Seam Carving for content aware image resizing: GUI implementation demo: SeamVector=findSeam(x); - File Exchange

Code covered by the BSD License

Highlights from
Seam Carving for content aware image resizing: GUI implementation demo

SeamCutGUI(varargin) SEAMCUTGUI M-file for SeamCutGUI.fig
Emean=findEnergy(x) FINDENERGY creates an gradient img from a given RGB or grayscale image.
SeamImg=findSeamImg(x) FINDSEAMIMG finds the seam map from which the optimal (vertical running)
SeamVector=findSeam(x); FINDSEAM returns a column vector of coordinates for the pixels to be
SeamVector=removalMap(X,lines... REMOVALMAP takes a given image and finds the ordered set of (vertical)
show(I,M) show - show content aware image resizing
x=SeamCut(x,SeamVector) SEAMCUT takes as input a RGB or grayscale image and SeamVector array to
x=SeamPlot(x,SeamVector) SEAMPLOT takes as input an image and the SeamVector array and produces
x=SeamPut(x,SeamVector) SEAMPUT takes as input a RGB or grayscale image and SeamVector array to
SandboxScript.m
View all files

from Seam Carving for content aware image resizing: GUI implementation demo by Danny Luong
Complete seam carving algorithm implementation with GUI demo

SeamVector=findSeam(x);

function SeamVector=findSeam(x);
% FINDSEAM returns a column vector of coordinates for the pixels to be
% removed (the seam), given the SeamImage resulting from findSeamImg.m
% each element "j" of SeamVector is the column number of the pixel in 
% row i to be removed.
%
% Author: Danny Luong
%         http://danluong.com
%
% Last updated: 12/20/07


[rows cols]=size(x);
for i=rows:-1:1
    if i==rows
        [value, j]=min(x(rows,:));  %finds min value of last row
    else    %accounts for boundary issues
        if SeamVector(i+1)==1
            Vector=[Inf x(i,SeamVector(i+1)) x(i,SeamVector(i+1)+1)];
        elseif SeamVector(i+1)==cols
            Vector=[x(i,SeamVector(i+1)-1) x(i,SeamVector(i+1)) Inf];
        else
            Vector=[x(i,SeamVector(i+1)-1) x(i,SeamVector(i+1)) x(i,SeamVector(i+1)+1)];
        end
        
        %find min value and index of 3 neighboring pixels in prev. row.
        [Value Index]=min(Vector);
        IndexIncrement=Index-2;
        j=SeamVector(i+1)+IndexIncrement;
    end
    SeamVector(i,1)=j;
end

Highlights from Seam Carving for content aware image resizing: GUI implementation demo

Highlights from
Seam Carving for content aware image resizing: GUI implementation demo