ILayoutDetector<T>
Interfaces · AiDotNet.Document.Interfaces
Interface for document layout detection models.
Implements: IDocumentModel<T>
Known implementations: DiT<T>, DocFormer<T>, DocGCN<T>, DocOwl<T>, LayoutGraph<T>, LayoutLM<T>, LayoutLMv2<T>, LayoutLMv3<T>, LayoutXLM<T>, LiLT<T>, UDOP<T>
For Beginners
Think of layout detection as drawing boxes around different parts of a document and labeling what each part is (title, paragraph, table, etc.). This helps computers understand the structure of a document just like humans do.
Example usage:
How It Works
Layout detection identifies and localizes different structural elements in a document, such as text blocks, tables, figures, headers, and footers.
Properties
| Property | Summary |
|---|---|
IReadOnlyList<LayoutElementType> SupportedElementTypes { get; } |
Gets the layout element types this detector can identify. |
Methods
| Method | Summary |
|---|---|
DocumentLayoutResult<T> DetectLayout(Tensor<T> documentImage) |
Detects layout regions in a document image. |
DocumentLayoutResult<T> DetectLayout(Tensor<T> documentImage, double confidenceThreshold) |
Detects layout regions with a specified confidence threshold. |