SVM classifier for image classification

I am currently working on an image processing project.I have trained SVM classifier to classify paddy leaf images into 3 categories.What should I do to so that any leaf image other than those corresponding to these three categories when given as input give a result that it doesnt belong to these 3 categories?

Answer 1

[Image Analyst]

First of all I'd convert your image into an N-by-3 matrix.

rgbImage = imread('peppers.png');
[r, g, b] = imsplit(rgbImage);
rgbValues = [r(:), g(:), b(:)];

Then I'd use Classification Learner on the Apps tab of the tool ribbon. Tell it to start a new session using your rgbValues from the workspace. Then train it with the SVM classifier of your choice.

Finally, have the Classification Learner export the code to an m-file.

I'm attaching various demos that you might find interesting.

[SWATHY Kakkooth]

My current SVM classifier works well in classifying rice leaf image into 3 categories. My problem is if any other image other than those belonging to these three is inputted how could I give a result that this leaf doesnt belong to those 3 categories?