Documentation of 'boofcv.alg.feature.describe.DescribePointBinaryCompare' Java class
DescribePointBinaryCompare
boofcv.alg.feature.describe

Class DescribePointBinaryCompare<T extends ImageGray>

  • Direct Known Subclasses:
    ImplDescribeBinaryCompare_F32, ImplDescribeBinaryCompare_U8


    public abstract class DescribePointBinaryCompare<T extends ImageGray>
    extends java.lang.Object

    For each bit in the descriptor it samples two points inside an image patch and compares their values. A value of 1 or 0 is assigned depending on their relative values. This type of descriptor is referred to as a random tree [1], random fern, and is used in BRIEF.

    [1] Y. Amit and D. Geman. Shape Quantization and Recognition with Randomized Trees. Neural Computation, 9(7):1545–1588, 1997.

    • Method Detail

      • setImage

        public void setImage(T image)
        Specifies the image from which feature descriptions are to be created.
        Parameters:
        image - Image being examined.
      • process

        public void process(int c_x,
                            int c_y,
                            TupleDesc_B feature)
        Computes the descriptor at the specified point. If the region go outside of the image then a description will not be made.
        Parameters:
        c_x - Center of region being described.
        c_y - Center of region being described.
        feature - Where the descriptor is written to.
      • processInside

        public abstract void processInside(int c_x,
                                           int c_y,
                                           TupleDesc_B feature)
        Called if the descriptor region is contained entirely inside the image
      • processBorder

        public abstract void processBorder(int c_x,
                                           int c_y,
                                           TupleDesc_B feature)
        Called if the descriptor region goes outside the image border

DataMelt 3.0 © DataMelt by jWork.ORG

Ads help maintain this website.