Hi there,

First of all, you seem to have adopted a global binarisation algorithm which does not take into account the over-saturated area around the bottom-right corner.

Second, I assume you are interested in extracting features from those Sanskrit symbols. For each ROI (a symbol) you can extract hundreds of spatial and frequency-based features. For instance in the spatial domain you can run after the following features: Area, Eccentricity, EulerNumber, Perimeter, Solidity

Vertical and horizontal projections' signature, Pseudo-Zernike moments, etc.

Hope that helps!

Thanks Abbas, the features which are are  mentioned are GLCM features and Zernike moments. can you please give any stuff regarding the Zernike moments which i can study and apply to my dataset. 

Thanks...!

Hi.. Satish Kumar

My be this interesting for you.

Arabic Character Recognition System Development - ResearchGate

https://www.researchgate.net/.../259290862_Arabic_Character_Recognition_System_Development.

The 4th International Conference on Electrical Engineering and Informatics (ICEEI 2013). ArabicCharacter Recognition System Development. Iping Supriana.

Check out these two links, they might be helpful:

Link1

A. Tahmasbi, F. Saki, S. B. Shokouhi, Classification of Benign and Malignant Masses Based on Zernike Moments, Comput. Biol. Med., vol. 41, no. 8, pp. 726-735, 2011

Link2

All the best,

Abbas

https://se.mathworks.com/matlabcentral/fileexchange/38900-zernike-moments

http://stackoverflow.com/questions/37382723/matlab-how-to-calculate-the-pseudo-zernike-moments

...and no, what I mentioned are not GLCM (gray-level co-occurrence matrix) derived features. I assume your image is binary not gray-scale, or?

What I mentioned are rather binary region/shape properties.

thanks abbas for your valuable information and suggestions. so is it correct if i use shape features for my binary image? 

hey, Zernike Moment are good descriptors in some cases, i used them in microscopy (not as feature descriptors per say) and using them now for SLAM with satisfactory results. see my publication "3d depth variant psf analysis and interpolation using zernike moments" where you can find how to calculate them. in your case a vector of n moment modulus is enough.

Well, that all boils down to the kind of task you are undertaking. You might need to resort to more sophisticated shape descriptors if the task at hand is deemed complex.

If you can achieve your goal with a simple algorithm just go for it.

oh sorry btw , i was thinking of using discrete haar moments in place of zrnike on images ... it might be a lot better.

I guess Fourier descriptors might be useful! I'm not sure.

Thankss Elie,  i look forward to discrete haar moments.