boofcv.alg.flow

Class UtilDenseOpticalFlow

java.lang.Object

boofcv.alg.flow.UtilDenseOpticalFlow

public class UtilDenseOpticalFlow
extends java.lang.Object

Useful functions when computing dense optical flow

Constructor Summary

Constructors

Constructor and Description
UtilDenseOpticalFlow()

Method Summary

Modifier and Type	Method and Description
static <T extends ImageGray> PyramidFloat<T>	standardPyramid(int width, int height, double scale, double sigma, int minSize, int maxLayers, java.lang.Class<T> imageType) Create a standard image pyramid used by dense optical flow parameters.

Methods inherited from class java.lang.Object

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

Constructor Detail

UtilDenseOpticalFlow

public UtilDenseOpticalFlow()

Method Detail

standardPyramid

public static <T extends ImageGray> PyramidFloat<T> standardPyramid(int width,
                                                                    int height,
                                                                    double scale,
                                                                    double sigma,
                                                                    int minSize,
                                                                    int maxLayers,
                                                                    java.lang.Class<T> imageType)

Create a standard image pyramid used by dense optical flow parameters. The first layer is the size of the input image and the last layer is ≥ the minSize. The sigma for each layer is computed using the following formula:

sigmaLayer = sigma*sqrt( scale^-2 - 1 )

If the scale is 1 then a single layer pyramid will be created. If the scale is 0 then the scale will be determined by the maxLayers parameter.

Type Parameters:

T - Image type

Parameters:

width - Width of input image.

height - Height of input image.

scale - Scale between layers. 0 ≤ scale ≤ 1. Try 0.7

sigma - Adjusts the amount of blur applied to each layer. If sigma ≤ 0 then no blur is applied.

minSize - The minimum desired image size in the pyramid

maxLayers - The maximum number of layers in the pyramid.

imageType - Type of image for each layer

Returns:

The image pyramid.