MLKernel Overview

Introduction

Module Library for Kernel Based Image Processing Algorithms

This library contains a set of classes implementing algorithms for kernel based image filtering using the ML concepts.

There are three types of classes:

• Base classes and files :
  • ml::Kernel : A template class managing a matrix of kernel elements.
  • KernelBaseOp : A class derived from Module to support kernel based image filtering.
  • KernelOp : The typical class to derive from when implementing a new kernel based image filter.
  • mlKernelMacros.h : A file containing a set of helper macros useful to implement kernel based image filtering.

• Helper classes :
  • KernelEditor : A module useful to create / build user defined 6D - kernels.
  • KernelTools : (Template) Classes with much functionality used in many kernel filters.
  • KernelLineApplicatorBase, KernelLineApplicator and LineApplicator :
    Old-style classes only to support older kernel modules implemented outside this library. Not used any more in this library.

• Module classes like RankFilter, ConvolutionFilter, Sobel3D, RobertsFilter, LocalMaxima and many more.