org.geotools.coverage.processing.operation
Class GradientMagnitude
java.lang.Object
org.geotools.coverage.processing.AbstractOperation
org.geotools.coverage.processing.Operation2D
org.geotools.coverage.processing.OperationJAI
org.geotools.coverage.processing.operation.GradientMagnitude
- All Implemented Interfaces:
- org.opengis.coverage.processing.Operation, java.io.Serializable
- public class GradientMagnitude
- extends OperationJAI
Edge detector which computes the magnitude of the image gradient vector in two orthogonal
directions. The result of the operation may be defined as:
dst[x][y][b] = sqrt(
SH(x,y,b)2 +
SV(x,y,b)2)
× scale
where and are the horizontal and vertical gradient images
generated from band b of the source image by correlating it with the supplied
orthogonal (horizontal and vertical) gradient masks.
Before to compute the gradients, the kernels are tested against artificials horizontal and
vertical gradients of one unit of sample / unit of localization. For example:
- For an image of elevation (meters) using a geographic coordinate system (degrees
of latitude and longitude), the units are meters/degree.
- For an image of temperature (?C) using a projected coordinate system (kilometers),
the units are ?C/km.
Kernels are normalized by dividing all their coefficients by the result of this test. In other
words, kernels are normalized in such a way that applying the
operation on a horizontal or vertical gradient of 1 such "geophysical" units will give a result
of 1. This is an attempt to give geophysical meaning to the numbers produced by the
operation. This normalization depends of the coverage's
grid geometry.
NOTE: When the masks are symetric (e.g. Sobel, Prewitt (or Smoothed),
isotropic, etc.), then the above-cited algorithm produces the same result
than the "normalization factor" × "spatial factor" published by:
Simpson, J.J. (1990), "On the accurate detection and enhancement of oceanic features
observed in satellite data" in Remote sensing environment, 33:17-33.
However, for non-symetric masks (e.g. Kirsch), then a difference is found.
Name: "GradientMagnitude"
JAI operator: "GradientMagnitude"
Parameters:
| Name |
Class |
Default value |
Minimum value |
Maximum value |
|
GridCoverage2D |
N/A |
N/A |
N/A |
|
KernelJAI |
KernelJAI.GRADIENT_MASK_SOBEL_HORIZONTAL |
N/A |
N/A |
|
KernelJAI |
KernelJAI.GRADIENT_MASK_SOBEL_VERTICAL |
N/A |
N/A |
- Since:
- 2.2
- Version:
- $Id: GradientMagnitude.java 17672 2006-01-19 00:25:55Z desruisseaux $
- Author:
- Martin Desruisseaux
- See Also:
Operations.gradientMagnitude(org.opengis.coverage.Coverage),
GradientMagnitudeDescriptor,
Serialized Form
|
Constructor Summary |
GradientMagnitude()
Constructs a default gradient magnitude operation. |
| Methods inherited from class org.geotools.coverage.processing.AbstractOperation |
ensureNonNull, getDescription, getDocURL, getName, getNumSources, getParameters, getProcessor, getVendor, getVersion, hashCode, toString |
| Methods inherited from class java.lang.Object |
clone, finalize, getClass, notify, notifyAll, wait, wait, wait |
GradientMagnitude
public GradientMagnitude()
- Constructs a default gradient magnitude operation.
deriveGridCoverage
protected GridCoverage2D deriveGridCoverage(GridCoverage2D[] sources,
OperationJAI.Parameters parameters)
- Applies the operation on grid coverage. The default implementation looks for kernels
specified in the
parameter block and divide them by the distance between pixels, in the grid coverage's
coordinate reference system.
- Overrides:
deriveGridCoverage in class OperationJAI
- Parameters:
sources - The source coverages.parameters - Parameters, rendering hints and coordinate reference system to use.
- Returns:
- The result as a grid coverage.
- See Also:
OperationJAI.doOperation(org.opengis.parameter.ParameterValueGroup, org.geotools.factory.Hints),
OperationJAI.deriveSampleDimension(org.geotools.coverage.GridSampleDimension[][], org.geotools.coverage.processing.OperationJAI.Parameters),
JAI.createNS(java.lang.String, java.awt.image.renderable.ParameterBlock, java.awt.RenderingHints)
deriveCategory
protected Category deriveCategory(Category[] categories,
OperationJAI.Parameters parameters)
- Derives the quantitative category for a band in the destination image.
This implementation computes the expected gradient range from the two
masks and the value range in the source grid coverage.
- Overrides:
deriveCategory in class OperationJAI
- Parameters:
categories - The quantitative categories from every sources. For unary operations
like , this array has a length of 1. For binary
operations like and , this array has a length of 2.parameters - Parameters, rendering hints and coordinate reference system to use.
- Returns:
- The quantative category to use in the destination image, or if unknow.
deriveUnit
protected javax.units.Unit deriveUnit(javax.units.Unit[] units,
OperationJAI.Parameters parameters)
- Derives the unit of data for a band in the destination image.
This method compute / where:
- is the sample unit in source image.
- is the coordinate reference system axis unit.
- Overrides:
deriveUnit in class OperationJAI
- Parameters:
units - The units from every sources. For unary operations like
, this array has a length of 1. For binary operations
like and , this array has a length of 2.parameters - Parameters, rendering hints and coordinate reference system to use.
- Returns:
- The unit of data in the destination image, or if unknow.
Copyright © GeoTools. All Rights Reserved.