Description: This readme describes the Biased and Filtered Point Sampling algorithm.
Before running the code, please run "compile_MEX.m" in the "dependent_files" directory.
BFPS.m is the main file that draws samples. A demo can be run from Matlab with the following command:
>> BFPS(double(imread( im.jpg )), 2, 1, 1, 0, 1)
For faster density estimates in non-parametric image segmentation, the fast Gauss transform can be used in place of Matlab s ksdensity function.
Copyright 2011. MIT. All Rights Reserved.
To Search:
File list (Check if you may need any files):
dependent_files
...............\generate_mask.m
...............\takemid.m
...............\calc_gradients.m
...............\calc_gradientsIMPORT.cpp
...............\set_pixel_colors.m
...............\calc_sdfIMPORT.cpp
...............\calc_curvature.m
...............\helperMEX.h
...............\calc_sdf.m
...............\compile_MEX.m
...............\curve_length_penalty.cpp
...............\binTree.cpp
...............\binTree.h
...............\distancePixel.cpp
...............\distancePixel.h
...............\linkedList.cpp
...............\linkedList.h
...............\calc_curvatureIMPORT.cpp
...............\is_border.m
...............\is_border.cpp
...............\is_borderIMPORT.cpp
...............\lognormpdf.m
...............\calc_sdf.cpp
...............\generate_initialization.m
...............\calculate_curve_length_gradients.m
...............\create_arrays.m
...............\quantize_index.m
...............\conv_perturbation.m
BFPS.m
im.jpg
README.txt
LICENSE.pdf
Download