plateform stringclasses 1
value | repo_name stringlengths 13 113 | name stringlengths 3 74 | ext stringclasses 1
value | path stringlengths 12 229 | size int64 23 843k | source_encoding stringclasses 9
values | md5 stringlengths 32 32 | text stringlengths 23 843k |
|---|---|---|---|---|---|---|---|---|
github | ojwoodford/ojwul-master | convert2gray.m | .m | ojwul-master/image/convert2gray.m | 394 | utf_8 | 20ef7eae75c457cf26bbaa3348131014 | %CONVERT2GRAY Convert an RGB image to grayscale
%
% B = convert2gray(A)
%
%IN:
% A - HxWxC input image, where C = 3 (RGB) or 1 (already grayscale).
%
%OUT:
% B - HxW grayscale output image, of the same class as A.
function im = convert2gray(im)
if size(im, 3) == 3
im = cast(reshape(double(reshape(im, [], 3))... |
github | ojwoodford/ojwul-master | imwarp.m | .m | ojwul-master/image/imwarp.m | 429 | utf_8 | e6216334a08f75eb5428bedae35469cf | %IMWARP Warp an image according to a homography
%
% im = imwarp(im, H)
%
%IN:
% im - HxWxC image
% H - 3x3 homography matrix from source to target
%
%OUT:
% im - HxWxC resampled output image
function im = imwarp(im, H)
% Compute the coordinates to sample at
X = proj(H \ homg(flipud(ndgrid_cols(1:size(im, 1), ... |
github | ojwoodford/ojwul-master | im2mov.m | .m | ojwul-master/image/im2mov.m | 5,227 | utf_8 | 41e469b5c3e2fb96e4aa5403e2c49598 | %IM2MOV Convert a sequence of images to a movie file
%
% Examples:
% im2mov infile outfile
% im2mov(A, outfile)
% im2mov(..., '-fps', n)
% im2mov(..., '-quality', q)
% im2mov(..., '-profile', profile)
% im2mov(..., '-nocrop')
%
% This function converts an image sequence to a movie.
%
% To cr... |
github | ojwoodford/ojwul-master | mov2im.m | .m | ojwul-master/image/mov2im.m | 641 | utf_8 | f623093ad1a8a6cae5684eb639972909 | %MOV2IM Convert a movie file to a sequence of images
%
% Examples:
% mov2im infile outfile_format
%
%IN:
% infile - string containing the name of the input video.
% outfile_format - format string for the movie frames. The filename for
% frame N is given by sprintf(outfile_format, N).
... |
github | ojwoodford/ojwul-master | imnorm.m | .m | ojwul-master/image/imnorm.m | 1,278 | utf_8 | a6b4caa2db5fa5166a2b34ff89d87eb9 | %IMNORM Spatially local image normalization
%
% B = imnorm(A, sigma, noise_variance)
% B = imnorm(A, [szy szx], noise_variance)
%
% Apply a local normalization operator (subtracting the mean and
% normalizing the variance) to an image, either with a Gaussian or window
% average weighting.
%
%IN:
% A - H... |
github | ojwoodford/ojwul-master | rng_seeder.m | .m | ojwul-master/utils/rng_seeder.m | 589 | utf_8 | 37ed4745dc7d52079ad2aaced05010da | %RNG_SEEDER Seed the random number generator, and print the seed if generated
%
% seed = rng_seeder()
% rng_seeder(seed)
%
% This function intializes the random number generator, and prints out the
% seed if one is not given or output.
%
%IN:
% seed - scalar seed for the random number generator.
%
%OUT:
% seed ... |
github | ojwoodford/ojwul-master | qfig.m | .m | ojwul-master/utils/qfig.m | 450 | utf_8 | 9f8367ca4007ca0eb5e1e401272d1009 | %QFIG Quietly select the figure
%
% fh = qfig(fn)
%
% Quietly selects the figure specified, without bringing it into focus
% (unless the figure doesn't exist yet).
%
% IN:
% fn - scalar positive integer, or figure handle indicating the figure
% to select.
%
% OUT:
% fh - handle to the figure.
functio... |
github | ojwoodford/ojwul-master | ojw_progressbar.m | .m | ojwul-master/utils/ojw_progressbar.m | 9,531 | utf_8 | ca29d395f2e8d30da82e8b0f31717b83 | %OJW_PROGRESSBAR Simple progress bar implementation
%
% [this, retval] = ojw_progressbar(tag, proportion, [total, [min_update_interval]])
%
% Starts, updates and closes a progress bar according to the proportion of
% time left. There are two ways of using the function:
%
% % Simple (one line) but more overhea... |
github | ojwoodford/ojwul-master | add_genpath_exclude.m | .m | ojwul-master/utils/add_genpath_exclude.m | 804 | utf_8 | ad7336761638201f97ff7c4fbd2a1b78 | %ADD_GENPATH_EXCLUDE Add a folder and subdirectories to the path, with exclusions
%
% add_genpath_exclude(folder_path, ...)
%
% For example:
% add_genpath_exclude('ojwul', '/.git', '\.git')
% adds ojwul and subdirectories to the path, excluding .git folders.
%
%IN:
% folder_path - Relative or absolute path to th... |
github | ojwoodford/ojwul-master | string2hash.m | .m | ojwul-master/utils/string2hash.m | 453 | utf_8 | 89a6f95ed4a295f057af5040c85cac81 | %STRING2HASH Convert a string to a 64 char hex hash string (256 bit hash)
%
% hash = string2hash(string)
%
%IN:
% string - a string!
%
%OUT:
% hash - a 64 character string, encoding the 256 bit SHA hash of string
% in hexadecimal.
function hash = string2hash(string)
persistent md
if isempty(md)
md ... |
github | ojwoodford/ojwul-master | col.m | .m | ojwul-master/utils/col.m | 391 | utf_8 | 7730db6cdeaee9ea0865f10334719fba | %COL Convert an array to a column vector along a particular dimension
%
% B = col(A, [dim])
%
%IN:
% A - Array of any size.
% dim - Positive integer indicating the dimension to arrange the elements
% of A along. Default: 1.
%
%OUT:
% B - Result of shiftdim(A(:), 1-dim).
function x = col(x, dim)
x = res... |
github | ojwoodford/ojwul-master | recurse_subdirs.m | .m | ojwul-master/utils/recurse_subdirs.m | 1,769 | utf_8 | 79eab70a46d565eaeeca552e4c34314e | %RECURSE_SUBDIRS Run a function recursively on a directory structure
%
% varargout = recurse_subdirs(func, base)
%
% This function calls a function, passing in the path to each subdirectory
% in the tree of the current directory (i.e. including subdirectories of
% subdirectories).
