Roadmap.txt

Catalano Framework 1.4 road map
-----------------------------------------


Version updates and fixes:

* Catalano.Android.Image
    - Added: Clear method in FastBitmap.
    - Added: get; setCoordinateSystem. (Now you can work with X and Y coordinates or I and J like the matrix.)
    - Added: getData method.
    - Added: indicateGrayscale in FastBitmap. (Used to optimized some filters to avoid clone in the memory.)

    - Optimized: Several operation filters (e.g grayscale, log, etc...)**
    - Optimized: Several filters that need to do a copy*.
* There are using recycle internally when needs some copies.
** There are directly accessing the pixels.

* Catalano.Core.Structs
   - New: Binary Heap.

* Catalano.Image
   - Added: get; setCoordinateSystem.
   - Added: get; setRGBData method.
   - Added: get; setGrayData method.

   - Fixed: some constructors in FastBitmap when are loading argb images.

* Catalano.Image.Filters
   - New: Separable Convolution.
   - New: Image Pyramids.
   - New: Weber Local Descriptor.
   - New: The Ovuscule (Active Contour).
   - New: CLAHE (Contrast Limited Adaptive Histogram Equalization).
   - New: Fast Radial Symmetry Transform.
   - New: Histogram Matching.
   - New: Alpha Trimmed Mean Filter.
   - New: Outline.
   - New: Mode.
   - New: Unsharp Mask.
   - New: Fast Variance.
   - New: Poisson Noise.

   - Added: getMaximumDistance in Distance Transform.
   - Added: support grayscale in Flood Fill.

   - Removed: get; setArithmetic in Sauvola/Niblack/WolfJoulion Threshold. (these filters are using separable convolution for to enhance the performance)

   - Renamed: Fast Gaussian Blur to Gaussian Box Blur.

   - Fixed: Replicate option in Convolution (fixed in Concurrent too).
   - Fixed: Gaussian Box Blur in grayscale images.
   - Fixed: Objective Fidelity was returning infinity values when occur 0 difference.

* Catalano.Image.Tools
   - New: Suk Flusser Moments.
   - New: Color Moments.
   - New: Kernel.
   - New: Chain Code.
   - New: Contour.
   - New: Filters Sequence.

   - Fixed: Smooth Noise Robust in Convolution Kernel.
   - Fixed: Blob Detection for grayscale images.
   - Added: new method Compute in Hu Moments.
   - Added: Ultimate Eroded Points in Shape Descriptors.
   - Added: Roughness in Shape Descriptors.

* Catalano.Image.Concurrent.Filters
   - New: Alpha Trimmed Mean.
   - New: Fast Variance.

* Catalano.Vision.Temporal
   - New: Mean Background Detector.
   - New: Median Background Detector.

* Catalano.MachineLearning
   - New: K Nearest Neighbour.

* Catalano.Math
    - New: Taylor Series.

    - Added: generic Transpose in Matrix.
    - Added: getReal, getImaginary in ComplexNumber for 1,2 dimensions.
    - Added: Sum in Matrix.
    - Added: CreateMatrix2D in Matrix.
    - Added: CreateMatrix1D in Matrix.
    - Added: Greatest common divisor in Tools.
    - Added: new overload in Chessboard distance.

* Catalano.Math.Decompositions
    - Added: Singular Value Decomposition.

* Catalano.Math.Graph
   - New: ISearch.

* Catalano.Math.Graph.Pathfinding
   - New: ANode.
   - New: NodeMap.
   - New: AStar.

* Catalano.Math.Geometry
   - New: Discrete Curve Evolution.

* Catalano.Math.Transforms
    - New: Fast Hilbert Transform.

    - Added: FFTShift in Fourier Transform.

    - Fixed: FFT, FFT2 in Fourier Transform. Now is working in Asymetric scaling. Removed power of 2 limitation in FFT and FFT2.

* Catalano.Statistics
   - New: Dissimilarity.
   - New: Approximation.

* Catalano.Statistics.Kernels
   - Added: Bessel.
   - Added: Circular.
   - Added: Dirichlet.
   - Added: Histogram Intersection.
   - Added: Hyper Secant.
   - Added: Inverse Multiquadric.
   - Added: Multiquadric.
   - Added: Rational Quadratic.
   - Added: Tensor.
   - Added: Wave.

   - Added: Alpha Trimmed Mean in Tools.
   - Added: Sum in Tools.

* General
   - Added: Comments in the filters inform for to developer about supported types (grayscale, rgb) and coodinate system used (cartesian, matrix or independ).