SegmentationMask<T>
Models & Types · AiDotNet.Augmentation.Image
Represents a segmentation mask for pixel-level annotations.
For Beginners
While a bounding box draws a rectangle around an object, a segmentation mask precisely outlines the object’s shape. When you rotate or flip an image, the mask must be transformed identically.
How It Works
Segmentation masks provide pixel-level object delineation, used for:
- Semantic segmentation (what class is each pixel)
- Instance segmentation (which object instance is each pixel)
- Panoptic segmentation (both semantic and instance)
Constructors
| Constructor | Summary |
|---|---|
SegmentationMask() |
Creates an empty segmentation mask. |
SegmentationMask(IList<IList<ValueTuple<double, double>>> polygons, int width, int height, MaskType type) |
Creates a segmentation mask from polygon vertices. |
SegmentationMask(T[] maskData, MaskType type, int classIndex) |
Creates a segmentation mask from dense data. |
SegmentationMask(int[] rleCounts, int width, int height, MaskType type) |
Creates a segmentation mask from polygon vertices. |
Properties
| Property | Summary |
|---|---|
IDictionary<string, object> Metadata { get; set; } |
Gets or sets additional metadata. |
IList<IList<ValueTuple<double, double>>> Polygons { get; set; } |
Gets or sets the polygon vertices (if using polygon encoding). |
MaskEncoding Encoding { get; set; } |
Gets or sets the current encoding format. |
MaskType Type { get; set; } |
Gets or sets the mask type. |
T Confidence { get; set; } |
Gets or sets the confidence score (if from a segmenter). |
T[] MaskData { get; set; } |
Gets or sets the mask data as a 2D array [height, width]. |
int ClassIndex { get; set; } |
Gets or sets the class index (for semantic/instance masks). |
int Height { get; set; } |
Gets or sets the mask height. |
int InstanceId { get; set; } |
Gets or sets the instance ID (for instance/panoptic masks). |
int Width { get; set; } |
Gets or sets the mask width. |
int[] RLECounts { get; set; } |
Gets or sets the RLE-encoded mask (if using RLE encoding). |
string ClassName { get; set; } |
Gets or sets the class name. |
Methods
| Method | Summary |
|---|---|
SegmentationMask<T> Clone() |
Creates a deep copy of this mask. |
T[] ToDense() |
Converts this mask to dense format. |
ValueTuple<int, int, int, int> GetBoundingBox() |
Calculates the bounding box of the mask. |
double Dice(SegmentationMask<T> other) |
Calculates the Dice coefficient with another mask. |
double IoU(SegmentationMask<T> other) |
Calculates the IoU (Intersection over Union) with another mask. |
int Area() |
Calculates the area (number of foreground pixels). |
int[] ToRLE() |
Converts this mask to RLE format. |