A Novel Feature Extraction Method for Unexploded Ordnance in Ultra-Wideband SAR

Morikawa M, Taylor S, and Persons M. Deaths and injuriesdue to unexploded ordnance (UXO) in northern Lao PDR(Laos)[J].Injury.1998, 29(4):301-304[2]Billings S D. Discrimination and classification of buriedunexploded ordnance using magnetometry[J].IEEE Trans. onGeoscience and Remote Sensing.2004, 42(6):1241-1251[3]Marr B, Torrione P, and Miller J, et al.. Parameterizedlikelihood ratio method for EMI unexploded detection[J].Proceeding of SPIE.2004, 5415:843-854[4]Zhang Y, Collins L, and Carin L. Physics model basedunexploded ordnance discrimination using wideband EMIdata[J].Proceeding of SPIE.2003, 5089:1023-1034[5]Carin L, Geng N, and McClure M, et al.. Wide-area detectionof land mines and unexploded ordnance[J].Inverse Problems.2002, 18(3):575-609[6]Dong Y, Runkle P R, and Carin L, et al.. Multi-aspectdetection of surface and shallow-buried unexploded ordnancevia ultra-wideband synthetic aperture radar[J].IEEE Trans. onGeoscience and Remote Sensing.2001, 39(6):1259-1270[7]Jin T, Zhou Z, and Chang W. Ultra-wideband SARtime-frequency representation image formation[J].IEEProceedings: Radar, Sonar and Navigation.2006, 153(5):389-395[8]Jin T, Zhou Z M, and Chang W G, et al.. Aspect-invariantfeature extraction and associated landmine detector in UWBSAR. CIE International Conference on Radar, Shanghai,China, 2006: 607-610.[9]Hu M K. Visual pattern recognition by moment invariants.IRE Trans. Information Theory, 1962, 8(2): 179-187.[10]Gonzalez R C and Woods R E. Digital Image Processing.Second Edition, Englewood Cliffs, NJ: Prentice Hall, 2002,ch.11.