%
%IN:
% func - A handle ... |
github | ojwoodford/ojwul-master | ndgrid_cols.m | .m | ojwul-master/utils/ndgrid_cols.m | 727 | utf_8 | 5cbe7e8b53cc0ad85a2e74377eeaac21 | %NDGRID_COLS Like NDGRID, but creates column vectors from the outputs
%
% [X, sz] = ndgrid_cols(...)
%
% This function applies passes its inputs directly to NDGRID, then converts
% the outputs to row vectors, which are stacked vertically, so each
% combination of inputs becomes a column vector in the output matrix.
%... |
github | ojwoodford/ojwul-master | compile.m | .m | ojwul-master/utils/compile.m | 15,693 | utf_8 | b2adcd1ca5e0fd46a553b939e43229cb | %COMPILE Mex compilation helper function
%
% Examples:
% compile func1 func2 ... -option1 -option2 ...
%
% This function can be used to (re)compile a number of mex functions, but
% is also a helper function enabling inline compilation.
function varargout = compile(varargin)
% There are two types of call:
%... |
github | ojwoodford/ojwul-master | temp_cd.m | .m | ojwul-master/utils/temp_cd.m | 486 | utf_8 | d5d5464b0d93132d4d3e0090e0ec7165 | %TEMP_CD Switch to a directory for the duration of the calling function
%
% cwd = temp_cd(dirname)
%
%IN:
% dirname - Full or relative path to the directory to switch to.
%
%OUT:
% cwd - Path string to current directory.
function cwd = temp_cd(dirname)
assert(evalin('caller', 'exist(''temp_cd_cleanupObj'', ''var... |
github | ojwoodford/ojwul-master | str2fun.m | .m | ojwul-master/utils/str2fun.m | 1,423 | utf_8 | cbdee8fc068af54566c8a6c414da697f | %STR2FUN Construct a function_handle from a function name or path.
% FUNHANDLE = STR2FUN(S) constructs a function_handle FUNHANDLE to the
% function named in the character vector S. The S input must be a
% character vector. The S input cannot be a character array with
% multiple rows or a cell array of cha... |
github | ojwoodford/ojwul-master | vgg_argparse.m | .m | ojwul-master/utils/vgg_argparse.m | 1,946 | utf_8 | 9831dc65a204ed7cf28af3acd5487ffd | %VGG_ARGPARSE Parse variable arguments into a structure
% opts = vgg_argparse(inopts,varargin)
% inopts: structure (cells array) listing valid members and default values
% varargin: variable arguments of form '<name>',<value>,...
% opts: opts modified by varargin
%
% Example:
% function f = foo(va... |
github | ojwoodford/ojwul-master | user_string.m | .m | ojwul-master/utils/user_string.m | 3,273 | utf_8 | 1001def19cdf03ef0095097e934b6640 | %USER_STRING Get/set a user specific string
%
% Examples:
% string = user_string(string_name)
% isSaved = user_string(string_name, new_string)
%
% Function to get and set a string in a system or user specific file. This
% enables, for example, system specific paths to binaries to be saved.
%
% The specified strin... |
github | ojwoodford/ojwul-master | bsxfun.m | .m | ojwul-master/utils/bsxfun.m | 4,309 | utf_8 | 2f5ffcbfa7ab333f15990ab47e3fe49f | % BSXFUN Binary Singleton Expansion Function
% C = BSXFUN(FUNC,A,B) applies the element-by-element binary operation
% specified by the function handle FUNC to arrays A and B, with singleton
% expansion enabled. FUNC can be one of the following built-in functions:
%
% @plus Plus
% ... |
github | ojwoodford/ojwul-master | maximize.m | .m | ojwul-master/utils/maximize.m | 884 | utf_8 | 2b1e724a67b717d50a29e1186d1e6612 | %MAXIMIZE Maximize a figure window to fill the entire screen
%
% Examples:
% maximize
% maximize(hFig)
%
% Maximizes the current or input figure so that it fills the whole of the
% screen that the figure is currently on. This function is platform
% independent.
%
%IN:
% hFig - Handle of figure to maxi... |
github | ojwoodford/ojwul-master | bitwise_hamming.m | .m | ojwul-master/bitwise/bitwise_hamming.m | 4,768 | utf_8 | 9f5bb67a39572295bc759ee01457d8c9 | %BITWISE_HAMMING Compute all hamming distances between two sets of bit vectors
%
% C = bitwise_hamming(A, B, [thresh])
%
% Given two sets of bit vectors (each column being a bit vector), compute
% the hamming distances between all pairs of vectors between the two sets.
% If a threshold is given, return only tho... |
github | ojwoodford/ojwul-master | DataHash.m | .m | ojwul-master/bitwise/DataHash.m | 22,490 | utf_8 | f8f52f3077dddaf31779b71355a47695 | function Hash = DataHash(Data, varargin)
% DATAHASH - Checksum for Matlab array of any type
% This function creates a hash value for an input of any type. The type and
% dimensions of the input are considered as default, such that UINT8([0,0]) and
% UINT16(0) have different hash values. Nested STRUCTs and CELLs are par... |
github | ojwoodford/ojwul-master | bitcount.m | .m | ojwul-master/bitwise/bitcount.m | 1,281 | utf_8 | 08975f47b04e6ab70dc6894320d1e49d | %BITCOUNT Count the number of set bits in each column of the input
%
% B = bitcount(A)
%
% Count the number of set bits in each column of the input array,
% typecast as a bit vector.
%
%IN:
% A - MxNx... input array.
%
%OUT:
% B - 1xNx... output array of bit counts.
function A = bitcount(A)
persi... |
github | ojwoodford/ojwul-master | fit_gaussian.m | .m | ojwul-master/stats/fit_gaussian.m | 647 | utf_8 | e4bfa184e5ae30b876c1a3415e6ddfbd | %FIT_GAUSSIAN Fit a multi-variate gaussian to data.
%
% [mu, whiten] = fit_gaussian(X)
%
% Fit a multi-variate gaussian to data. This function can handle
% under-constrained data.
%
% IN:
% X - MxN matrix of N vectors of dimension M.
%
% OUT:
% mu - Mx1 distribution mean.
