reduce_domain (ImageData, Region, ImageReduced)
dev_clear_window ()
dev_display (ImageReduced)
stop()
*
* Select only border pixels for the interpolation
dilation_circle (Circle, CircleDilation, 1.5)
intersection (CircleDilation, Region, RegionBorderData)
dev_clear_window ()
dev_display (ImageReduced)
dev_display (RegionBorderData)
stop()
*
* Interpolate pixels
reduce_domain (ImageData, RegionBorderData, ImageReducedBorder)
interpolate_scattered_data_image (ImageReducedBorder, Circle, \
ImageInterpolated, \
'thin_plate_splines', [], [])
paint_gray (ImageInterpolated, ImageData, ImageFilled)
dev_clear_window ()
dev_display (ImageFilled)
Result🔗
If the parameters are valid, the operator
interpolate_scattered_data_image returns the value 2 (H_MSG_TRUE).
If necessary, an exception is raised.
Combinations with other operators🔗
Module🔗
Foundation