Package org.photonvision.jni

Class RknnObjectDetector

java.lang.Object

org.photonvision.jni.RknnObjectDetector

All Implemented Interfaces:

ObjectDetector, Releasable

public class RknnObjectDetector
extends Object
implements ObjectDetector

Manages an object detector using the rknn backend.

Constructor Summary

Constructors

Constructor	Description
RknnObjectDetector(RknnModel model, org.opencv.core.Size inputSize)	Creates a new RknnObjectDetector from the given model.

Method Summary

Modifier and Type	Method	Description
List<NeuralNetworkPipeResult>	detect(org.opencv.core.Mat in, double nmsThresh, double boxThresh)	Detects objects in the given input image using the RknnDetector.
List<String>	getClasses()	Returns the classes that the detector can detect
RknnModel	getModel()	Returns the model in use by this detector.
void	release()	Thread-safe method to release the detector.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

RknnObjectDetector

public RknnObjectDetector(RknnModel model,
 org.opencv.core.Size inputSize)

Creates a new RknnObjectDetector from the given model.

Parameters:

model - The model to create the detector from.

inputSize - The required image dimensions for the model. Images will be Letterboxed to this shape.

Method Details

getModel

public RknnModel getModel()

Returns the model in use by this detector.

Specified by:

getModel in interface ObjectDetector

getClasses

public List<String> getClasses()

Returns the classes that the detector can detect

Specified by:

getClasses in interface ObjectDetector

Returns:

The classes

detect

public List<NeuralNetworkPipeResult> detect(org.opencv.core.Mat in,
 double nmsThresh,
 double boxThresh)

Detects objects in the given input image using the RknnDetector.

Specified by:

detect in interface ObjectDetector

Parameters:

in - The input image to perform object detection on.

nmsThresh - The threshold value for non-maximum suppression.

boxThresh - The threshold value for bounding box detection.

Returns:

A list of NeuralNetworkPipeResult objects representing the detected objects. Returns an empty list if the detector is not initialized or if no objects are detected.

release

public void release()

Thread-safe method to release the detector.

Specified by:

release in interface Releasable