% whiten - Mx(min(M,N)) ... |
github | ojwoodford/ojwul-master | qpbo.m | .m | ojwul-master/optimize/qpbo.m | 4,377 | utf_8 | 5a8872c3010e87e8bc72194a2040c0d7 | %QPBO Binary MRF energy minimization on non-submodular graphs
%
% [L stats] = qpbo(UE, PI, PE, [TI, TE], [options])
%
% Uses the Quadratic Pseudo-Boolean Optimization (QPBO - an extension of
% graph cuts that solves the "roof duality" problem, allowing graphs with
% submodular edges to be solved) to solve binary, pa... |
github | ojwoodford/ojwul-master | cd_learn_normal.m | .m | ojwul-master/optimize/cd_learn_normal.m | 6,890 | utf_8 | 938e9df7096531476e5785896d25e2be | function estim = cd_learn_normal(varargin)
%CD_LEARN_NORMAL Demos contrastive divergence learning
%
% params = cd_learn_normal(options)
%
% Uses contrastive divergence to learn the parameters of a normal
% distribution that training data is generated from, and displays the
% results on completion. Assumes we don't k... |
github | ojwoodford/ojwul-master | dp_pair_chain.m | .m | ojwul-master/optimize/dp_pair_chain.m | 708 | utf_8 | cfd16b503833f9e7ed38686b86431031 | %DP_PAIR_CHAIN Dynamic programming on a chain of pairwise links
%
% Examples:
% L = dp_pair_chain(U, E)
% [L en] = dp_pair_chain(U, E)
%
% Minimize the cost of a set of pairwise chains.
%
%IN:
% U - PxQxR array of unary costs (double, single or uint32)
% E - PxP or PxPx(Q-1)xR array of pairwise costs (same type as U).... |
github | ojwoodford/ojwul-master | simulated_annealing.m | .m | ojwul-master/optimize/simulated_annealing.m | 4,238 | utf_8 | f7934e45f1d5640779962e7f527c9e5f | %SIMULATED_ANNEALING Perform simulated annealing on conditional energies
%
% [X en] = simulated_annealing(X0, energy, T, varargin)
%
% Simulated annealing for discrete energy minimization labelling problems
% for which conditional energy distributions can be computed. Given an
% initial labelling, a normalized... |
github | ojwoodford/ojwul-master | trw_bp.m | .m | ojwul-master/optimize/trw_bp.m | 2,447 | utf_8 | 57d5a8ab7552ea350909384ce697dec5 | %TRW_BP Multi-label MRF energy minimization using TRW-S & LBP
%
% [L energy lower_bound] = vgg_trw_bp(UE, PI, PE, [options])
%
% Uses the message passing algorithms TRW-S or LBP to solve an MRF energy
% minimization problem with binary or multiple labels.
%
% This function uses mexified C++ code written by Vladimir ... |
github | ojwoodford/ojwul-master | global_basin.m | .m | ojwul-master/optimize/global_basin.m | 718 | utf_8 | 4ed7d9117482aaf974cadcde31d9c101 | %GLOBAL_BASIN Output binary mask of watershed for the global min of array
%
% B = global_basin(A)
%
% This function computes the binary mask of all those points in an array
% from which local minimization (e.g. gradient descent with small steps)
% would lead to the global minimum of the array. This is the water... |
github | ojwoodford/ojwul-master | load_field.m | .m | ojwul-master/io/load_field.m | 317 | utf_8 | 0c31b7afe4727eefccf1f04a9c95a46f | %LOAD_FIELD Load only one field from a file
%
% x = load_field(name, field)
%
%IN:
% name - Filename string of the file containing the field.
% field - Name string of the field to be loaded.
%OUT:
% x - Loaded field.
function x = load_field(name, field)
x = load(name, field);
x = x.(field);
end |
github | ojwoodford/ojwul-master | read_wobj.m | .m | ojwul-master/io/read_wobj.m | 15,880 | utf_8 | 6ae16c6146e94ba4869abdc7d72d8084 | function OBJ=read_wobj(fullfilename)
% Read the objects from a Wavefront OBJ file
%
% OBJ=read_wobj(filename);
%
% OBJ struct containing:
%
% OBJ.vertices : Vertices coordinates
% OBJ.vertices_texture: Texture coordinates
% OBJ.vertices_normal : Normal vectors
% OBJ.vertices_point : Vertice data used for poi... |
github | ojwoodford/ojwul-master | read_float32.m | .m | ojwul-master/io/read_float32.m | 306 | utf_8 | ca7c4beb289a078ff916ee21d5220d62 | %READ_FLOAT32 Read an entire file in as an array of 32-bit floats
%
% A = read_float32(fname)
%
%IN:
% fname - string containing the filename of the file to be read.
%
%OUT:
% A - Nx1 single array of the values in the file.
function A = read_float32(fname)
A = read_bin(fname, 'float32'); |
github | ojwoodford/ojwul-master | write_text.m | .m | ojwul-master/io/write_text.m | 1,376 | utf_8 | 220bb0478703b9cb82ce59c35f18ab32 | %WRITE_TEXT Write out an array to a text file
%
% write_text(A, fname, append)
%
% Writes out an array to a text file using the minimum number of
% significant figures required to reconstruct the exact binary number when
% read in.
%
% The array is written out row by row, with dimensions 3 and higher
% conca... |
github | ojwoodford/ojwul-master | read_ply.m | .m | ojwul-master/io/read_ply.m | 5,687 | utf_8 | b34b22a96aec386a8a7c5cbf579eb898 | %% read ply
% Read mesh data from ply format mesh file
%
%% Syntax
% [face,vertex]= read_ply(filename)
% [face,vertex,color] = read_ply(filename)
%
%% Description
% filename: string, file to read.
%
% face : double array, nf x 3 array specifying the connectivity of the mesh.
% vertex: double array, nv x 3 arra... |
github | ojwoodford/ojwul-master | get_user_path.m | .m | ojwul-master/io/get_user_path.m | 1,844 | utf_8 | be1d8bac23899f0064233e1635f9aee4 | %GET_USER_PATH Get a user/computer-specific directory or file path
%
% path_str = get_user_path(name, check_path, type [append])
%
% Ask a user to select a specific directory or file, and store its path.
% If a valid path already exists, use this.
%
%IN:
% name - Name of the directory to be found.
% check_path - ... |
github | ojwoodford/ojwul-master | json_write.m | .m | ojwul-master/io/json_write.m | 794 | utf_8 | 88d106691d8319da9c329c12fcc0a16d | %JSON_WRITE Write a MATLAB variable to a JSON file
%
% json_write(var, fname)
%
% This function wraps the JSON for Modern C++ class in a mex wrapper, for
% fast writing of MATLAB variables into JSON files.
%
%IN:
% var - MATLAB variable to be written to a JSON file.
% filename - String of filename (if... |
github | ojwoodford/ojwul-master | stlwrite.m | .m | ojwul-master/io/stlwrite.m | 10,820 | utf_8 | c0d2afd34d9e64039055c0120d2a0800 | function stlwrite(filename, varargin)
%STLWRITE Write STL file from patch or surface data.
