CANNYEDGEFUSE - Two-scale Canny edge detection.

Description
Syntax
Inputs
Property [propertyname propertyvalues]
Output
Reference
Acknowledgment
See also
Function implementation

Description

Perform two-scale Canny edge detection through (logical) multiplication of the binary edge maps calculated using two Canny edge detectors at different scales.

Syntax

   edge = CANNYEDGEFUSE(I, sig2);
   edge = CANNYEDGEFUSE(I, sig2, sig1, ser, ...
                    'propertyname',propertyvalue,... );

Inputs

I : input image.

sig2, sig1 : standard-deviations of, resp., the coarsest and the finest Gaussian filters used for edge detection; default: sig2 needs always to be passed as input and sig1=1.

serad : radius of the disk-shaped structuring element used for dilating the edge map obtained at the coarsest scale; default: serad=3.

Property [propertyname propertyvalues]

'der' : string setting the name of the function used for Canny edge detection; it can be:

'vista' for the function CANNYEDGES used in Radius project,
'matlab' or 'edge' for the standard EDGE detection implemented in matlab (the same with nbin=64 instead of 128 when estimating automatically the threshold values),
'kovesi' for the function CANNY implemented by P.Kovesi;

default: der='matlab'; see CANNYEDGEMAP.

'hyst' : vector [low high] setting low and high hystheresis threshold values with 0<=low<high<1 used with der='edge'; default: hyst=[], and the threshold values are chosen automatically (see EDGE).

'reduce' : logical value or string defining the way the different channels of a multispectral image into the edge map (see CANNYEDGE_BASE); it can be either:

'igray': the input RGB image I is converted to a gray image using the function RGB2GRAY,
'imax', 'isum' : the input image (any dimension) is converted to a gray image by taking the sum and the max over the different channels resp.,
'gmax' : gradients are computed for the different channels and their local pixelwise max is given as a single input to the edge detector;
'eor' : calculations are made like for a multispectral image (as if reduce=false), but the final edge map is taken as the logical OR of the output edge maps of the different channels;

in true case, it is set to 'sum'; default: reduce=false, ie. no 'combination' is used, a multichannel map is output; this parameter is naturally ignored when I is a scalar image.

Output

edgemap : a binary edge map.

Reference

[PCPN06] G. Papari P. Campisi, N. Petkov and A. Neri: "A multiscale approach to contour detection by texture suppression", Proc. SPIE 2006 Image processing: Algorithm and Systems, vol. 6064A, 2006. http://spiedigitallibrary.org/proceedings/resource/2/psisdg/6064/1/60640D_1

Acknowledgment

This function calls the function CANNYEDGES of RADIUS project.

Function implementation

function edgemap = cannyedgefuse(I, sig2, varargin)

parsing and checking parameters

error(nargchk(1, 18, nargin, 'struct'));
error(nargoutchk(1, 1, nargout, 'struct'));

if ~isnumeric(I)
    error('cannyedgefuse:inputparameter','a matrix is required in input');
end

p = createParser('CANNYEDGEFUSE');
% function handling the function used for estimating Canny edges
p.addRequired('sig2', @(x)x>0.15);
p.addOptional('sig1', 1., @(x)isscalar(x) && isfloat(x) && x>0);
p.addOptional('serad', 3, @(x)isscalar(x) && isinteger(x) && x>1);
p.addParamValue('der', 'matlab', @(x)ischar(x) && ...
    any(strcmpi(x,{'vista','matlab','kovesi'})));
p.addParamValue('hyst', [], ...
    @(x)isvector(x) && length(x)<=2 && all(x>=0) && all(x<1));
p.addParamValue('reduce', false, @(x)islogical(x) || ...
    (ischar(x) && any(strcmpi(x,{'igray','imax','isum','gmax','eor'}))));

% parse and validate all input arguments
p.parse(sig2,varargin{:});
p = getvarParser(p);

internal variables

C = size(I,3);

% preliminary test on the existence of the function called for edge detection
if (strcmpi(p.der,'vista') && ~exist('cannyedges','file')) || ...
    (strcmpi(p.der,'kovesi') && (~exist('nonmaxsup','file') || ...
    ~exist('hysthresh','file')))
   error('cannyedgefuse:errormethod', ...
       ['function associated to ' p.der ' approach not found']);
end

% parameters hyst can be provided only with the function edge
if ~isempty(p.hyst) && strcmpi(p.der,'vista')
    warning('cannyedgefuse:inputwarning',['method ''matlab'' selected']);
    p.der = 'matlab';

elseif strcmpi(p.der,'kovesi') && (isempty(p.hyst) || length(p.hyst)<2)
     error('cannyedgefuse:errorinput', ...
       ['a vector with [low,high] hysteresis values is required' ...
         ' with ' p.der ' approach']);
end

if C==1 && ~(islogical(p.reduce) && ~p.reduce),
    warning('cannyedgefuse:inputwarning', ...
        'ignored option ''reduce'' with scalar image');
    p.reduce = false;

elseif C~=3 && ischar(p.reduce) && strcmpi(p.reduce,'igray')
    error('cannyedgefuse:inputerror', ...
         ['option ''reduce'' set to ' p.reduce ' for RGB images only']);
end

if strcmpi(p.der,'edge') && ischar(p.reduce) && strcmpi(p.reduce,'gmax')
        warning('cannyedgefuse:inputwarning', ...
            'option ''gmax'' ignored with method ''edge''');
        p.reduce = false;
end

setting variables: check that the sigma variables are set properly, ie. that sigma1<sigma2, otherwise invert their roles for furhter processing

if p.sig1>p.sig2
    tmp = p.sig1; p.sig1 = p.sig2; p.sig2 = tmp;
end

main computation

edgemap = cannyedgefuse_base(I, p.sig2, p.sig1, p.serad, p.der, p.hyst, p.reduce);

display

if p.disp
    figure, imagesc(edgemap), axis image off;
    if size(edgemap,3)==1, colormap gray;  end;
    title(['fused edge map at scales ' num2str([p.sig1 p.sig2])]);
end

end % end of cannyedgefuse