EDGECORNER - Wrapper function for various edge/corner detection techniques.
Contents
Syntax
[Edge, Corner] = EDGECORNER(I);
[Edge, Corner] = EDGECORNER(I, met_edge, met_corner,
'Property', propertyvalue, ...);
Inputs
I : an input image with size (X,Y,C), where |C>1 when I is multichannel.
met_edge : logical variable or string setting the method used for detecting edges; it can be either:
- 'canny' or 'log' when performing Canny or log-Laplacian edge detection [GW02] with function EDGE; in that case, the detector is applied channel by channel,
- 'vista' for detecting edges likewise [PS06], a slightly variant of the Canny detector implemented in Matlab,
- 'black' for the anisotropic edge detector of [BSMH98] using the function ANISOEDGE,
- 'rothwell' for the subpixel edge detector of [RMHN95] using the function ROTHWELLEDGE,
- 'koethe' for the tensor-based edge detector of [Koth03] using the function KOETHEDGE,
- 'compass' for the compass edge/corner detector of [RT01] using the function COMPASSEDGE,
- 'elder' for the blur-based edge detector of [EZ98] using the function ELDERZUCKEREDGE,
- any of the strings 'prewitt', 'sobel', 'fast', 'derivative5', 'derivative7', 'fleck' or 'luengo' for applying the corresponding edge detectors using the function GRDSMOOTH,
- 'congrue' for the edge/corner detection of [Koves03] based on phase congruency and implemented in CONGRUENCYEDGE;
in the case met_edge is logical and false, then no edge are estimated, it is expected that corners only are detected (see variable met_corner below); in the case met_edge is logical and true, it is assigned its default value; default: met_edge='vista'.
met_corner : logical variable or string defining the method used for extracting corners; it can be either:
- 'harris' or 'noble' for Harris-Forster and Noble detections resp. using the function HARRISCORNER,
- 'susan' for Susan detection [SB97] with function SUSANCORNER,
- 'fast9', 'fast10', 'fast11', 'fast12', for any FAST corner detection using the function FASTCORNER,
- 'cpda' for curvature-based detection [ALFR09] using the function FASTCPDA;
in the case met_corner is logical and false, then no corners are extracted, it is expected that edges only are detected (see variable met_edge above); in the case met_edge is logical and true, it is assigned its default value; default: met_corner='harris'.
Note that in some cases, the implemented edge detectors (see met_edge above) also enable to extract the corners ('congrue', 'compass' or 'koethe').
Property [propertyname propertyvalues]
'rho' : post-smoothing standard deviation, used either by either of the detectors 'koethe' and 'harris' (see corresponding functions); default: rho=1.
'sig' : pre-smoothing standard deviation, used by various of the methods above.
'reduce' : logical value or string defining the way the different channels of a multispectral image are combined into the output edge map (see CANNYEDGE_BASE); it can be either:
- 'igray': the input RGB image is converted to a gray image using 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; note that reduce should be set to true when met_edge='compass'.
'int', 'samp' : parameters used for GST estimation in the method 'koethe'; see corresponding function and GRD2GST; default: int='fast' and samp=1.
Outputs
Edge : logical map of edges.
Corner : logical map of corners.
Note that if met_edge is set to false and only one lhs is required, then Corner is automatically passed in the output lhs variable (ie. output variables' positions of Edge and Corner are exchanged).
References
[Canny86] J.F. Canny: "A computational approach to edge detection", IEEE Trans. on Pattern Analysis and Machine Intelligence, 8(6):679-698, 1986. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4767851&tag=1
[Zenzo86] S. Di Zenzo: "A note on the gradient of a multi-image", Computer Vision and Graphical Image Processing, 33:116-125, 1986. http://www.sciencedirect.com/science/article/pii/0734189X86902239
[HS88] C.G. Harris and M.J. Stephens: "A combined corner and edge detector", Proc. Vision Conference, pp 147-151, 1988. http://www.bmva.org/bmvc/1988/avc-88-023.pdf
[RMHN95] C. Rothwell, J. Mundy, B. Hoffman and V.-D. Nguyen: "Driving Vision by Topology", Techn. report 2444, INRIA, 1995. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=477034&tag=1
[HSSB97] M. Heath, S. Sarkar, T. Sanocki, T. and K.W. Bowyer: "A robust visual method for assessing the relative performance of edge-detection algorithms, IEEE Trans. on Pattern Analysis and Machine Intelligence, 19(12):1338-1359, 1997. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=643893
[SB97] S.M. Smith and J.M. Brady: "{SUSAN} - A new approach to low level image processing", International Journal of Computer Vision, 23(1):45-78, 1997. http://www.lems.brown.edu/vision/courses/image-processing/Readings/smith95susan.pdf
[EZ98] J.H. Elder and S.W. Zucker: "Local scale control for edge detection and blur Estimation", IEEE Trans. on Pattern Analysis and Machine Intelligence, 20(7), 1998. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=689301&tag=1
[BSMH98] M. Black, G. Sapiro, D. Marimont and D. Heeger: "Robust anisotropic Diffusion", IEEE Trans. Image Processing, 7(3,):421-432, 1998. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=661192
[HSSB98] M. Heath, S. Sarkar, T. Sanocki, and K. Bowyer: "Comparison of edge detectors: a methodology and initial study", Computer Vision and Image Understanding, 69(1):38-54, 1998. http://www.sciencedirect.com/science/article/pii/S1077314297905877
[RT01] M. Ruzon and C. Tomasi: "Edge, junction, and corner detection using color distributions", IEEE Transactions on Pattern Analysis and Machine Intelligence, 23(11):1281-1295, 2001. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=969118
[GW02] R.C. Gonzales and R.E. Woods: "Digital Image Processing", Prentice Hall, 2002.
