A new feature fusion method at decision level and its application

Abstract

To solve the problem that using only one feature is poor for the terrestrial environment classification, a new feature fusion method at decision level is proposed in this paper. An eigenvector is obtained by using a Gaussian mixture model (GMM) firstly, and the probabilities of the eigenvector belonging to a certain class of the texture and color are computed. Then the probabilities are multiplied by different weights according to the contribution, and summed to get the maximal likelihood probability to achieve feature fusion. Experimental results demonstrate that the method in this paper is better than other current methods and the classification performance is superior to a single feature obviously.

Keywords

Texture Feature / Discrete Cosine Transform / Gaussian Mixture Model / Color Feature / Fusion Method

References

[1] ManduchiR.. Autonomous Robots, 2005, 18: 81

[2] SchechnerY. Y., AverbuchY.. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29: 1655

[3] PietikäinenM., NurmelaT., MäenpääT.. Pattern Recognition, 2004, 37: 313

[4] SunJ., MehtaT., WoodenD.. Journal of Robotic Systems, 2007, 23: 1019

[5] ManduchiR.. Proceedings of the International Conference on Computer Vision, 1999, 2: 956

[6] ShiX. J., ManduchiR.. Image and Vision Computing, 2007, 25: 1748

[7] PermuterH., JermynJ.. Pattern Recognition, 2006, 39: 695

[8] BianconiF., FernándezA.. Pattern Recognition, 2007, 40: 3325

[9] LeungT., MalikJ.. International Journal of Computer Vision, 2001, 43: 29

[10] KimS. C., KangT. J.. Pattern Recognition, 2007, 40: 1207

[11] LiX.-w., HeP.-l., ZhangX.-r.. Journal of Optoelectronics Laser, 2008, 19: 1127

[12] YangX.-m., HeX.-h., WuW., XueL., ChenM.. Journal of Optoelectronics Laser, 2007, 18: 487