ComputeRadiiOfInfluence.m | Given a set of input points and a set of base points, compute the radii of influence for each point pair, determined by each point's 12 nearest neighbors. These are the same radii used by the locally weighted mean technique. The output parameter <radii> is an array of these radius values |

CorrectDistortion.m | This m-file pertains to the functionality behind the GUI interface defined in CorrectDistortion.fig. Here, all GUI elements are created and initialized. Further, back-end functionality is defined for loading data from SetupDistortionCorrection files and applying it to correct distorted sequences |

CreateControlPoints.m | This m-file is in charge of marching through the grid image, extracting centroids from each of the cells and computing the corresponding location of each centroid in the idealized, reconstructed space |

CreateLookupTable.m | Given an image transformation and the image resolution (which |

documentation.H [code] | |

FastDistortionCorrection.m | Given a pixel lookup table created from SetupDistortionCorrection and a distorted image, this function quickly transforms and resamples the image, yielding the corrected result which is stored in <newImage> |

GetErrorArray.m | Given two sets corresponding ordered pairs, return an array of the squared straight line distances between corresponding points. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |

GetErrorBins.m | This function takes two sets of corresponding point pairs as input and produces a binned error histogram. Currently, there are eight pre-defined bins: < 0.05 pixels, < 0.1, < 0.15, < 0.2, < 0.25, < 0.5, < 1.0, and >= 1.0. This can be made more flexible in the future |

MeasurePrecision.m | (Should be renamed MeasureAccuracy.m) This function measure and provides various accuracy statistics on a set of corresponding control point pairs. Mean error, max error, sum of squared errors, standard deviation are all provided for both x and y individually, and their combination as a Euclidean distance |

MinimizeError.m | This m-file contains the core functionality of the distortion correction algorithm. Given a set of control point pairs to operate on the algorithm uses a hierarchical correction technique to minimize the residual error between the point sets |

ReduceControlPairsForLWM.m | This m-file provides functionality for reducing the number of control point pairs used to create a locally weighted mean image transformation. Points are reduced with respect to a uniform grid to ensure consistent spacing in the reduced point sets |

RemovePointSubset.m | This function allows for set subtraction. Given a set A and a set B, where B is a subset of A, the difference of the two sets, A - B is returned and stored in <differenceSet> |

rnorm.m | Rnorm returns a column of norm values. Given an input matrix of X rows and Y columns it returns an X by 1 column of norms |

SetupDistortionCorrection.m | This m-file provides the underlying functionality that corresponds to the GUI defined in SetupDistortionCorrection.fig In this m-file, all elements of the GUI are created and initialized, and an interface is provided for applying distortion correction techniques to registration images and storing the necessary pieces of information for subsequent image correction |

Generated on Thu Jun 12 17:50:18 2008 for VideoCorrection by 1.5.1