%
% STLWRITE(FILE, FV) writes a stereolithography (STL) file to FILE for a
% triangulated patch defined by FV (a structure with fields 'vertices'
% and 'faces').
%
% STLWRITE(FILE, FACES, VERTICES) takes faces an... |
github | ojwoodford/ojwul-master | write_wobj.m | .m | ojwul-master/io/write_wobj.m | 9,798 | utf_8 | 3004ab65ebe5ca6b284c74c7b326861d | function write_wobj(OBJ,fullfilename)
% Write objects to a Wavefront OBJ file
%
% write_wobj(OBJ,filename);
%
% OBJ struct containing:
%
% OBJ.vertices : Vertices coordinates
% OBJ.vertices_texture: Texture coordinates
% OBJ.vertices_normal : Normal vectors
% OBJ.vertices_point : Vertice data used for point... |
github | ojwoodford/ojwul-master | write_bin.m | .m | ojwul-master/io/write_bin.m | 715 | utf_8 | a0c3679b8d3d39851291017cce94c39d | %WRITE_BIN Write out an array to a binary file
%
% write_bin(A, fname)
%
% Writes out an array to a binary file in the format of the data in the
% array. I.e. if the array is of type uint32 then the values are saved to
% file as 32-bit unsigned integers. This function cannot save complex
% numbers.
%
%IN:
%... |
github | ojwoodford/ojwul-master | json_read.m | .m | ojwul-master/io/json_read.m | 812 | utf_8 | bab4c1c2c0d67a3e0d2f0d01b385fa9a | %JSON_READ Mex wrapper to C++ class for fast reading of JSON files
%
% var = json_read(filename)
%
% This function wraps the JSON for Modern C++ class in a mex wrapper, for
% fast loading of JSON files into MATLAB variables.
%
%IN:
% filename - String of filename (if in current directory) or full or
% ... |
github | ojwoodford/ojwul-master | read_bin.m | .m | ojwul-master/io/read_bin.m | 498 | utf_8 | 9b2d92e5eda0ed038bf0e108342fbac4 | %READ_BIN Read an entire file in as an array of a given type
%
% A = read_bin(fname, type)
%
%IN:
% fname - string containing the filename of the file to be read.
% type - string indicating the datatype of the values in the file.
%
%OUT:
% A - Nx1 array of the values in the file, of datatype type.
f... |
github | ojwoodford/ojwul-master | fopens.m | .m | ojwul-master/io/fopens.m | 5,918 | utf_8 | b712bcfc2438ae7217099a12ba390fc6 | %FOPENS Open file which is always closed when the function exits.
% FID = FOPENS(FILENAME) opens the file FILENAME for read access.
% FILENAME is a string containing the name of the file to be opened.
% (On PC systems, FOPENS opens files for binary read access.)
%
% FILENAME can be a MATLABPATH relative part... |
github | ojwoodford/ojwul-master | svm_read_sparse.m | .m | ojwul-master/classify/svm_read_sparse.m | 254 | utf_8 | 2156b44a43e178996d2b95d5723a22f5 | % SVM_READ_SPARSE Mex wrapper interface to the svm library
function varargout = svm_read_sparse(varargin)
sourceList = {'svm_read_sparse.c'}; % Cell array of source files
[varargout{1:nargout}] = compile(varargin{:}); % Compilation happens here
return
|
github | ojwoodford/ojwul-master | roc_curve.m | .m | ojwul-master/classify/roc_curve.m | 1,184 | utf_8 | ee9f65b707b7f495ee759487bafaa1d2 | %ROC_CURVE Compute the x and y parameters of an ROC curve
%
% [X, Y, auc] = roc_curve(scores, ground_truth)
%
% Compute the parameters of a Receiver Operating Characteristic curve,
% which plots true positive rate against false positive rate, over a range
% of classification thresholds.
%
%IN:
% scores - Mx1 vector... |
github | ojwoodford/ojwul-master | svm_predict.m | .m | ojwul-master/classify/svm_predict.m | 1,026 | utf_8 | 688b6ffb6f6bdb0e0f1d9e065a701fd1 | % SVM_PREDICT Mex wrapper interface to the svm library
%
% [predicted_label, accuracy, decision_values/prob_estimates] = svmpredict(testing_label_vector, testing_instance_matrix, model [,'libsvm_options']);
%
% -testing_label_vector:
% An m by 1 vector of prediction labels. If labels of test
... |
github | ojwoodford/ojwul-master | fiksvm_predict.m | .m | ojwul-master/classify/fiksvm_predict.m | 971 | utf_8 | 619c7d52f9ec6e4903b099fe3295215d | % FIKSVM_PREDICT Mex wrapper interface to the svm library
%
% Usage: [exact_values, pwconst_values, pwlinear_values,[times]] = ...
% fiksvm_predict(testing_label_vector, testing_instance_matrix, model,'libsvm_options')
%
% Output:
% exact_values : predictions using binary search
% pwconst_values ... |
github | ojwoodford/ojwul-master | svm_demo.m | .m | ojwul-master/classify/svm_demo.m | 2,009 | utf_8 | e635ff91ea3db75613b2d90885d71a9e | %SVM_DEMO A simple demo of SVM classification
%
% svm_demo(N)
%
% Train different binary SVM classifiers on 2D data, and visualize the
% results.
%
%IN:
% N - Integer number of 2D points to train and test on. Default: 500.
function svm_demo(N)
if nargin < 1
N = 500;
end
% Create some classifica... |
github | ojwoodford/ojwul-master | linear_predict.m | .m | ojwul-master/classify/linear_predict.m | 1,333 | utf_8 | 51ebe96e9f67b299e7822af80f846381 | % LINEAR_PREDICT Mex wrapper interface to the linear svm library
%
% [predicted_label, accuracy, decision_values/prob_estimates] = linear_predict(testing_label_vector, testing_instance_matrix, model [, 'liblinear_options', 'col']);
%
% -testing_label_vector:
% An m by 1 vector of prediction label... |
github | ojwoodford/ojwul-master | linear_train.m | .m | ojwul-master/classify/linear_train.m | 1,135 | utf_8 | 9d26e561f17d09f7178f3a3c68b33e85 | % LINEAR_TRAIN Mex wrapper interface to the linear svm library
%
% model = linear_train(training_label_vector, training_instance_matrix [,'liblinear_options', 'col']);
%
% -training_label_vector:
% An m by 1 vector of training labels. (type must be double)
% -training_instance_matrix:
% ... |
github | ojwoodford/ojwul-master | svm_precomp_model.m | .m | ojwul-master/classify/svm_precomp_model.m | 414 | utf_8 | 93448d99ffc7328ebb914386e77e8eb8 | % SVM_PRECOMP_MODEL Mex wrapper interface to the svm library
function varargout = svm_precomp_model(varargin)
sourceDir = 'private/libsvm/';
sourceList = {['-I' sourceDir], 'svm_precomp_model.cpp', [sourceDir 'svm.cpp'], ...
