% CLIQUEMIX Multi-image blending and swiping over a clique % % Function allows blending and swiping between any pair within a collection of images % % Usage: cliquemix(im, B, figNo, nodeLabel) % % Arguments: % im - A cell array of images to be blended. If omitted, or % empty, the user is prompted to select images via a file % dialog. % B - Parameter controlling the weighting function when % blending between two images. % B = 0 Linear transition between images (default) % B = 2^4 Sigmoidal transition at midpoint. % B = 2^10 Near step-like transition at midpoint. % figNo - Optional figure number. % nodeLabel - Optional cell array of strings specifying the labels to % be associated with the nodes on the blending interface. % % This function sets up an interface consisting of a regular polygon with nodes % corresponding to the input images to be blended. Each node is conected to % every other node forming a clique. Positioning the cursor along any of the % edges joining two nodes will generate an interactive blend formed from the % 2-image blend of the images connected by the edge. As the mouse is moved % around the interface the blend snaps to the edge closest to the mouse position % (with some hysteresis to reduce unwanted transitions). A right-click on any % edge will lock the blending to that edge irrespective of the mouse position. % A second right-click will unlock the edge. % % Hitting 's' toggles between blending and swiping mode. % % This tool allows you to compare any image with any other image in a collection % with ease. However once you go beyond about 8 images the difference in edge % lengths on the interface starts to make the tool less natural to use. % % See also: LINIMIX, BILINIMIX, TERNARYMIX, BINARYMIX, CYCLEMIX, LOGISTICWEIGHTING % Reference: % Peter Kovesi, Eun-Jung Holden and Jason Wong, 2014. % "Interactive Multi-Image Blending for Visualization and Interpretation", % Computers & Geosciences 72 (2014) 147-155. % http://doi.org/10.1016/j.cageo.2014.07.010 % Copyright (c) 2011-2014 Peter Kovesi % Centre for Exploration Targeting % The University of Western Australia % peter.kovesi at uwa edu au % % Permission is hereby granted, free of charge, to any person obtaining a copy % of this software and associated documentation files (the "Software"), to deal % in the Software without restriction, subject to the following conditions: % % The above copyright notice and this permission notice shall be included in % all copies or substantial portions of the Software. % % The Software is provided "as is", without warranty of any kind. % % August 2011 - Original version % May 2012 - General rework and cleanup % March 2014 - More tidying up % December 2014 - Incorporation of swiping option function cliquemix(im, B, figNo, nodeLabel) if ~exist('B', 'var') || isempty(B), B = 0; end % linear blending if ~exist('im', 'var'), im = []; end if ~exist('figNo', 'var') || isempty(figNo) fig = figure; else fig = figure(figNo); end [im, nImages, fname] = collectncheckimages(im); % If interface node names have not been supplied use default image names if ~exist('nodeLabel', 'var') nodeLabel = fname; end % Display 1st image and get handle to Cdata in figure(2) % Suppress display warnings for images that are large S = warning('off'); figure(fig), clf; imposition = [0.35 0.0 0.65 1.0]; subplot('position', imposition); imshow(im{1},'border', 'tight'); drawnow imHandle = get(gca,'Children'); warning(S) % Draw the interface clique and label vertices ah = subplot('position',[0.0 0.4 0.35 0.35]); % Generate vertices of blending polygon interface deltaTheta = 2*pi/nImages; theta = [0:(nImages-1)] * deltaTheta + pi/2; v = [cos(theta) sin(theta)]; % Construct list of edges that link the vertices eNo = 0; for n = 1:(nImages-1) for m = (n+1):nImages eNo = eNo+1; edge{eNo}.n1 = n; edge{eNo}.n2 = m; edge{eNo}.v1 = v(:,n); edge{eNo}.v2 = v(:,m); edge{eNo}.h = ... line([edge{eNo}.v1(1) edge{eNo}.v2(1)], ... [edge{eNo}.v1(2) edge{eNo}.v2(2)], ... 'LineWidth', 2); end end hold on plot(v(1,:), v(2,:), '.', 'color', [0 0 0], 'Markersize', 30') hold off radsc = 1.2; for n = 1:nImages if v(1,n) < -0.1 text(v(1,n)*radsc, v(2, n)*radsc, namenpath(nodeLabel{n}), ... 'FontSize', 16, 'FontWeight', 'bold',... 'HorizontalAlignment','right'); elseif v(1,n) > 0.1 text(v(1,n)*radsc, v(2, n)*radsc, namenpath(nodeLabel{n}), ... 