Class YoloDetector

Namespace

VisioForge.Core.AI

Assembly

VisioForge.Core.AI.dll

Object detector built on top of VisioForge.Core.AI.OnnxInferenceEngine that supports multiple model families (YOLOv8/v11, YOLOX, RT-DETR / D-FINE) selected via VisioForge.Core.Types.X.AI.YoloDetectorSettings.Model.

public class YoloDetector : IDisposable

Inheritance

object

YoloDetector

Implements

IDisposable

Inherited Members

object.Equals(object?)

object.Equals(object?, object?)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object?, object?)

object.ToString()

Remarks

On construction it asks the family-specific decoder for the right preprocessing options (cached for the detector's lifetime). Then on every frame it letterboxes/resizes the frame accordingly via the engine, runs the model, hands the raw outputs to the decoder, and optionally draws the resulting boxes onto the RGBA frame. The underlying engine is lazily initialized on the first VisioForge.Core.AI.YoloDetector.Process(VisioForge.Core.Types.X.VideoFrameX) call (before the first frame is preprocessed, so the model's real input size is authoritative). The class name is retained for API compatibility although it now drives non-YOLO models too.

Constructors

YoloDetector(YoloDetectorSettings)

Initializes a new instance of the VisioForge.Core.AI.YoloDetector class.

public YoloDetector(YoloDetectorSettings settings)

Parameters

settings YoloDetectorSettings

The detector settings.

Exceptions

ArgumentNullException

Thrown when settings is null.

Properties

ActiveProvider

Gets the execution provider the underlying inference session actually engaged (valid after VisioForge.Core.AI.YoloDetector.Init).

public OnnxExecutionProvider ActiveProvider { get; }

Property Value

OnnxExecutionProvider

Methods

Dispose()

Releases the inference engine.

public void Dispose()

Dispose(bool)

Releases unmanaged and - optionally - managed resources.

protected virtual void Dispose(bool disposing)

Parameters

disposing bool

true to release both managed and unmanaged resources; false to release only unmanaged resources.

Init()

Loads the model and prepares the detector for inference.

public bool Init()

Returns

bool

true if initialization succeeded; otherwise, false.

Process(VideoFrameX)

Runs detection on a single RGBA frame. The engine is lazily initialized on first use. If that initialization fails (missing/corrupt model, provider error) the failure is latched and detection is disabled for this detector: subsequent frames return empty without retrying, until a later explicit VisioForge.Core.AI.YoloDetector.Init call succeeds (or the detector is recreated). See _initFailed.

public OnnxDetection[] Process(VideoFrameX frame)

Parameters

frame VideoFrameX

The source RGBA frame.

Returns

OnnxDetection[]

The detected objects, in source-frame coordinates; empty when nothing was found or when initialization has failed.

SetContext(BaseContext)

Sets the processing context used for logging and propagates it to the inference engine.

public void SetContext(BaseContext context)

Parameters

context BaseContext

The context.