[sourceDir 'svm_model_matlab.cpp'], [sourceDir 'fiksvm.cpp']}; % Cell array of ... |
github | ojwoodford/ojwul-master | svm_train.m | .m | ojwul-master/classify/svm_train.m | 824 | utf_8 | 6e6b58d056e746c8dd4e981e19013be6 | % SVM_TRAIN Mex wrapper interface to the svm library
%
% model = svm_train(training_label_vector, training_instance_matrix, [,'libsvm_options']);
%
% -training_label_vector:
% An m by 1 vector of training labels.
% -training_instance_matrix:
% An m by n matrix of m training i... |
github | ojwoodford/ojwul-master | separable_steerable_filter.m | .m | ojwul-master/filter/separable_steerable_filter.m | 3,348 | utf_8 | e3883fff343cc4c0c1a56aa00bbac3be | %SEPARABLE_STEERABLE_FILTER Compute the basis and weights of separable steerable filters
%
% [basis, weight_fun] = separable_steerable_filter(r, coeff, X)
%
% Decompose a steerable filter formed from an odd or even parity polynomial
% times a Gaussian window into a separable basis using the equations in
% Appen... |
github | ojwoodford/ojwul-master | find_first.m | .m | ojwul-master/numeric/find_first.m | 1,001 | utf_8 | f48f512bfb5b53e2022805b0c86b9125 | %FIND_FIRST Fast, vectorized version of find(A, 1, 'first')
%
% B = find_first(A)
% B = find_first(A, start)
% B = find_first(A, operator, value)
% B = find_first(A, operator, value, start)
%
% Find the first element in each column vector of an array which meets a
% particular comparison criterion.
%
% Example:
% fin... |
github | ojwoodford/ojwul-master | dimsel.m | .m | ojwul-master/numeric/dimsel.m | 1,511 | utf_8 | 21dccb0ea6801c98b303e4707b05f5f8 | %DIMSEL Select one indexed element from each vector along a given dimension
%
% B = dimsel(A, I)
%
% Given as input a numeric array, A, and an index array, I, which contains
% indices along a particular dimension of A, output the indexed elements of
% A.
%
% For example, in the code:
% A = rand(3);
% [B... |
github | ojwoodford/ojwul-master | first.m | .m | ojwul-master/numeric/first.m | 819 | utf_8 | 6656ebac8a73f0f3965ebb5eb3161baa | %FIRST Returns indices of the first non-zero elements along the given dimension
%
% B = first(A, [dim])
%
% For each vector along the given dimension, this function returns the
% index of the first non-zero element along that vector, or 0 if there is
% no non-zero element.
%
%IN:
% A - MxNx... input array
... |
github | ojwoodford/ojwul-master | gauss_mask.m | .m | ojwul-master/numeric/gauss_mask.m | 1,101 | utf_8 | fb9150428ba19cc33aa5d2f465a0a953 | %GAUSS_MASK Compute 1D Nth derivative of a Gaussian
%
% Examples:
% F = gauss_mask(sigma)
% F = gauss_mask(sigma, deriv)
% F = gauss_mask(sigma, deriv, X)
%
% This function computes the Nth derivative of a Gaussian, in 1D.
%
% IN:
% sigma - Standard deviation of the Gaussian.
% deriv - The derivative to... |
github | ojwoodford/ojwul-master | zero_mean.m | .m | ojwul-master/numeric/zero_mean.m | 523 | utf_8 | b8edf0ff597b9b20b5cf70e48aeaee55 | %ZERO_MEAN Subtract the means from a set of vectors to make them zero mean
%
% Y = zero_mean(X)
% Y = zero_mean(X, dim)
%
% Subtract the mean along a specified dimension from all vectors in an
% array, making them zero mean.
%
%IN:
% X - Array containing vectors to zero mean.
% dim - Dimension along ... |
github | ojwoodford/ojwul-master | isapprox.m | .m | ojwul-master/numeric/isapprox.m | 1,409 | utf_8 | 5ebc8a650be39b533bbbfe311488be94 | %ISAPPROX Check if A and B are approximately equal
%
% [tf, d] = isapprox(A, B, tol)
%
% Determines whether two input arrays are approximately equal. If a
% tolerance is given, and no outputs are requested, the function asserts if
% the inputs aren't approximately equal.
%
%IN:
% A - Numeric array.
% B - Nume... |
github | ojwoodford/ojwul-master | extremum.m | .m | ojwul-master/numeric/extremum.m | 614 | utf_8 | 51a4fafafae5abcb7335319637a6da18 | %EXTREMUM Compute the extreme value along a given dimension
%
% B = extremum(A, [dim])
%
% Output the most extreme value (furthest from 0) along a specified
% dimension of an input array.
%
%IN:
% A - Numeric input array.
% dim - Dimension along which to compute the extremum. Default: first
% no... |
github | ojwoodford/ojwul-master | sqdist.m | .m | ojwul-master/numeric/sqdist.m | 784 | utf_8 | 56bd1dee4c97472b5aebb3d958a26641 | %SQDIST Squared Euclidean distance between sets of vectors
%
% D = sqdist(A, [B])
%
% IN:
% A - MxJ matrix of columnwise vectors.
% B - MxK matrix of columnwise vectors. Default: B = A.
%
% OUT:
% D - JxK Squared distance between each vector in A and each vector in B.
function D = sqdist(A, B)
... |
github | ojwoodford/ojwul-master | normalize.m | .m | ojwul-master/numeric/normalize.m | 582 | utf_8 | a11f36e044a61612a4bb2b0651a2062b | %NORMALIZE Set vectors in an array to be of unit length
%
% Y = normalize(X)
% Y = normalize(X, dim)
%
% Set all the non-zero vectors in an array, along a specified dimension, to
% be of unit length.
%
%IN:
% X - Array containing vectors to normalize.
% dim - Dimension along which to normalize X. Defa... |
github | ojwoodford/ojwul-master | all_finite.m | .m | ojwul-master/numeric/all_finite.m | 291 | utf_8 | 1f4977902fa9382cafc15932c81c65f7 | %ALL_FINITE Checks if all the elements in an array are finite
%
% tf = all_finite(x)
%
%IN:
% x - Numeric array.
