Calculating mean for lines generated from Hough Transform

Hello everyone,

I'm trying to determine the average RGB colour of lines detected in an image through a Hough transform. I have calcuated the mean of each individual line successful, but I want the mean of all of the lines that are found. The amount of lines could change each time.

for k = 1:length(lines)
   xy = [lines(k).point1; lines(k).point2];
   plot(xy(:,1),xy(:,2),'LineWidth',2,'Color','green');
   % Plot beginnings and ends of lines
   plot(xy(1,1),xy(1,2),'x','LineWidth',2,'Color','yellow');
   plot(xy(2,1),xy(2,2),'x','LineWidth',2,'Color','red');
   % find average road colour - get all co-ords between endpoints
   [x, y]=bresenham(xy(1,1),xy(1,2),xy(2,1),xy(2,2));
   LineCoOrds{k}=[x, y]; 
   % Get RGB values for these points & their mean
   P{k} = impixel(I,x,y);
   Meanpix{k}=mean(P{k});
end

I have tried methods such as below, but havent got it quite right.

TotalMeanPixel=cellfun(@mean, Meanpix);

I'm fairly new to Matlab so any help is appreciated a lot.

EDIT - Just to clarify, I want the mean of each individual column, e.g. first column, for all the cells in Meanpix, so I ended up with one value for RGB. I need the average colour of all the lines detected.

Many thanks!

Liz

Answer 1

[Paul]

Maybe something like this is what you need:

    mean_vec=[];
    for ii=1:length(Meanpix)
        x=cell2mat(Meanpix{ii});
        col1 = x(:,1);
        mean_vec=[mean_vec; col1];
    end

    TotalMeanPixel=mean(mean_vec)

[Liz]

Thanks a lot for your help, that works for calculating the R component of RGB, but I also need the other 2. I originally tried just following the same example for G and B but at the end it gave me one mean value, whereas I need a 1x3 output. I'm now trying something similar to this but I get 'Subscripted assignment dimension mismatch.'

mean_vec=[];
for ii=1:length(Meanpix)
    x=cell2mat(Meanpix(ii));
    col1 = x(:,1);
    mean_vec(:,1)=[mean_vec; col1];
    col2 = x(:,2);
    mean_vec(:,2)=[mean_vec; col2];
    col3 = x(:,3);
    mean_vec(:,3)=[mean_vec; col3];
end
TotalMeanPixel(:,1)=mean(mean_vec(:,1));
TotalMeanPixel(:,2)=mean(mean_vec(:,2));
TotalMeanPixel(:,3)=mean(mean_vec(:,3));

Any help is greatly appreciated! If I work it out I'll post it and give you 'answered' :)

[Liz]

I believe this code works (I calculated by hand too). Thanks for your help!

mean_vec1=[];
mean_vec2=[];
mean_vec3=[];
for ii=1:length(Meanpix)
    x=cell2mat(Meanpix(ii));
    col1 = x(:,1);
    mean_vec1=[mean_vec1; col1];
    col2 = x(:,2);
    mean_vec2=[mean_vec2; col2];
    col3 = x(:,3);
    mean_vec3=[mean_vec3; col3];
end
TotalMeanPixelR=mean(mean_vec1);
TotalMeanPixelG=mean(mean_vec2);
TotalMeanPixelB=mean(mean_vec3);

[Paul]

You can do this too which is easier:

    mean_mat=[];
    for ii=1:length(Meanpix)
        x=cell2mat(Meanpix(ii));        
        mean_mat=[mean_mat; x];
    end
    TotalMeanPixel=mean(mean_mat);
    R_mean=TotalMeanPixel(1);
    G_mean=TotalMeanPixel(2);
    B_mean=TotalMeanPixel(3);

[Liz]

Even better, thanks a lot!