boofcv.alg.feature.orientation.impl

Class ImplOrientationSlidingWindowIntegral<T extends ImageGray,G extends GradientValue>

java.lang.Object

boofcv.alg.feature.orientation.OrientationIntegralBase<T,G>

boofcv.alg.feature.orientation.impl.ImplOrientationSlidingWindowIntegral<T,G>

All Implemented Interfaces:

OrientationIntegral<T>, RegionOrientation

public class ImplOrientationSlidingWindowIntegral<T extends ImageGray,G extends GradientValue>
extends OrientationIntegralBase<T,G>

Implementation of OrientationSlidingWindow for integral images. TODO comment how this implementation works

Constructor Summary

Constructors

Constructor and Description
ImplOrientationSlidingWindowIntegral(double radiusToScale, double samplePeriod, double windowSize, int sampleRadius, double weightSigma, int sampleKernelWidth, java.lang.Class<T> integralType) Specifies configuration parameters and initializes data structures

Method Summary

Modifier and Type	Method and Description
double	compute(double c_x, double c_y) Computes the orientation of a region about its center.

Methods inherited from class boofcv.alg.feature.orientation.OrientationIntegralBase

getImageType, setImage, setObjectRadius

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ImplOrientationSlidingWindowIntegral

public ImplOrientationSlidingWindowIntegral(double radiusToScale,
                                            double samplePeriod,
                                            double windowSize,
                                            int sampleRadius,
                                            double weightSigma,
                                            int sampleKernelWidth,
                                            java.lang.Class<T> integralType)

Specifies configuration parameters and initializes data structures

Parameters:

samplePeriod - How often the image is sampled. This number is scaled. Typically 1.

windowSize - Angular window that is slide across

sampleRadius - Radius of the region being considered in terms of samples. Typically 6.

weightSigma - Sigma for weighting distribution. Zero for unweighted.

sampleKernelWidth - Size of kernel doing the sampling. Typically 4.

integralType - Type of integral image being processed.

Method Detail

compute

public double compute(double c_x,
                      double c_y)

Description copied from interface: RegionOrientation

Computes the orientation of a region about its center.

Parameters:

c_x - Center of the region in image pixels.

c_y - Center of the region in image pixels.

Returns:

Orientation in radians. Angle zero points along x-axis and pi/2 along y-axis.