 |
 |
 |
 |
 |
 |
Sains
Malaysiana 38(2): 249-260(2009)
Penyaringan dan Pemilihan Fitur Statistik Asas untuk Pengecaman Spesimen Forensik Balistik (Extraction and Selection of
Basic Statistical Features for
Forensic Ballistic Specimen
Identification)
Nor Azura Md Ghani
Pusat Pengajian Statistik
Fakulti Teknologi Maklumat dan Sains Kuantitatif
Universiti Teknologi MARA, 40450 Shah Alam, Selangor D.E..
Liong Choong-Yeun & Abdul Aziz Jemain
Pusat Pengajian Sains
Matematik, Fakulti Sains dan Teknologi
Universiti Kebangsaan
Malaysia, 43600 UKM Bangi, Selangor D.E.
Diserahkan:
12 Mac 2008 / Diterima: 16 Julai 2008
ABSTRAK
Pengecaman senjata api
semakin serius dan amat penting di dalam bidang penyiasatan jenayah khususnya
untuk tempoh dua dekad ini. Dalam makalah
ini, suatu sistem untuk pengecaman senjata api berasaskan tapak kelongsong
peluru telah diperkenalkan. Tapak kelongsong peluru adalah satu daripada
penunjuk yang amat penting dalam membantu menyelesaikan masalah pengecaman
pistol. Peluru yang telah digunakan akan
meninggalkan lebih daripada 30 ciri yang amat berharga pada tapak kelongsong
bagi membantu pihak tertentu mengecam pistol yang telah diguna. Ciri-ciri
tersebut sebagai suatu gabungan membentuk apa yang dikenali sebagai kesan cap
jari bagi pistol. Lantaran itu, kajian
ini adalah untuk mendapatkan fitur yang sesuai bagi pengecaman senjata
api. Terlebih dahulu imej tapak
kelongsong peluru ini telah ditemberengkan kepada tiga bahagian, membentuk tiga
set imej yang berlainan. Imej-imej ini juga dilakukan prapemprosesan untuk
membentuk tiga set imej lagi. Fitur-fitur seterusnya disaring daripada imej asal tapak kelongsong
peluru dan imej yang telah melalui prapemprosesan. Dua puluh fitur yang berbeza secara
signifikan telah diperoleh dan dikirakan untuk imej-imej asal dan yang telah
dilakukan prapemprosesan. Kesemua
pemprosesan telah dilakukan menggunakan pengaturcaraan MATLAB. Suatu skim
berdasarkan analisis korelasi seterusnya telah diperkenalkan untuk pencarian
fitur berdasarkan konsep meminimumkan lewahan data tetapi memastikan ciri-ciri
unik tersimpan. Fitur-fitur yang
berkorelasi tinggi akan digugurkan pasangannya dan hasilnya tinggal cuma tujuh
fitur sahaja. Ketujuh-tujuh fitur ini
telah diuji sebagai suatu vektor fitur untuk mengelaskan lima pistol daripada
model yang sama menggunakan analisis diskriminan. Hasil pengelasan menunjukkan lebih 80%
daripada imej-imej tapak kelongsong peluru itu telah dikelaskan dengan jayanya.
Kata kunci: Analisis korelasi; analisis diskriminan; fitur;
imej tapak kelongsong peluru; senjata api
ABSTRACT
Firearms identification has been getting very important in crime
investigation in the last two decades. In this paper, a recognition system for firearms identification based on
cartridge case image is introduced. Cartridge case is one of the important clues towards solving the gun
file. There are more than 30 marks left
on the surface of the cartridge case when a gun is fired which are invaluable
towards identifying the firearm used. These marks in combination produces a “fingerprint” for identification
of a firearm. Therefore, the aim of this
research work is towards extraction and identification of suitable features for
firearms recognition. Firstly the cartridge case images are segmented into
three parts, forming three sets of images. These images were also preprocessed
to form three more sets of images. Features were extracted from these original
and preprosessed. Twenty significant
features each were identified and computed for the original and the
preprocessed images. All processing were done using MATLAB programming. A
scheme based on correlation analysis were introduced towards features selection
based on the concept of minimising data redundancy but maximising classes’
differences. Features that are highly correlated were dropped and eventually
there are only seven significant features. The seven features formed a feature
vector for the fireams recognition and were tested on five pistols of the same
model using discriminant analysis. The classification results show that more
than 80% of the cartridge case images were classified correctly.
Keywords: cartridge case image; correlation analysis;
discriminant analysis; firearms; feature
RUJUKAN
Burrard,
G. 1951. The Identification of Firearms and Forensic Ballistics. London: Herbert Jenkins.
Chapman,
S.J. 2004. MATLAB Programming for
Engineers. Edisi ke-3. Sydney: Thomson Learning.
Coakes, S.J.,
Steed, L. & Dzidic, P. 2006. SPSS Version 13.0 for Windows: Analysis
Without Anguish. Milton: John Wiley
& Sons.
French, A
& Poulsen, J. 2002. Multivariate
analysis of variance (MANOVA). [atas talian] http://online.sfsu.edu/~efc/classes/biol710/manova/manova.htm (21 Ogos 2006).
