Background distribution of multi-spectral detector output
-
摘要: 在进行防火监测、泄露监测时,被检测信息往往比较微弱。这时研究检测结果的分布规律,对检测门限的确定、统计检测和恒虚警检测等具有十分重要的作用。从统计混合模型出发,对多光谱图像CEM滤波器背景部分输出的分布规律进行了分析,认为滤波结果的概率密度函数服从高斯混合分布。并利用ETM卫星数据进行了相应的验证。结果表明实际概率密度曲线与EM算法参数估计结果吻合很好。虚警概率曲线也表明,在虚警概率Pfa>10-4条件下,采用高斯混合分布模型确定的门限是准确的。Abstract: In the detection of field fire or leakage, the target signature is very weak. Analyzing statistical models of actual spectral detector output distribution is crucial for setting detection thresholds, selecting statistical classifiers and designing constant false alarm rate detectors. Multispectral image data is processed with the CEM algorithm. Statistical model of detector output is discussed based on the stochastic mixing model. It is concluded that Gaussian mixture model is suitable. This is demonstrated with the ETM Satellite date. EM algorithm is adopted to estimate parameters of Gaussian mixture model. Computer simulation results show that the probability density function of statistics derived from Gaussian mixture model agrees with that of actual data. Curves of false alarm rate also show that Gaussian mixture model is effective at the condition of Pfa>10-4.
-
Keywords:
- target detection /
- multispectral /
- background distribution /
- Gaussian mixture model
-
-
[1]YING LI, ANTHONY VODACEK, ROBERT L. KREMENS, et al. A hybrid contextual approach to wildland fire detection using multispectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2005,43(9):2115-2126.
[2]INGEBJORG KASEN, PAL ERIK GOA, TORB-JORN SKAULI. Target detection in hyperspectral images based on multi-component statistical models for representation of background clutter[J].SPIE,2004,5612:258-264.
[3]DI M, BISCEGLIE,EPISCOPO R, GALDI C, et al. Multiband CFAR detection of thermal anomalies using principal component analysis Geoscience and Remote Sensing Symposium[J]. IEEE,2007,23:4822-4825.
[4]IRVING S REED, YU Xiao-li. Adaptive multiple-band CFAR detection of an optical pattern with unknown spectral distribution[J]. IEEE Trans on Acoust Speech Signal Process, 1990,38:1760-1770.
[5]MANOLAKIS D, ROSSACCI M, CIPAR J, et al. Statistical characterization of natural hyperspectral backgrounds using t-elliptically contoured distributions[J]. SPIE, 2005,5806:56-65.
[6]ACITO N, CORSINI G, DIANI M. A stochastic mixing model approach to sub-pixel target detection in hyper-spectral images[J]. IEEE International Conference, ICIP, 2005,1:653-656.
[7]CHANG CHEN L, LIU JIH MING, CHIEU BIN CHANG, et al. Generalized constrained energy minimization approach to subpixel target detection for multispectral imagery[J]. Optical Engineering, 2000,39(5):1275-1281.
计量
- 文章访问数: 2209
- HTML全文浏览量: 110
- PDF下载量: 716