Class OpticalFlow

Namespace

Bonsai.Vision

Assembly

Bonsai.Vision.dll

Represents an operator that computes dense optical flow using Gunnar Farneback’s algorithm across all images in the sequence.

public class OpticalFlow : Transform<Tuple<IplImage, IplImage>, IplImage>

Inheritance

object

Combinator<Tuple<IplImage, IplImage>, IplImage>

Transform<Tuple<IplImage, IplImage>, IplImage>

OpticalFlow

Inherited Members

Combinator<Tuple<IplImage, IplImage>, IplImage>.Process(IObservable<Tuple<IplImage, IplImage>>)

object.ToString()

object.Equals(object)

object.Equals(object, object)

object.ReferenceEquals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

Properties

Flags

Gets or sets a value specifying the operation flags for the optical flow algorithm.

public FarnebackFlowFlags Flags { get; set; }

Property Value

FarnebackFlowFlags

Iterations

Gets or sets the number of iterations of the algorithm at each pyramid level.

[Range(0, 2147483647)]
public int Iterations { get; set; }

Property Value

int

Levels

Gets or sets the number of pyramid layers, including the initial image.

[Range(1, 2147483647)]
public int Levels { get; set; }

Property Value

int

PolyN

Gets or sets the size of the pixel neighborhood used to find polynomial expansion in each pixel.

[Range(0, 2147483647)]
public int PolyN { get; set; }

Property Value

int

PolySigma

Gets or sets the standard deviation of the Gaussian used to smooth the derivatives used as a basis for the polynomial expansion.

[Precision(2, 0.01)]
public double PolySigma { get; set; }

Property Value

double

PyramidScale

Gets or sets a value specifying the image scale (less than 1) used to build the pyramids for each image.

[Range(0, 1)]
public double PyramidScale { get; set; }

Property Value

double

WindowSize

Gets or sets the averaging window size. Larger values increase robustness to noise and fast motion, but yield a blurred motion field.

[Range(0, 2147483647)]
public int WindowSize { get; set; }

Property Value

int

Methods

Process(IObservable<IplImage>)

Computes dense optical flow using Gunnar Farneback’s algorithm across all images in an observable sequence.

public IObservable<IplImage> Process(IObservable<IplImage> source)

Parameters

source IObservable<IplImage>

The sequence of images across which to compute dense optical flow.

Returns

IObservable<IplImage>

A sequence of IplImage objects representing the flow field between the current and the previous image in the source sequence.

Process(IObservable<Tuple<IplImage, IplImage>>)

Computes dense optical flow using Gunnar Farneback’s algorithm across all image pairs in an observable sequence.

public override IObservable<IplImage> Process(IObservable<Tuple<IplImage, IplImage>> source)

Parameters

source IObservable<Tuple<IplImage, IplImage>>

The sequence of image pairs across which to compute the dense optical flow. The optical flow is computed from the first to the second image in the pair.

Returns

IObservable<IplImage>

A sequence of IplImage objects representing the flow field between the first and the second image of each pair in the source sequence.