Multiple Features Fusion Based on PCA in Liver CT Image Recognition

Abstract:

The contribution of various features for image recognition is different. In order to comprehensively and complementarily express the contents of liver CT image, a method of multiple features fusion based on PCA (Principal Component Analysis) is proposed in this paper. First, the features of each medical image include color, texture and shape are extracted, which constitute a feature vector that can represent the image. Then, for the wide gap of the value among the features, the features are standardized. Finally, the method of PCA is used to fuse the standardized features which can reduce the dimension and eliminate the redundant information among the features. The results through the recognition based on SVM (Support Vector Machine) proves that the fused features can express the contents of the liver CT image better and a higher recognition rate can be got in the image recognition.

• Periodical: Applied Mechanics and Materials (Volumes 484-485)
• Pages: 916-921
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.484-485.916
• Online since: January 2014
• Authors: Peng Peng Jiao, Yi Zheng Guo
• Keywords: Feature Fusion, Feature Standardization, Medical Image Recognition, Principal Component Analysis (PCA)

References

[1] Tarjoman, Mana; Fatemizadeh, Emad; Badie, Kambiz. An implementation of a CBIR system based on SVM learning scheme. Journal of Medical Engineering and Technology, 2013, 37(1): 43-47.

DOI: 10.3109/03091902.2012.742157

[2] Ahmad, Wan Siti Halimatul Munirah Wan; Fauzi, Mohammad Faizal Ahmad. Comparison of different feature extraction techniques in content-based image retrieval for CT brain images. Proceedings of the 2008 IEEE 10th Workshop on Multimedia Signal Processing, MMSP 2008, pp.503-508, (2008).

DOI: 10.1109/mmsp.2008.4665130

[3] Kludas, Jana; Bruno, Eric; Marchand-Maillet, Stephane. Exploiting document feature interactions for efficient information fusion in high dimensional spaces. 2008 1st International Workshops on Image Processing Theory, Tools and Applications, IPTA (2008).

DOI: 10.1109/ipta.2008.4743798

[4] Tahoun, Mohamed A; Nagaty, Khaled A; El-Arief, Taha I. A robust content-based image retrieval system using multiple features representations. Networking, Sensing and Control, v2005, pp.116-122, (2005).

DOI: 10.1109/icnsc.2005.1461172

[5] Yu, Sheng; Zeng, Jie-xian; Xie, Li. Image retrieval algorithm based on the multiple features fusion. Computer Engineering, 2012, 38(24): 216-224.

[6] Kludas Jana, Bruno Eric, Marchand-Maillet Stéphane. Information fusion in multimedia information retrieval. Lecture Notes in Computer Science, 2008(4918), 147-159.

DOI: 10.1007/978-3-540-79860-6_12

[7] Song, Yu-qing; Chen, Jian-mei; Guo, Yi-zheng. Medical image recognition Research based on multi-feature fusion. Application Research of computer. 2008, 25(6): 1750-1752.

[8] Wang, Da-wei; Chen, Ding-rong; He, Yi-zheng. A survey of feature-level image fusion based on target recognition. Avionics Technology, 2011, 42(2): 6-12.

[9] Gadelmawla, E.S. A vision system for surface roughness characterization using the gray level co-occurrence matrix. NDT and E International, 2004, 37(7): 577-588.

DOI: 10.1016/j.ndteint.2004.03.004