%
%OUT:
% tf - Boolean indicating whether all elements of x are finite.
function x = all_finite(x)
if issparse(x)
[~, ~, x] = find(x);
end
x = all(isfinite(x(:)));
end |
github | ojwoodford/ojwul-master | array_snake_indices.m | .m | ojwul-master/numeric/array_snake_indices.m | 921 | utf_8 | 1036fca344d56d2f1a94ebffb868560d | %ARRAY_SNAKE_INDICES List of indices snaking through an array
%
% I = array_snake_indices(sz)
%
% Produces a list of all the indices into an array of size sz, with each
% consecutive index referencing a neighbouring element to the previous
% index.
%
%IN:
% sz - 1xN vector of the size of array to index.
%
%OUT:
% ... |
github | ojwoodford/ojwul-master | range01.m | .m | ojwul-master/numeric/range01.m | 665 | utf_8 | e05a9fe686b93d39d6355913b9be9179 | %RANGE01 Apply gain and bias so range of data is exactly [0, 1]
%
% Y = range01(X, [dim])
%
% Set all vectors in X along the specified dimension to be in the range
% [0,1].
%
%IN:
% X - Array containing vectors to rescale to range [0, 1].
% dim - Dimension along which to zero-mean X. Default: first
% ... |
github | ojwoodford/ojwul-master | normd.m | .m | ojwul-master/numeric/normd.m | 483 | utf_8 | 15ade47822ea81b1839ab3528ecf0fa4 | %NORMD Compute the 2-norms of vectors in an array along a specific dimension
%
% Y = normd(X)
% Y = normd(X, dim)
%
% Compute the 2-norms of vectors in an array, along a specific dimension.
%
%IN:
% X - Array containing vectors to compute the norms of.
% dim - Dimension along which to compute the 2-nor... |
github | ojwoodford/ojwul-master | inv44n.m | .m | ojwul-master/linear/inv44n.m | 2,251 | utf_8 | 14bff2f4c1537e18cea4be0b8fc0b81c | %INV44N Compute the inverse of an array of 4x4 matrices
%
% [B, d] = inv44n(A)
%
% Vectorized computation of the inverse of multiple 4x4 matrices.
%
%IN:
% A - 4x4xN array.
%
%OUT:
% B - 4x4xN array, where B(:,:,a) = inv(A(:,:,a)).
% d - 1xN array, where d(a) = det(A(:,:,a)).
function [B, det] = i... |
github | ojwoodford/ojwul-master | cross.m | .m | ojwul-master/linear/cross.m | 2,275 | utf_8 | 730d29b6bf45181cf2d3a230017ba4af | function c = cross(a,b,dim)
%CROSS Vector cross product.
% C = CROSS(A,B) returns the cross product of the vectors
% A and B. That is, C = A x B. A and B must be 3 element
% vectors.
%
% C = CROSS(A,B) returns the cross product of A and B along the
% first dimension of length 3.
%
% C = CROSS(A,B,DIM), w... |
github | ojwoodford/ojwul-master | isposdef.m | .m | ojwul-master/linear/isposdef.m | 266 | utf_8 | 0c7220c903df17ddae3c6d1f2f6d7544 | %ISPOSDEF Check if a square matrix is positive definite
%
% tf = isposdef(A)
%
%IN:
% A - NxN matrix.
%
%OUT:
% tf - boolean indicating whether A is positive definite.
function tf = isposdef(A)
[~, tf] = chol(A);
tf = (tf == 0) && (rank(A) == size(A, 1));
end |
github | ojwoodford/ojwul-master | inv33n.m | .m | ojwul-master/linear/inv33n.m | 564 | utf_8 | cc713280e9ef287893c776d0bb44b625 | %INV33N Compute the inverse of an array of 3x3 matrices
%
% [B, d] = inv33n(A)
%
% Vectorized computation of the inverse of multiple 3x3 matrices.
%
%IN:
% A - 3x3xN array.
%
%OUT:
% B - 3x3xN array, where B(:,:,a) = inv(A(:,:,a)).
% d - 1xN array, where d(a) = det(A(:,:,a)).
function [T, det] = i... |
github | ojwoodford/ojwul-master | invSE3n.m | .m | ojwul-master/linear/invSE3n.m | 435 | utf_8 | b18da849ddbbaed68c17c33dbcb4a630 | %INVSE3N Compute the inverse of an array of 3x4 SE3 matrices
%
% [B, d] = invSE3n(A)
%
% Vectorized computation of the inverse of multiple 3x4 SE3 matrices.
%
%IN:
% A - 3x4xN array of SE3 matrices.
%
%OUT:
% B - 3x4xN array of inverse SE3 matrices.
function T = invSE3n(T)
R = permute(T(1:3,1:3,:), ... |
github | ojwoodford/ojwul-master | det33n.m | .m | ojwul-master/linear/det33n.m | 584 | utf_8 | 3d07d701f15024db467c402a4600390a | %DET33N Compute the determinant of an array of 3x3 matrices
%
% d = det33n(A)
%
% Vectorized computation of the determinant of multiple 3x3 matrices.
%
%IN:
% A - 3x3xN array.
%
%OUT:
% d - 1xN array, where d(a) = det(A(:,:,a)).
function T = det33n(T)
T = reshape(T, 9, []);
cond = zeros(size(T));
... |
github | ojwoodford/ojwul-master | pca.m | .m | ojwul-master/linear/pca.m | 1,299 | utf_8 | 41e51e264202aacba6887d551cf4746e | %PCA Principal component analysis
%
% [T, Y, V] = pca(X)
%
% IN:
% X - MxN matrix of N vectors of dimension M.
%
% OUT:
% T - Mx(M+1) Projection matrix for PCA transformation
% Y - MxN matrix of transformed vectors, where Y = T * [X; ones(1, N)].
% V - Mx1 list of eigen values associated with ea... |
github | ojwoodford/ojwul-master | tmult.m | .m | ojwul-master/linear/tmult.m | 2,711 | utf_8 | a0e531042d0030663fd3f37c2dc1fbc2 | %TMULT Tensor matrix multiply
%
% C = tmult(A, B, [transpose])
%
% Matrix multiplication over tensor arrays (i.e. arrays of matrices), with
% the ability to transpose matrices first.
%
% C = tmult(A, B) is equivalent to:
%
% sz = [size(B) 1];
% sz(1) = size(A, 1);
% C = zeros(sz);
% for a = 1:pr... |
github | ojwoodford/ojwul-master | chol22n.m | .m | ojwul-master/linear/chol22n.m | 525 | utf_8 | cac431a9a2f1d476ea0b607172597a6f | %CHOL22N Compute the Cholesky decomposition of an array of 2x2 matrices
%
% B = chol22n(A)
%
% Vectorized computation of the Cholesky decomposition of multiple 2x2
% matrices.
