Models an image object detection represented by a label, a geometry, and optionally a confidence value.

interface AnnotationsObjectDetection {
    attributes?: Record<string, AnnotationsBoolean | AnnotationsNumerical>;
    boundingBox?: AnnotationsBoundingBox;
    confidence?: number;
    label: string;
    polygon?: AnnotationsPolygon;
    polyline?: AnnotationsPolyLine;
}

Properties

attributes?: Record<string, AnnotationsBoolean | AnnotationsNumerical>

Additional attributes data for a compound.

A plain rectangle

confidence?: number

The confidence score for the primitive. It should be between 0 and 1.

0

1

label: string

The label describing what type of object it is

A closed polygon represented by n vertices. In other words, we assume that the first and last vertex are connected.

A polygonal chain consisting of n vertices