Geradts, Z.,
Bijhold, J. & Hermsen, R. 1999. Pattern recognition in a database of
cartridge cases. Proceedings
SPIE: Investigation and Forensic Science
Technologies 3576:104-105.
Geradts, Z.,
Bijhold, J., Hermsen, R. & Murtagh, F. 2001. Image matching algorithms
for breech face marks and firing pins in a database of spent cartridge cases of
firearms. Forensic Science
International 119(1): 97-106.
Gonzales,
R.C., Wood, R.E. & Eddins, S.L. 2004. Digital
Image Processing Using MATLAB. Upper
Saddle River, NJ: Pearson Prentice-Hall.
Gonzales,
R.C. & Wood, R.E. 2002. Digital
Image Processing. Edisi ke-2. Upper Saddle River, NJ: Prentice-Hall.
Haddad, R.A.
& Parsons, T.W. 1991. Digital Signal Processing: Theory, Applications and Hardware. Upper Saddle River, NJ: W.H. Freeman and Co.
Hair, J. F.,
Black, W. C., Babin, J. B., Anderson, R.
E. & Tatham, R. L. 2006. Multivariate
Data Analysis. Edisi ke-6. Upper
Saddle River, NJ: Pearson Prentice Hall.
Hall, M.
A. 1999. Correlation-based Feature Selection for Machine Learning. Disertasi doktor falsafah yang tidak
diterbitkan. University of Waikato, New
Zealand.
Heard,
B.J. 1997. Handbook of Firearms and Ballistics:
Examining and Interpreting Forensic Evidence. Chichester: John Wiley & Sons Ltd.
Jain, A.K.
1989. Fundamentals of Digital Image
Processing. Upper Saddle River, NJ:
Prentice-Hall.
Jahne, B. 1997. Digital
Image Processing: Concepts, Algorithms and Scientific Applications. Edisi
ke-4. Berlin: Springer-Verlag.
Johnson, R.A.
& Wichern, D.W. 2002. Applied Multivariate
Statistical Analysis. Edisi
ke-5. Upper Saddle River, NJ: Prentice Hall.
Kou, C., Tung, C.-T. & Fu, H.C. 1994. FISOFM:
Firearms Identification based on SOFM Model of Neural Network. Proceedings of the Institute of Electrical
and Electronics Engineers 28th Annual International Carnahan
Conference on Security Technology, 12 – 14 Oktober: 120-125.
Lattin, J., Carroll. J.D. & Green. P.E. 2003. Analyzing Multivariate Data. Toronto: Thomson
Learning.
Liong,
C.-Y. 2001. Investigation of Vision Based System for
Passenger Position Recognition. Disertasi doktor falsafah yang tidak diterbitkan. Cranfield University, United Kingdom.
Manly, B. 1986. Multivariate Statistical
Methods: A Primer. London: Chapman
and Hall.
MathWorks. 2003. Image
Processing Toolbox for Use with MATLAB: User’s Guide Version 4. Cetakan ke-4. The MathWorks, Inc.
Morrison,
D.F. 2005. Multivariate Statistical
Methods. Edisi ke-4. Southbank: Thomson Learning.
Smith, C.L.
& Cross, J.M. 1995. Optical Imaging Techniques for Ballistics
Specimens to Identify Firearms. Proceedings
of the IEEE International Carnahan Conference on Security Technology, Oktober:
275-289.
Smith,
C.L. 1997. Fireball: A Forensic Ballistics Imaging
System. Proceedings of The
Institute of Electrical and Electronics Engineers 31st Annual
International Carnahan Conference on Security Technology, 15–17
Oktober: 64–70.
Smith, C.L.
2001. An Intelligent Imaging Approach to
the Identification of Forensic Ballistics Specimens. Proceedings of IEEE International
Conferences on Info-tech and Info-net 3: 390-396.
S-PLUS. 2002. S-PLUS Essentials II: The Command
Line. Seattle: Insightful
Corporation.
Swiniarski,
R. W. & Skowron, A. 2002. Rough Set Methods in Feature Selection and
Recognition. Pattern Recognition 24: 833 – 849.
Wulder, M. 2005. Multivariate Statistics: Multicollinearity
and Singularity. [atas talian] http://www.pfc.cfs.nrcan.gc.ca/profiles/wulder/mvstats/multicol_e.html (4 April 2008).
Xin, L.-P.,
Zhou, J. & Rong, G. 2000. A Cartridge Identification System for Firearms Authentication. Signal Processing
Proceedings, IEEE 21-25 Ogos: 1405-1408.
Zhang, M.
& Yao, J.T. 2004. A Rough Sets Based Approach to Feature
Selection. Processing NAFIPS ’04,
IEEE Annual Meeting of the Fuzzy Information, 27 – 30 June, 1: 434 –
439.
Zhou, J.,
Xin, L.-P., Gao, D.-S., Zhang, C.-S. & Zhang, D. 2001. Automated Identification for Firearms Authentication. Proceedings of the IEEE Computer Society
Conference 1: 749-754.
|
|||
|
