Package boofcv.alg.fiducial.square

Class DetectFiducialSquareBinary<T extends ImageGray<T>>

java.lang.Object
boofcv.alg.fiducial.square.BaseDetectFiducialSquare<T>
boofcv.alg.fiducial.square.DetectFiducialSquareBinary<T>
All Implemented Interfaces:
VerbosePrint
public class DetectFiducialSquareBinary<T extends ImageGray<T>> extends BaseDetectFiducialSquare<T>

Square fiducial that encodes numerical values in a binary N by N grids, where N ≥ 3. The outer border is entirely black while the inner portion is divided into a grid of equally sized back and white squares. Typical grid sizes are 3x3, 4x4, and 5x5, which can encode up to 32, 4096, and 2,097,152 unique values respectively. In other words, a grid of size N can encode N*N-4 bits, or a number with 2N*N-4 values. The lower left corner is always back and while all the other corners are always white. This allows orientation to be uniquely determined.

The above image is an example of a 4x4 grid and visually shows the fiducials internal coordinate system. The center of the fiducial is the origin of the coordinate system, e.g. all sides are width/2 distance away from the origin. +x is to the right, +y is up, and +z out of the paper towards the viewer. The black orientation corner is pointed out in the above image. The fiducial's width refers to the width of each side along the black border NOT the internal encoded image. The size of each square is the same and has a width of (fiducal width)*(1.0 - 2.0*(border fractional width))/N.

NOTE: While a larger grid size will allow you to encode more numbers it will increase the rate at which ID numbers are incorrectly identified.
NOTE: The size of the border can be adjusted, but 0.25 is recommended. The thinner the black border is the worse it will perform when viewed at an angle. However, the closer the fiducial is the less this is an issue allowing for thinner borders.

  • Field Details

  • Constructor Details

    • DetectFiducialSquareBinary

      public DetectFiducialSquareBinary(int gridWidth, double borderWidthFraction, double minimumBlackBorderFraction, InputToBinary<T> inputToBinary, DetectPolygonBinaryGrayRefine<T> quadDetector, Class<T> inputType)
      Configures the fiducial detector
      Parameters:
      gridWidth - Number of elements wide the encoded square grid is. 3,4, or 5 is recommended.
      borderWidthFraction - Fraction of the fiducial's width that the border occupies. 0.25 is recommended.
      inputToBinary - Converts the input image into a binary image
      quadDetector - Detects quadrilaterals in the input image
      inputType - Type of image it's processing

  • Method Details

    • processSquare

      protected boolean processSquare(GrayF32 gray, BaseDetectFiducialSquare.Result result, double edgeInside, double edgeOutside)
      Description copied from class: BaseDetectFiducialSquare
      Processes the detected square and matches it to a known fiducial. Black border is included.
      Specified by:
      processSquare in class BaseDetectFiducialSquare<T extends ImageGray<T>>
      Parameters:
      gray - Image of the undistorted square
      result - Which target and its orientation was found
      edgeInside - Average pixel value along edge inside
      edgeOutside - Average pixel value along edge outside
      Returns:
      true if the square matches a known target.

    • extractNumeral

      protected int extractNumeral()
      Extract the numerical value it encodes
      Returns:
      the int value of the numeral.

    • rotateClockWise

      protected void rotateClockWise()

    • thresholdBinaryNumber

      protected boolean thresholdBinaryNumber()
      Sees how many pixels were positive and negative in each square region. Then decides if they should be 0 or 1 or unknown

    • findBitCounts

      protected void findBitCounts(GrayF32 gray, double threshold)
      Converts the gray scale image into a binary number. Skip the outer 1 pixel of each inner square. These tend to be incorrectly classified due to distortion.

    • setLengthSide

      public void setLengthSide(double lengthSide)

    • getGridWidth

      public int getGridWidth()
      Number of elements wide the grid is

    • setAmbiguityThreshold

      public void setAmbiguityThreshold(double ambiguityThreshold)
      parameters which specifies how tolerant it is of a square being ambiguous black or white.
      Parameters:
      ambiguityThreshold - 0 to 1, inclusive

    • getNumberOfDistinctFiducials

      public long getNumberOfDistinctFiducials()

    • getGrayNoBorder

      public GrayF32 getGrayNoBorder()

    • getBinaryInner

      public GrayU8 getBinaryInner()

    • printClassified

      public void printClassified()