'FontSize', 16, 'FontWeight', 'bold',... 'HorizontalAlignment','left'); else text(v(1,n)*radsc, v(2, n)*radsc, namenpath(nodeLabel{n}), ... 'FontSize', 16, 'FontWeight', 'bold',... 'HorizontalAlignment','center'); end end r = 1.2; axis ([-r r -r r]), axis off, axis equal % Set callback function and window title [rows,cols,chan] = size(im{1}); setwindowsize(fig, [rows cols], imposition); set(fig, 'WindowButtonDownFcn',@wbdcb); set(fig, 'KeyReleaseFcn', @keyreleasecb); set(fig, 'NumberTitle', 'off') set(fig, 'name', ' CliqueMix') set(fig, 'Menubar','none'); % Set up custom pointer myPointer = [circularstruct(7) zeros(15,1); zeros(1, 16)]; myPointer(~myPointer) = NaN; blending = 0; locked = 0; currentEdge = 1; blendswipe = 'blend'; % Start in blending mode fprintf('\nLeft-click to toggle blending on and off.\n'); fprintf('Right-click to toggle edge locking on and off.\n'); fprintf('A right-click will also turn blending on if it was off\n'); fprintf('Hit ''s'' to toggle between swiping and blending modes.\n\n'); %-------------------------------------------------------------------- % Window button down callback function wbdcb(src,evnt) click = get(src,'SelectionType'); % Left click toggles blending if strcmp(click,'normal') if ~blending % Turn blending on blending = 1; set(src,'Pointer','custom', 'PointerShapeCData', myPointer,... 'PointerShapeHotSpot',[9 9]) set(src,'WindowButtonMotionFcn',@wbmcb) else % Turn blending off and unlock blending = 0; locked = 0; set(src,'Pointer','arrow') set(src,'WindowButtonMotionFcn','') set(edge{currentEdge}.h, 'color', [0 0 1]); end % Right click toggles edge locking elseif strcmp(click,'alt') if blending && ~locked % Lock to closest edge e = closestedge; if e ~= currentEdge set(edge{currentEdge}.h, 'color', [0 0 1]); currentEdge = e; end set(edge{currentEdge}.h, 'color', [1 0 0]); locked = 1; elseif ~blending && ~locked % Turn blending on and lock to closest edge blending = 1; locked = 1; set(src,'Pointer','custom', 'PointerShapeCData', myPointer,... 'PointerShapeHotSpot',[9 9]) set(src,'WindowButtonMotionFcn',@wbmcb) currentEdge = closestedge; set(edge{currentEdge}.h, 'color', [1 0 0]); elseif locked % Unlock locked = 0; set(edge{currentEdge}.h, 'color', [0 0.5 0]); end end end %-------------------------------------------------------------------- % Window button move call back function wbmcb(src,evnt) cp = get(ah,'CurrentPoint'); x = cp(1,1); y = cp(1,2); [e, n1, n2, frac] = activeedge(x,y); if strcmp(blendswipe, 'blend') w = logisticweighting(frac, B,[0 1 0 1]); blend = w*im{n1} + (1-w)*im{n2}; else blend = swipe(n1, n2, frac); end set(imHandle,'CData', blend); end %--------------------------------------------------------------------- % ACTIVEEDGE % Establish the edge in the graph that is currently 'active'. Return the node % numbers, n1 and n2, that define the edge and the fractional distance of the % projected cursor point along the edge from n1 to n2 % % This version maintains a ring buffer that records the edge that the cursor % is closest to for the last N instantiations of this function. The 'active' % edge is set to the edge number that appears most frequently in the ring % buffer (the mode). Seems to work quite well however the bahaviour is % dependent on the speed of the computer. It would be good to have something % that was tied to actual time. function [e, n1, n2, frac] = activeedge(x,y) N = 30; % Buffer size persistent ebuf; % Buffer for recording which edge we have been on persistent ptr; % Pointer into buffer. if isempty(ebuf), ebuf = currentEdge*ones(1,N); end if isempty(ptr), ptr = 0; end if locked v1 = edge{currentEdge}.v1; v2 = edge{currentEdge}.v2; [eDist, pfrac] = dist2segment(v1, v2, x, y); e = currentEdge; n1 = edge{currentEdge}.n1; n2 = edge{currentEdge}.n2; frac = 1-pfrac; else % Not locked. Find edge closest to x,y eDist = zeros(1,length(edge)); pfrac = zeros(1,length(edge)); for n = 1:length(edge) v1 = edge{n}.v1; v2 = edge{n}.v2; [eDist(n), pfrac(n)] = dist2segment(v1, v2, x, y); end [minDist, ind] = min(eDist); ptr = mod(ptr+1,N-1); ebuf(ptr+1) = ind; % Set edge to be the one which is most common within ebuf e = mode(ebuf); if e ~= currentEdge set(edge{currentEdge}.h, 'color', [0 0 1]); currentEdge = e; end set(edge{currentEdge}.h, 'color', [0 