Abstract—In this paper, we proposed the new texture features called Haralick in Contourlet Sub-bands based Feature- HCSF-for 2D bone structure evaluation. Our proposed bone texture features are built on co-occurrences matrix for outside wedges and Gabor features for coarsest level of contourlet transform. The classification experiments are tested with linear SVM engine and the precision ratio is 70%. The proposed method is also compared to SFTA algorithm, Haralick, and Gabor features. The classification precision of proposed method is always higher than those in the related methods.
Index Terms—Bone texture characterization, contourlet, gabor transform.
Viet Quoc Ngo is with University of Education, HCMC, Vietnam (email: thanhbinhsp@gmail.com).
Thuc Nguyen Dinh is with Hochiminh City University of Natural Science, HCMC, Vietnam (e-mail: ndthuc@fit.hcmus.edu.vn).
[PDF]
Cite:Viet Quoc Ngo and Thuc Nguyen Dinh, "Bone Texture Characterization Based on Contourlet and Gabor Tranforms," International Journal of Computer Theory and Engineering vol. 8, no. 2, pp. 177-181, 2016.