Table of Contents

Class VLMRegion

Namespace
VisioForge.Core.Types.X.AI
Assembly
VisioForge.Core.dll

A single grounded region produced by a Florence-2 task: a text label (an object category, a region description, or a recognized text block) together with its bounding box in source-frame pixel coordinates.

public class VLMRegion

Inheritance

Inherited Members

Remarks

Regions are only produced by the grounding tasks (VisioForge.Core.Types.X.AI.VLMTask.ObjectDetection, VisioForge.Core.Types.X.AI.VLMTask.DenseRegionCaption, VisioForge.Core.Types.X.AI.VLMTask.OcrWithRegion, VisioForge.Core.Types.X.AI.VLMTask.PhraseGrounding). Caption-only tasks return an empty region set and carry their result in VisioForge.Core.Types.Events.VLMResultGeneratedEventArgs.Text.

Constructors

VLMRegion()

Initializes a new instance of the VisioForge.Core.Types.X.AI.VLMRegion class.

public VLMRegion()

VLMRegion(string, Rect)

Initializes a new instance of the VisioForge.Core.Types.X.AI.VLMRegion class.

public VLMRegion(string label, Rect box)

Parameters

label string

The region label.

box Rect

The region bounding box in source-frame pixels.

Properties

Box

Gets or sets the region bounding box in source-frame pixel coordinates. For quadrilateral OCR regions this is the axis-aligned bounding rectangle of the quad.

public Rect Box { get; set; }

Property Value

Rect

Label

Gets or sets the region label: the object category, region description, or recognized text.

public string Label { get; set; }

Property Value

string