Package org.opencv.dnn

Class TextDetectionModel

java.lang.Object

org.opencv.dnn.Model

org.opencv.dnn.TextDetectionModel

Direct Known Subclasses:

TextDetectionModel_DB, TextDetectionModel_EAST

public class TextDetectionModel
extends Model

Base class for text detection networks

Field Summary

Fields inherited from class org.opencv.dnn.Model

nativeObj

Constructor Summary

Constructors

Modifier	Constructor	Description
protected	TextDetectionModel(long addr)

Method Summary

Modifier and Type	Method	Description
static TextDetectionModel	__fromPtr__(long addr)
void	detect(Mat frame, List<MatOfPoint> detections)
void	detect(Mat frame, List<MatOfPoint> detections, MatOfFloat confidences)	Performs detection Given the input frame, prepare network input, run network inference, post-process network output and return result detections.
void	detectTextRectangles(Mat frame, MatOfRotatedRect detections)
void	detectTextRectangles(Mat frame, MatOfRotatedRect detections, MatOfFloat confidences)	Performs detection Given the input frame, prepare network input, run network inference, post-process network output and return result detections.
protected void	finalize()

Methods inherited from class org.opencv.dnn.Model

getNativeObjAddr, predict, setInputCrop, setInputMean, setInputParams, setInputParams, setInputParams, setInputParams, setInputParams, setInputParams, setInputScale, setInputSize, setInputSize, setInputSwapRB, setPreferableBackend, setPreferableTarget

Methods inherited from class java.lang.Object

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

Constructor Detail

TextDetectionModel

protected TextDetectionModel(long addr)

Method Detail

__fromPtr__

public static TextDetectionModel __fromPtr__(long addr)

detect

public void detect(Mat frame,
                   List<MatOfPoint> detections,
                   MatOfFloat confidences)

Performs detection Given the input frame, prepare network input, run network inference, post-process network output and return result detections. Each result is quadrangle's 4 points in this order: - bottom-left - top-left - top-right - bottom-right Use cv::getPerspectiveTransform function to retrive image region without perspective transformations. Note: If DL model doesn't support that kind of output then result may be derived from detectTextRectangles() output.

Parameters:

frame - The input image

detections - array with detections' quadrangles (4 points per result)

confidences - array with detection confidences

detect

public void detect(Mat frame,
                   List<MatOfPoint> detections)

detectTextRectangles

public void detectTextRectangles(Mat frame,
                                 MatOfRotatedRect detections,
                                 MatOfFloat confidences)

Performs detection Given the input frame, prepare network input, run network inference, post-process network output and return result detections. Each result is rotated rectangle. Note: Result may be inaccurate in case of strong perspective transformations.

Parameters:

frame - the input image

detections - array with detections' RotationRect results

confidences - array with detection confidences

detectTextRectangles

public void detectTextRectangles(Mat frame,
                                 MatOfRotatedRect detections)

finalize

protected void finalize()
                 throws Throwable

Overrides:

finalize in class Model

Throws:

Throwable