ObjectDetectionResult

[androidJvm]
class ObjectDetectionResult(val awarenessParams: AwarenessImageParams, val numDetections: Int, val numClasses: Int, val boundingBoxLocations: FloatArray?, val probabilities: FloatArray?, val trackingIds: IntArray?, val imageParams: ObjectDetectionImageParams)

Wrapper for the native object detection result buffer.

Constructors


ObjectDetectionResult	[androidJvm] constructor(awarenessParams: AwarenessImageParams, numDetections: Int, numClasses: Int, boundingBoxLocations: FloatArray?, probabilities: FloatArray?, trackingIds: IntArray?, imageParams: ObjectDetectionImageParams)

Name	Summary
awarenessParams	[androidJvm] val awarenessParams: AwarenessImageParams Inference meta-data.
boundingBoxLocations	[androidJvm] val boundingBoxLocations: FloatArray? Corner points of the bounding box. Length: numDetections * 4
imageParams	[androidJvm] val imageParams: ObjectDetectionImageParams Dimensions of the source image and models input dimensions.
numClasses	[androidJvm] val numClasses: Int The number of possible classifications per object.
numDetections	[androidJvm] val numDetections: Int The number of detected objects.
probabilities	[androidJvm] val probabilities: FloatArray? The probabilities (detection confidence) buffer. Length: numDetections * numClasses
trackingIds	[androidJvm] val trackingIds: IntArray? Optional Ids for tracking the same object through inference frames. Length = numDetections

Name	Summary
toDetectedObjects	[androidJvm] fun ObjectDetectionResult.toDetectedObjects(confidenceThreshold: Float = 0.4f): List<DetectedObject> Converts this ObjectDetectionResult into a list of strongly typed DetectedObject instances.