%
%IN:
% A - 2x2xN array.
%
%OUT:
% B - 2x2xN array, where B(:,:,a) = chol(A(:,:,a), 'lower').
% Formula from here: http://m... |
github | ojwoodford/ojwul-master | whiten_srt.m | .m | ojwul-master/linear/whiten_srt.m | 693 | utf_8 | 81a53816a156dc60778e232690a33aa9 | %WHITEN_SRT Transform data to be zero mean and close to identity covariance
%
% [Y, T] = whiten_srt(X)
%
% Apply a similarity transform to the data such that the mean is zero and
% the covariance is as close to the identity as possible.
%
%IN:
% X - MxN array of N vectors of dimension M to be whitened
%
% Y - (M+... |
github | ojwoodford/ojwul-master | whiten.m | .m | ojwul-master/linear/whiten.m | 762 | utf_8 | 74a208f31dba9cc99791735e5ef5310d | %WHITEN Transform data to be zero mean and identity covariance
%
% [Y, T] = whiten(X, [epsilon])
%
%IN:
% X - MxN array of N vectors od dimension M to be whitened
% epsilon - scalar value to add to eigen values to avoid amplifying
% noise. Default: 1e-4.
%
% Y - MxN array of whitened data.
% T - (... |
github | ojwoodford/ojwul-master | linear_regressor.m | .m | ojwul-master/linear/linear_regressor.m | 2,399 | utf_8 | 2b1edd7949fb50b4886bc0009e8a8527 | classdef linear_regressor < handle
properties (SetAccess = private, Hidden = true)
parameters;
regularization_lambda = 0;
isquadratic = false;
end
methods
function this = linear_regressor(quad, reg)
if nargin > 0
this.isquadratic = quad;
... |
github | ojwoodford/ojwul-master | eig22n.m | .m | ojwul-master/linear/eig22n.m | 710 | utf_8 | d4eb84365738b510a317624640f6432c | %EIG22N Compute the eigenvalues and eigenvectors of 2x2 matrices
%
% [e, V] = eig22n(A)
%
% Vectorized computation of the eigenvalues and eigenvectors of multiple
% 2x2 matrices.
%
%IN:
% A - 2x2xN array.
%
%OUT:
% e - 2x1xN array, where e(:,a) = eig(A(:,:,a)).
% V - 2x2xN array, where [V(:,:,a), ~] = eig(A(:,:... |
github | ojwoodford/ojwul-master | inv22n.m | .m | ojwul-master/linear/inv22n.m | 509 | utf_8 | d3988bcba15c758442d58d984be90c23 | %INV22N Compute the inverse of an array of 2x2 matrices
%
% [B, d] = inv22n(A)
%
% Vectorized computation of the inverse of multiple 2x2 matrices.
%
%IN:
% A - 2x2xN array.
%
%OUT:
% B - 2x2xN array, where B(:,:,a) = inv(A(:,:,a)).
% d - 1xN array, where d(a) = det(A(:,:,a)).
function [T, det] = i... |
github | ojwoodford/ojwul-master | edge_demo.m | .m | ojwul-master/edges/edge_demo.m | 1,554 | utf_8 | 9a4ed09075dbf09302013c950184e2b6 | %EDGE_DEMO Compute and visualize the gradient and edgels of an image
%
% edge_demo
% edge_demo(A, scale, thresh)
%
%IN:
% A - HxWxC image. Default: Use peppers.png
% scale - sigma of Gaussian blur to apply during edge detection.
% Default: 0.7.
% thresh - Threshold on edgel suppression. If... |
github | ojwoodford/ojwul-master | edge_orient.m | .m | ojwul-master/edges/edge_orient.m | 1,697 | utf_8 | 15cb5f6b3da227debeff758ef7043493 | %EDGE_ORIENT Calculate edge filter responses at any orientation
%
% [R, O, G] = edge_orient(I, sigma, [mcm])
%
% Returns a handle to a function that generates an edge filter response
% image using angularly adaptive filtering, and also outputs the angle that
% gives maximal edge response. The implementation us... |
github | ojwoodford/ojwul-master | edge_grad_max.m | .m | ojwul-master/edges/edge_grad_max.m | 2,128 | utf_8 | 59bbc912b76006eae4bc3b0b297e10d2 | %EDGE_GRAD_MAX Mask of gradient maxima
%
% M = edge_grad_max(G, [thresh, [M]])
%
% Given a gradient image, computes the magnitude of locally maximal (in the
% direction of maximum gradient) gradients.
%
% IN:
% G - HxWx2 array of gradient images in x and y directions, along third
% dimension.
% t... |
github | ojwoodford/ojwul-master | edge_scale.m | .m | ojwul-master/edges/edge_scale.m | 3,642 | utf_8 | 2f8b378bdf47530ade9e24ab3a115f5a | %EDGE_SCALE Compute an edge scale image
%
% [S, G, E] = edge_scale(I, [max_octaves, [levels_per_octave, [thresh]]])
%
% Output the edge scale of each pixel, and a list of edgels, if requested.
% The scale of an edge is determined by the amount of image smoothing
% required before it stops being a locally maxim... |
github | ojwoodford/ojwul-master | compute_edgels.m | .m | ojwul-master/edges/compute_edgels.m | 3,118 | utf_8 | f0cfd6d776fe3d002ffd49a5f7164962 | %COMPUTE_EDGELS Extract edgels from a gradient image
%
% E = compute_edgels(G, thresh)
%
% Returns a list of edgel positions and normal angles, each edgel being a
% pixel long, given a gradient image and threshold parameter.
%
% IN:
% G - HxWx2 array of gradient images in x and y directions, along third
%... |
github | ojwoodford/ojwul-master | tensor_voted_edges.m | .m | ojwul-master/edges/tensor_voted_edges.m | 2,114 | utf_8 | 0f69548b15f818bc7e694233d37f96d7 | % Author: Emmanuel Maggiori. March 2014.
%
% Complimentary material for the literature review:
% "Perceptual grouping by tensor voting: a comparative survey of recent approaches". E Maggiori, HL Manterola, M del fresno. To be published in IET Computer Vision.
%
% Implementation of Steerable Tensor Voting as published... |
github | ojwoodford/ojwul-master | fast_kmeans.m | .m | ojwul-master/cluster/fast_kmeans.m | 1,324 | utf_8 | 6cf23c5f257a6576ed755100821fcb1f | %FAST_KMEANS Compute k-means cluster centres
%
% [CX sse I] = fast_kmeans(X, params[, CX])
%
% This function computes k-means cluster centres. It will work faster if
% data is projected onto principle components first.
%
% IN:
% X - MxN matrix of N input vectors of dimension M.
