Concatenate SURF features and Radon Features to train SVM

Question

Is it possible to concatenate 2 bagoffeatures objects to train a classifier?

I have trained a classifier using SURF points by the following :

extractorFcn = @SURFBOW;
bag = bagOfFeatures(trainingSets,'CustomExtractor',extractorFcn);

where SURFBOW contains :

[height,width,numChannels] = size(I);
if numChannels > 1
    grayImage = rgb2gray(I);
else
    grayImage = I;
end
multiscaleSURFPoints = detectSURFFeatures(grayImage,'MetricThreshold',100);
features = extractFeatures(grayImage, multiscaleSURFPoints,'Upright',true);
featureMetrics = multiscaleSURFPoints.Metric;

and followed Matlab's example : http://www.mathworks.com/help/vision/examples/image-category-classification-using-bag-of-features.html?refresh=true

Next, i have done something similar to extract the Radon Features of the image instead using another extractor function but with RadonBOW(I) as follows :

[height,width,numChannels] = size(I);
if numChannels > 1
    grayImage = double(rgb2gray(I));
else
    grayImage = double(I);
end



dx = imfilter(grayImage,fspecial('sobel') ); % x, 3x3 kernel
dy = imfilter(grayImage,fspecial('sobel')'); % y
gradmag = sqrt( dx.^2 + dy.^2 );

% mask by disk
R = min( size(grayImage)/2 ); % radius
disk = insertShape(zeros(size(grayImage)),'FilledCircle', [size(grayImage)/2,R] );
mask = double(rgb2gray(disk)~=0);
gradmag = mask.*gradmag;

% radon transform
theta = linspace(0,180,179);
vars = zeros(size(theta));
for u = 1:length(theta)
    [rad,xp] =radon( gradmag, theta(u) );
    indices = find( abs(xp)<R );
    % ignore radii outside the maximum disk area
    % so you don't sum up zeroes into variance
    vars(u) = var( rad( indices ) );
end
features = vars/norm(vars);
featureMetrics = var(features);

I receive fair results with each. is there anyway to combine these to train the classifier using both the radon and SURF points?

(I have also tried to do the Radon BOW method manually by use of kmeans, however, i got extremely poor results, so I believe that was incorrect)

Thank you!

Dima Dima · Accepted Answer · 2016-03-03T20:21:29

When you use bagOfFeatures you call the encode method, which takes an image an returns a bag-of-features histogram. You can concatenate that histogram with your Radon features for the same image, and train on that.

Concatenate SURF features and Radon Features to train SVM

1 Answers