boofcv.alg.fiducial.calib.chess

Class DetectChessboardSquarePoints<T extends boofcv.struct.image.ImageGray<T>>

java.lang.Object

boofcv.alg.fiducial.calib.chess.DetectChessboardSquarePoints<T>

public class DetectChessboardSquarePoints<T extends boofcv.struct.image.ImageGray<T>>
extends java.lang.Object

Given a binary image it detects the presence of chess board calibration grids. 1) Detect blobs in binary image and select square like ones. 2) Create two grids, inner and outer, 3) Put the grids together, 4) extract initial estimate for corners

Constructor Summary

Constructors

Constructor and Description
DetectChessboardSquarePoints(int numRows, int numCols, boofcv.struct.ConfigLength maxCornerDistance, boofcv.alg.shapes.polygon.DetectPolygonBinaryGrayRefine<T> detectorSquare) Configures chess board detector.

Method Summary

Modifier and Type	Method and Description
void	adjustBeforeOptimize(georegression.struct.shapes.Polygon2D_F64 polygon, org.ddogleg.struct.GrowQueue_B touchesBorder, boolean clockwise) The polygon detected from the contour is too small because the binary image was eroded.
org.ddogleg.struct.FastQueue<georegression.struct.point.Point2D_F64>	getCalibrationPoints()
boofcv.alg.shapes.polygon.DetectPolygonBinaryGrayRefine<T>	getDetectorSquare()
java.util.List<java.util.List<SquareNode>>	getGraphs()
SquareCrossClustersIntoGrids	getGrids()
int	getNumCols()
int	getNumRows()
SquaresIntoCrossClusters	getShapeToClusters()
boolean	process(T input, boofcv.struct.image.GrayU8 binary) Detects chessboard in the binary image.

Methods inherited from class java.lang.Object

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

Constructor Detail

DetectChessboardSquarePoints

public DetectChessboardSquarePoints(int numRows,
                                    int numCols,
                                    boofcv.struct.ConfigLength maxCornerDistance,
                                    boofcv.alg.shapes.polygon.DetectPolygonBinaryGrayRefine<T> detectorSquare)

Configures chess board detector.

Parameters:

numRows - Number of rows in square grid

numCols - Number of columns in square grid

maxCornerDistance - Maximum distance in pixels that two "overlapping" corners can be from each other.

Method Detail

process

public boolean process(T input,
                       boofcv.struct.image.GrayU8 binary)

Detects chessboard in the binary image. Square corners must be disconnected. Returns true if a chessboard was found, false otherwise.

Parameters:

input - Original input image.

binary - Binary image of chessboard

Returns:

True if successful.

adjustBeforeOptimize

public void adjustBeforeOptimize(georegression.struct.shapes.Polygon2D_F64 polygon,
                                 org.ddogleg.struct.GrowQueue_B touchesBorder,
                                 boolean clockwise)

The polygon detected from the contour is too small because the binary image was eroded. This expand the size of the polygon so that it fits the image edge better

getGraphs

public java.util.List<java.util.List<SquareNode>> getGraphs()

getShapeToClusters

public SquaresIntoCrossClusters getShapeToClusters()

getGrids

public SquareCrossClustersIntoGrids getGrids()

getDetectorSquare

public boofcv.alg.shapes.polygon.DetectPolygonBinaryGrayRefine<T> getDetectorSquare()

getCalibrationPoints

public org.ddogleg.struct.FastQueue<georegression.struct.point.Point2D_F64> getCalibrationPoints()

getNumRows

public int getNumRows()

getNumCols

public int getNumCols()