% params - [nclusters max_it... |
github | ojwoodford/ojwul-master | isautodiff.m | .m | ojwul-master/autodiff/isautodiff.m | 104 | utf_8 | 114480bc63ab1389236fc7fac0cbbea5 | %ISAUTODIFF Helper for autodiff
function tf = isautodiff(varargin)
tf = false(1, max(nargin, 1));
end |
github | ojwoodford/ojwul-master | var_indices.m | .m | ojwul-master/autodiff/var_indices.m | 90 | utf_8 | 343c2c6be93b20eb08c12d5780c90d43 | %VAR_INDICES Dummy helper function for autodiff
function c = var_indices(a)
c = [];
end |
github | ojwoodford/ojwul-master | autodiff.m | .m | ojwul-master/autodiff/autodiff.m | 25,193 | utf_8 | 27a380b7799c8c8099d41e76ab04c2e2 | % A class for doing autodifferentiation
classdef autodiff
properties (SetAccess = private, Hidden = true)
value; % Function values Mx...xN
deriv; % Jacobian values VxMx...xN
varind; % Variable indices 1xV
end
methods
function obj = autodiff(a, v, b)
... |
github | ojwoodford/ojwul-master | ojw_interp2_alt.m | .m | ojwul-master/autodiff/ojw_interp2_alt.m | 850 | utf_8 | d8690ddac7c190d31d55d7348871f54d | %OJW_INTERP2_ALT Fast 2d interpolation for images
%
% V = ojw_interp2_alt(A, X)
% V = ojw_interp2_alt(A, X, interp_mode)
% V = ojw_interp2_alt(A, X, interp_mode, oobv)
%
% Wrapper to ojw_interp2 which has horizontal and vertical coordinates
% concatenated along the third dimension into one array.
%
%IN:
% A - HxWxC do... |
github | ojwoodford/ojwul-master | grad.m | .m | ojwul-master/autodiff/grad.m | 221 | utf_8 | 401e020bd53921b653fe42f0dc0629b9 | %GRAD Dummy helper function for autodiff
function c = grad(a, vars)
if nargin < 2
c = 0;
elseif isscalar(vars) && vars < 0
c = sparse(-vars, numel(a));
else
c = zeros([numel(vars) size(a)]);
end
end
|
github | ojwoodford/ojwul-master | refine_subpixel.m | .m | ojwul-master/features/refine_subpixel.m | 1,323 | utf_8 | 2fb324b22128ffd02507ef1746c82bb5 | %REFINE_SUBPIXEL N-dimensional sub-pixel refinement
%
% [offset, val] = refine_subpixel(A, M)
%
% Computes the offsets and values of the refined positions of maxima/minima
% in an N-dimensional array, by fitting a quadratic around the points.
%
%IN:
% A - An N-dimensional array of size sz.
% M - A binary... |
github | ojwoodford/ojwul-master | fast_corners.m | .m | ojwul-master/features/fast_corners.m | 892 | utf_8 | 51b22d38c75dfc9d3eeca1cd61e2fae2 | %FAST_CORNERS Call mexed FAST corner detector
%
% [XY, scores] = fast_corners(I, thresh, type)
%
% FAST corner detection using Ed Rosten's C implementation. Method
% published in:
% "Machine learning for high-speed corner detection",
% E. Rosten & T. Drummond, ECCV 2006.
%
%IN:
% I - HxW uint8 grayscale... |
github | ojwoodford/ojwul-master | extract_features.m | .m | ojwul-master/features/extract_features.m | 1,610 | utf_8 | 5e25a2a041b3867b15db27af4864c6bb | %EXTRACT_FEATURES Extract interest points from a detector score image
%
% [X, s] = extract_features(score, [radius, [thresh, [subpixel]]])
%
% Finds the local maxima in the input detector score image.
%
%IN:
% score - HxW interest point detector score.
% radius - Scalar indicating the radius of non-maxima... |
github | ojwoodford/ojwul-master | corners.m | .m | ojwul-master/features/corners.m | 1,326 | utf_8 | cb352371b69acdedf06ea6f61d8d00b4 | %CORNERS Compute corner detector score
%
% score = corners(I, [sigma], method)
%
%IN:
% I - HxWxC image
% sigma - Scalar value determing the scale of the gradient filter.
% Default: 1.
% method - String determing the method to use: 'harris', 'noble'
% or 'shi-tomasi'. Default: 'shi-... |
github | ojwoodford/ojwul-master | dog.m | .m | ojwul-master/features/dog.m | 279 | utf_8 | a1b8d4163884aa428f3e0c43500dd4a6 | %DOG Difference of Gaussians blob detector
%
% features = dog(I)
%
%IN:
% I - HxWxC image
%
%OUT:
% features - 4xN frames of N features: [X; Y; scale; orientation].
function features = dog(I)
if size(I, 3) == 3
I = rgb2gray(I);
end
features = vl_sift(single(I));
end
|
github | ojwoodford/ojwul-master | interest_point_demo.m | .m | ojwul-master/features/interest_point_demo.m | 3,385 | utf_8 | 8516f6ee54d76a2f85743454d8f59737 | %INTEREST_POINT_DEMO Compute and visualize the interest points of an image
%
% interest_point_demo
% interest_point_demo(A, [scale, [thresh, [detector]]])
%
%IN:
% A - HxWxC image. Default: Use peppers.png
% scale - scale of the interest points to be detected (if applicable).
% Default: 2.
%... |
github | ojwoodford/ojwul-master | vl_dsift.m | .m | ojwul-master/features/vlfeat/vl_dsift.m | 5,724 | utf_8 | a5e67611f562f0b69ababcc52b234260 | % VL_DSIFT Dense SIFT
% [FRAMES,DESCRS] = VL_DSIFT(I) extracts a dense set of SIFT
% keypoints from image I. I must be of class SINGLE and grayscale.
% FRAMES is a 2 x NUMKEYPOINTS, each colum storing the center (X,Y)
% of a keypoint frame (all frames have the same scale and
% orientation). DESCRS is a 128 x... |
github | ojwoodford/ojwul-master | vl_ubcmatch.m | .m | ojwul-master/features/vlfeat/vl_ubcmatch.m | 1,416 | utf_8 | 9fa45a4773984b53f3103700b3b9b2e0 | % VL_UBCMATCH Match SIFT features
% MATCHES = VL_UBCMATCH(DESCR1, DESCR2) matches the two sets of SIFT
% descriptors DESCR1 and DESCR2.
%
% [MATCHES,SCORES] = VL_UBCMATCH(DESCR1, DESCR2) retuns the matches and
% also the squared Euclidean distance between the matches.
%
% The function uses the algorithm s... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.