PageSegmentationResult<T>

Models & Types · AiDotNet.Document

Represents the result of page segmentation.

For Beginners

Page segmentation divides a document page into different types of regions (paragraphs, figures, tables, etc.). This result contains all detected regions with their types and locations.

Constructors

Constructor Summary
PageSegmentationResult()

Properties

Property Summary
IReadOnlyList<DocumentRegion<T>> Regions { get; set; } Gets the detected document regions.
IReadOnlyList<int> ReadingOrder { get; set; } Gets the reading order of text regions (indices into Regions list).
Tensor<T> ClassProbabilities { get; set; } Gets the class probabilities for each pixel (shape: [height, width, num_classes]).
Tensor<T> SegmentationMask { get; set; } Gets the pixel-level segmentation mask (class index per pixel).
double ProcessingTimeMs { get; set; } Gets the processing time in milliseconds.
int RegionCount { get; } Gets the total number of detected regions.