Share Email Print
cover

Proceedings Paper

BP network identification technology of infrared polarization based on fuzzy c-means clustering
Author(s): Haifang Zeng; Guohua Gu; Weiji He; Qian Chen; Wei Yang
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Infrared detection system is frequently employed on surveillance operations and reconnaissance mission to detect particular targets of interest in both civilian and military communities. By incorporating the polarization of light as supplementary information, the target discrimination performance could be enhanced. So this paper proposed an infrared target identification method which is based on fuzzy theory and neural network with polarization properties of targets. The paper utilizes polarization degree and light intensity to advance the unsupervised KFCM (kernel fuzzy C-Means) clustering method. And establish different material pol1arization properties database. In the built network, the system can feedback output corresponding material types of probability distribution toward any input polarized degree such as 10° 15°, 20°, 25°, 30°. KFCM, which has stronger robustness and accuracy than FCM, introduces kernel idea and gives the noise points and invalid value different but intuitively reasonable weights. Because of differences in characterization of material properties, there will be some conflicts in classification results. And D - S evidence theory was used in the combination of the polarization and intensity information. Related results show KFCM clustering precision and operation rate are higher than that of the FCM clustering method. The artificial neural network method realizes material identification, which reasonable solved the problems of complexity in environmental information of infrared polarization, and improperness of background knowledge and inference rule. This method of polarization identification is fast in speed, good in self-adaption and high in resolution.

Paper Details

Date Published: 8 September 2011
PDF: 9 pages
Proc. SPIE 8193, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Infrared Imaging and Applications, 81931Z (8 September 2011); doi: 10.1117/12.900239
Show Author Affiliations
Haifang Zeng, Nanjing Univ. of Science and Technology (China)
Guohua Gu, Nanjing Univ. of Science and Technology (China)
Weiji He, Nanjing Univ. of Science and Technology (China)
Qian Chen, Nanjing Univ. of Science and Technology (China)
Wei Yang, Nanjing Univ. of Science and Technology (China)


Published in SPIE Proceedings Vol. 8193:
International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Infrared Imaging and Applications
Jeffery J. Puschell; Junhao Chu; Haimei Gong; Jin Lu, Editor(s)

© SPIE. Terms of Use
Back to Top