Class DISK

Namespace

OpenCvSharp

Assembly

OpenCvSharp.dll

DISK: deep learning based local feature detector and descriptor (OpenCV 5). Runs an ONNX model through the DNN module. Requires OpenCV built with the dnn module.

public class DISK : Feature2D, IDisposable

Inheritance

object

CvObject

CvPtrObject

Algorithm

Feature2D

DISK

Implements

IDisposable

Inherited Members

Feature2D.DescriptorSize

Feature2D.DescriptorType

Feature2D.DefaultNorm

Feature2D.Empty()

Feature2D.Detect(Mat, Mat)

Feature2D.Detect(InputArray, InputArray)

Feature2D.Detect(IEnumerable<Mat>, IEnumerable<Mat>)

Feature2D.Compute(InputArray, ref KeyPoint[], OutputArray)

Feature2D.Compute(IEnumerable<Mat>, ref KeyPoint[][], IEnumerable<Mat>)

Feature2D.DetectAndCompute(InputArray, InputArray, out KeyPoint[], OutputArray, bool)

Feature2D.Write(string)

Feature2D.Read(string)

Feature2D.GetDefaultName()

Algorithm.Write(FileStorage)

Algorithm.Read(FileNode)

Algorithm.Save(string)

CvPtrObject.RawPtr

CvObject.IsDisposed

CvObject.ptr

CvObject.SetSafeHandle(OpenCvSafeHandle)

CvObject.Dispose()

CvObject.Dispose(bool)

CvObject.DisposeManaged()

CvObject.DisposeUnmanaged()

CvObject.ThrowIfDisposed()

CvObject.CvPtr

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Properties

ImageSize

Target input size fed to the network.

public Size ImageSize { get; set; }

Property Value

Size

MaxKeypoints

Maximum number of keypoints to return per image; -1 keeps all detections.

public int MaxKeypoints { get; set; }

Property Value

int

ScoreThreshold

Keypoints with network score strictly below this value are discarded.

public float ScoreThreshold { get; set; }

Property Value

float

Methods

Create(string, int, float, Size, int, int)

Creates a DISK detector from a model file path.

public static DISK Create(string modelPath, int maxKeypoints = -1, float scoreThreshold = 0, Size imageSize = default, int backendId = 0, int targetId = 0)

Parameters

modelPath string

Path to the DISK ONNX model.

maxKeypoints int

Maximum number of keypoints to return per image; -1 keeps all detections.

scoreThreshold float

Discard keypoints with network score strictly below this value.

imageSize Size

Target input size fed to the network. Default (empty) falls back to the network's fixed 1024x1024 input. When overriding, both dimensions must be positive multiples of 16.

backendId int

DNN backend identifier (see Backend); 0 = default.

targetId int

DNN target identifier (see Target); 0 = CPU.

Returns

DISK

CreateFromMemory(byte[], int, float, Size, int, int)

Creates a DISK detector from an in-memory ONNX model buffer.

public static DISK CreateFromMemory(byte[] bufferModel, int maxKeypoints = -1, float scoreThreshold = 0, Size imageSize = default, int backendId = 0, int targetId = 0)

Parameters

bufferModel byte[]

A buffer containing the contents of the DISK ONNX model.

maxKeypoints int

Maximum number of keypoints to return per image; -1 keeps all detections.

scoreThreshold float

Discard keypoints with network score strictly below this value.

imageSize Size

Target input size fed to the network. Default (empty) falls back to the network's fixed 1024x1024 input. When overriding, both dimensions must be positive multiples of 16.

backendId int

DNN backend identifier (see Backend); 0 = default.

targetId int

DNN target identifier (see Target); 0 = CPU.

Returns

DISK