HOW to combine between SURF features and HOG features

im doing my experiment using matlab i want code thats combine between SURF features and HOG features

[abdur rahman]

hey friend,did you finish your experiment? I am also working on the same project.

[Chidiebere Ike]

Hello Ahmed, can you share the code (combination of SURF features and HOG features) to my email [e-mail]. Thank you

[chetna kaushal]

hy . can you please help me out too

Answer 1

[Dima Lisin]

Hi Ahmed,

It would help if you could tell us more about what you are trying to do.

The extractHOGFeatures function in the Computer Vision System Toolbox has two different modes. The most common mode is to compute a single HOG descriptor for the entire image. The second mode is to compute multiple HOG descriptors centered around a given set of image points, similar to what extractFeatures does for SURF or FREAK descriptors.

As to how to combine SURF and HOG, that depends on what it is you are trying to do. If you are using the bag-of-features approach for image classification, you may want to concatenate the bag-of-features histogram with a HOG vector computed over the entire image. If you want to match features, you can try concatenating individual SURF descriptors with HOG descriptors. Or you can extract SURF descriptors from one set of points, and HOG descriptors from another.

[sonia carole]

Thanks Dima Lisin for your answer I want to extract HOG features for image classification using bag-of features. How can make the concatenation between the queries HOG feature vector and database HOG feature vector

Answer 2

[swapnali ware]

can I use comination for HOG and SURF features for character recognition for better performance??

Answer 3

[SATISH KUMAR]

how to combine SURF and HOG

here is the simple solution as per my knowledge and i have used also

surf = [surf_features];

HOG = [hog_features];;

combined_Features = [surf_features,hog_features];

now you have the simple concatenated feature set.

hope this will help you.