0.5 0]); % Identify the indices of the images associated with this edge and % also calculate the fractional position we are along this edge for % the blending n1 = edge{e}.n1; n2 = edge{e}.n2; frac = 1-pfrac(e); end end % of activeedge %------------------------------------------------------------------------ % Find edge closest to cursor function e = closestedge cp = get(ah,'CurrentPoint'); x = cp(1,1); y = cp(1,2); eDist = zeros(1,length(edge)); for n = 1:length(edge) v1 = edge{n}.v1; v2 = edge{n}.v2; eDist(n) = dist2segment(v1, v2, x, y); end [~, e] = min(eDist); end %------------------------------------------------------------------------ % Given an x,y coordinate and a line segment with end points v1 and v2. Find the % closest point on the line segment. Return the distance to the line segment % and the fractional distance of the closest point on the segment from v1 to v2. function [d, frac] = dist2segment(v1, v2, x, y) xy = [x;y]; v1v2 = v2-v1; % Vector from v1 to v2 D = norm(v1v2); % Distance v1 to v2 v1xy = xy-v1; % Vector from v1 to xy % projection of v1xy on unit vector v1v2 proj = dot(v1v2/D, v1xy); if proj < 0 % Closest point is v1 d = norm(v1xy); frac = 0; elseif proj > D % Closest point is v2 d = norm(xy-v2); frac = 1; else % Somewhere in the middle pt = v1 + proj*v1v2/D; d = norm(xy-pt); frac = proj/D; end end % of dist2segment %------------------------------------------------------------------------ % Function to set up window size nicely function setwindowsize(fig, imsze, imposition) screen = get(0,'ScreenSize'); scsze = [screen(4) screen(3)]; window = get(fig, 'Position'); % Set window height to match image height window(4) = imsze(1); % Set window width to match image width allowing for fractional width of % image specified in imposition window(3) = imsze(2)/imposition(3); % Check size of window relative to screen. If larger rescale the window % size to fit 80% of screen dimension. winsze = [window(4) window(3)]; ratio = max(winsze./scsze); if ratio > 1 winsze = winsze/ratio * 0.8; end window(3:4) = [winsze(2) winsze(1)]; set(fig, 'Position', window); end %----------------------------------------------------------------------- % Generate swipe image given the two end nodes of the active edge and the % fractional distance along the edge. function swipeim = swipe(n1, n2, frac) % Form vectors from the centre of the image to the corners and construct % the dot product between v(:,n2)-v(:,n1) to determine the extreme % corners of the image that are relevant corner = [1 cols cols 1 1 1 rows rows]; % Vectors from centre of image to top-left, top-right, bottom-right, % bottom-left corners. c = corner - repmat([cols/2; rows/2], 1, 4); % Vector from node 1 to node 2 d = v(:,n2) - v(:,n1); d(2) = -d(2); % Negate y to match image coordinate frame d1 = d/norm(d); % Find c2, the index of corner direction maximally alligned with d dotcd = c'*d; % Dot product between c and d [~,c1] = max(dotcd); c2 = mod((c1-1)-2, 4) + 1; % Opposite corner to c2 c1c2 = corner(:,c2)-corner(:,c1); % Compute distance from c1 to c2 in the direction of d dist = dot(c1c2, d1); % Construct location that is 'frac' units along line parallel to % v(:,n2)-v(:,n1) between the two extreme image corners % This has to be frac*dist from c1 to c2 p = corner(:,c1) + frac*dist*d1; % Cut the image in two at the point along a line perpendicular to % v(:,n2)-v(:,n1) and construct a mask for defining the image parts to be % combined. % Form equation (p - [c;r]).d % mask is where this value is > 0 giving all image points closest to n2 [x,y] = meshgrid(1:cols, 1:rows); x = p(1) - x; y = p(2) - y; mask = (x*d(1) + y*d(2)) < 0; swipeim = im{n2}; if chan == 1 swipeim(mask) = im{n1}(mask); else % Colour swipeim = zeros(rows,cols,chan); for ch = 1:chan swipeim(:,:,ch) = im{n1}(:,:,ch) .* mask + im{n2}(:,:,ch) .* ~mask; end end end %----------------------------------------------------------------------- % Key Release callback % If 's' is pressed system toggles between swiping and blending modes function keyreleasecb(src,evnt) if evnt.Character == 's' % Swipe/Blend toggle if strcmp(blendswipe, 'swipe') blendswipe = 'blend'; else blendswipe = 'swipe'; end end wbmcb(src,evnt); % update the display end %------------------------------------------------------------------------ end % of everything