[Koves03] P.D. Kovesi: "Phase congruency detects corners and edges", Proc. DICTA, pp. 309-318, 2003. www.csse.uwa.edu.au/~pk/research/pkpapers/phasecorners.pdf
[Koth03] U. Kothe: "Edge and junction detection with an improved structure tensor", Proc. of DAGM Symposium, LNCS 2781, pp. 25-32, Springer, 2003. http://hci.iwr.uni-heidelberg.de/Staff/ukoethe/papers/structureTensor.pdf
[HY04] X.C. He and N.H.C. Yung: "Curvature scale space corner detector with adaptive threshold and dynamic region of support", Proc. ICPR, vol. 2, pp. 791-794, 2004. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1334377
[PS06] L. Prasad and A. Skourikhine: "Vectorized image segmentation via trixel agglomeration", Pattern Recognition, 39:501-514, 2006. http://www.sciencedirect.com/science/article/pii/S0031320305003857
[ALFR09] M. Awrangjeb, G. Lu, C.S. Fraser and M. Ravanbakhsh: "A fast corner detector based on the chord-to-point distance accumulation Technique", Proc. DICTA, pp. 519-525, 2009. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5384897
See also
EDGE, CORNER, CANNYEDGE, CANNYEDGES, CANNYEDGEPROD, CANNYEDGEMAP, SDGDEDGE, CONGRUENCYEDGE, COMPASSEDGE, ANISOEDGE, ELDERZUCKEREDGE, KOETHEDGE, PETROUEDGE, SUSANCORNER, HARRISCORNER, FASTCORNER, FASTCPDA. Requires: EDGECORNER_BASE.
Function implementation
function [Edge, Corner] = edgecorner(I, varargin)
parsing parameters
error(nargchk(1, 23, nargin, 'struct')); error(nargoutchk(0, 2, nargout, 'struct')); % mandatory parameter if ~isnumeric(I) error('edgecorner:inputerror','matrix required in input'); end p = createParser('EDGECORNER'); % principal optional parameters p.addOptional('edge', 'vista', @(x) islogical(x) || (ischar(x) && ... any(strcmpi(x,{'canny','log','vista','koethe','black','rothwell', ... 'sobel','prewitt','congrue','compass','elder', ... 'fast','derivative5','derivative7','luengo','fleck'})))); p.addOptional('corner', 'harris', @(x) islogical(x) || (ischar(x) && ... any(strcmpi(x,{'harris','noble','susan','cpda', ... 'fast9','fast10','fast11','fast12'})))); % other optional parameters p.addParamValue('sig', 0.5, @(x)isscalar(x) && isfloat(x) && x>=0); p.addParamValue('rho', 1, @(x)isscalar(x) && isfloat(x) && x>=0); p.addParamValue('reduce', false, @(x)islogical(x) || ... (ischar(x) && any(strcmpi(x,{'igray','imax','isum','gmax','eor'})))); % used by all corner detections p.addParamValue('thres', 1, @(x)(isscalar(x) || numel(x)==2) && all(x>=0)); % used for method 'Koethe' only p.addParamValue('int', 'fast', @(x)islogical(x) || (ischar(x) && ... any(strcmpi(x,{'matlab','conv','fast','ani'})))); p.addParamValue('samp', 1, @(x)isscalar(x) && round(x)==x && x>=1 && x<=5); % parse and validate all input arguments p.parse(varargin{:}); p = getvarParser(p);
check parameters
if islogical(p.edge) && ~p.edge && islogical(p.corner) && ~p.corner error('edgecorner:inputerror', ... 'at least one of the variables met_corner or met_edge must be defined'); end if islogical(p.edge) && p.edge p.edge = 'vista'; % reset to default end if islogical(p.corner) && p.corner p.corner = 'harris'; end if strcmpi(p.edge,'compass') && size(I,3)==3 && ... ~(islogical(p.reduce) && p.reduce) warning('edgecorner:inputwarning', ... ['compass not using full color information like in original method - '... 'set ''reduce'' to true for original approach']); end if strcmpi(p.edge,'congrue') && any(p.thres>1), warning('edgecorner:inputwarning', ... ['threshold should be in [0,1] ''with congrue'' method - '... 'threshold set to 0.1']); p.thres = [0.1 0.1]; end
main calculation
[Edge, Corner] = edgecorner_base(I, p.edge, p.corner, ...
p.rho, p.sig, p.thres, p.reduce, p.int, p.samp);
display
if p.disp nout = ~isempty(Edge) + ~isempty(Corner); figure, if ~isempty(Edge) subplot(1,nout,1), imagesc(Edge), axis image off, title(['edge map - method ' p.edge]); if size(Edge,3)==1, colormap gray; end end if ~isempty(Corner) subplot(1,nout,1+~isempty(Edge)), imagesc(Corner), axis image off, title(['corner map - method ' p.corner]); if size(Corner,3)==1, colormap gray; end end end
if only one lhs and Edge is empty, then exchange the variables
if nargout==1 && isempty(Edge) Edge = Corner; Corner = []; end
end % end of edgecorner