org.firstinspires.ftc.robotcore.external.tfod

Interface TFObjectDetector

All Superinterfaces:

CameraStreamSource

public interface TFObjectDetector
extends CameraStreamSource

Interface for TensorFlow Object Detector.

Nested Class Summary

Nested Classes

Modifier and Type	Interface and Description
static class	TFObjectDetector.Parameters TFObjectDetector.Parameters provides configuration information for instantiating the TFObjectDetector

Method Summary

Modifier and Type	Method and Description
void	activate() Activates this TFObjectDetector so it starts recognizing objects.
void	deactivate() Deactivates this TFObjectDetector so it stops recognizing objects.
java.util.List<Recognition>	getRecognitions() Returns the list of recognitions.
java.util.List<Recognition>	getUpdatedRecognitions() Returns the list of recognitions, but only if they are different than the last call to getUpdatedRecognitions().
void	loadModelFromAsset(java.lang.String assetName, java.lang.String... labels) Loads a TFLite model from the indicated application asset, which must be of type .tflite.
void	loadModelFromFile(java.lang.String absoluteFileName, java.lang.String... labels) Loads a TFLite model from the indicated file, which must be a .tflite file and contain the full file path.
void	setClippingMargins(int left, int top, int right, int bottom) Sets the number of pixels to obscure on the left, top, right, and bottom edges of each image passed to the TensorFlow object detector.
void	setZoom(double magnification, double aspectRatio) Indicates that only the zoomed center area of each image will be passed to the TensorFlow object detector.
void	shutdown() Perform whatever cleanup is necessary to release all acquired resources.

Methods inherited from interface org.firstinspires.ftc.robotcore.external.stream.CameraStreamSource

getFrameBitmap

Method Detail

loadModelFromAsset

void loadModelFromAsset(java.lang.String assetName,
                        java.lang.String... labels)

Loads a TFLite model from the indicated application asset, which must be of type .tflite.

Parameters:

assetName - the name of the .tflite model asset to load

labels - the labels of the objects in the model

loadModelFromFile

void loadModelFromFile(java.lang.String absoluteFileName,
                       java.lang.String... labels)

Loads a TFLite model from the indicated file, which must be a .tflite file and contain the full file path.

Parameters:

absoluteFileName - the full path to the .tflite model file to load

labels - the labels of the objects in the model

activate

void activate()

Activates this TFObjectDetector so it starts recognizing objects.

deactivate

void deactivate()

Deactivates this TFObjectDetector so it stops recognizing objects.

setClippingMargins

void setClippingMargins(int left,
                        int top,
                        int right,
                        int bottom)

Sets the number of pixels to obscure on the left, top, right, and bottom edges of each image passed to the TensorFlow object detector. The size of the images are not changed, but the pixels in the margins are colored black.

setZoom

void setZoom(double magnification,
             double aspectRatio)

Indicates that only the zoomed center area of each image will be passed to the TensorFlow object detector. For no zooming, set magnification to 1.0. For best results, the aspect ratio should match the aspect ratio of the images that were used to train the TensorFlow model (1.7777 for 16/9).

getUpdatedRecognitions

java.util.List<Recognition> getUpdatedRecognitions()

Returns the list of recognitions, but only if they are different than the last call to getUpdatedRecognitions().

getRecognitions

java.util.List<Recognition> getRecognitions()

Returns the list of recognitions.

shutdown

void shutdown()

Perform whatever cleanup is necessary to release all acquired resources.