Preparation GUIDE for Final-Submission Manuscript An Automatic PCB matching system That is insensitive to image scaling or mirroring
Author Name Author Department Author Institute City, ST zipcode Tel : 123-456-7890 Fax : 123-456-0987 e-mail :
[email protected]
(MS(MS-WORD 2000 version) Coauthor Name Coauthor Department Coauthor Institute City, ST zipcode Tel : +81-3-333-1234 Fax : +81-3-333-5678 e-mail :
[email protected].
Abstract - A forensic tool has been developed that allows a software or hardware system to be built that is capable of probing a large database of digital images of circuit boards and compare them for similarities to provide investigation leads for electronic crimes digital forensic science investigations. The developed system has been simulated and proved to be very efficient in detection similarities between a target image and a large image database even when the target image is noisy, scaled or mirrored.
I Introduction Digital forensics [13], also referred to as computer forensic analysis, electronic discovery, electronic evidence discovery, digital discovery, data recovery, data discovery, computer analysis, and computer examination, is the process of methodically examining computer media (hard disks, diskettes, tapes, etc.) for evidence. A thorough analysis by a skilled examiner can result in the reconstruction of the activities of a computer user. Unfortunately, crime methods evolve with technology. New forms of theft started to appear in the industrial evolution era where some companies steal the innovations and designs of their competitors and start manufacturing these products for their own interest without licensing or benefiting the original innovators or manufacturers. This work created a scientific tool (system) algorithm that can aid the investigation of electronic design crimes. A
technique has been developed that allows a software or hardware tool to be built that is capable of probing a large database of digital images of circuit boards and compare them for similarities to provide investigation leads for digital forensic science investigations. The PCB image matching system is composed of three components: • Storage component, • Target image processing component and • Query component. The matching system was implemented in Matlab with the following characteristics: • A library of 100 images was constructed. • The image enhancement block was implemented as function that could call any image enhancement method of choice (Image enhancement algorithm 8 was used to implement this block). • The feature vector extraction method was implemented in 5 different methods: 1. FFT feature vectors 2. FFT feature vectors with invariant moments in the FFT domain 3. DWT feature vectors 4. DWT feature vectors with invariant moments in the DWT domain 5. Image invariant moments combined with DWT feature vectors with invariant moments in the DWT domain The simulation results showed that the system with Image
invariant moments combined with DWT feature vectors with invariant moments in the DWT domain gave the best results that are insensitive to image mirroring or scaling..
1. Discreet Fourier transform (DFT). 2. Discreet Wavelet transform “db4” (DWT) E.
II. System Details A. Block Diagram The following Figure shows the block diagram for the proposed system solution:
Feature Vector Extraction
The feature vectors of choice are calculated off-line and stored for each image so efficient computation is not a critical criterion for the matching process in the query stage. The following methods for feature vectors were used: 1. Statistics of the FFT magnitude coefficients. 2. Invariant moments of the FFT magnitude coefficients. 3. Statistics of the DWT decomposition coefficients. 4. Invariant moments of the DWT decomposition coefficients. 5. Invariant moments of the image spatial domain data combined with invariant moments of the DWT decomposition coefficients. TABLE I Fonts for Camera-Ready Papers
Figure 1– System block diagram
The PCB image matching system is composed of
Font Size
Style
Text
14pt
bold
Paper title
12pt
Authors’ names
11pt
Section title
10pt
Authors’ affiliations, main text, equations
10pt
italic
Subheadings
9pt
bold
Abstract
9pt
Section titles, table names, table captions, tables figure captions, references, Footnotes, text subscripts and superscripts
6pt
table superscripts
three components: 1. Storage component, 2. Target image processing component and 3. Query component. B. Storage Component This component is used to store the database of the library images. The input images go through necessary enhancement, proper transformation, feature vectors extraction and the database storage for the library images and their feature vector matrix.
F. DatabaseSstorage The library PCB images are stored features of choice are calculated off-line and stored for each image so efficient computation is not a critical criterion for the matching process in the query stage.
C. Target Image Processing Component The image passes through an automatic image enhancement (algorithm 8) that was developed in chapter 5; the goal of this step is to reduce the effect of noise, to stress the important features of the image and to ease the process of automatic image recognition. D. Transformation In this step a suitable transformation is performed to allow the process of extraction of the feature vectors of choice. In this simulation two types of transformation were used:
Figure 2– Database Storage system.
G. Query Component
V. Summary and Conclusions The inputs for this component are composed by the PCB images library feature vector matrix, the target image feature vector and the matching threshold. It computes the distance between the target image feature vector and the rows of the matrix and reports the images with distance below input matching threshold. H. Query Component The inputs for
III. Figures and Tables Position figures and tables at the tops and bottoms of columns. Avoid placing them in the middle of columns. Large figures and tables may span across both columns. Figure captions should be below the figures; table captions should be above the tables. Avoid placing figures and tables before their first mention in the text. Use the abbreviation “Fig.1”, even at the beginning of a sentence. • The paper's graphics should have resolutions of 600dpi for monochrome, 300 dpi for grayscale. But, please do not include too much high-resolution images or photos in your paper. • Please do not include color figures in the PDF file.
IV. Helpful Hints A. References List and number all references at the end of the paper. When referring to them in the text, type the corresponding reference number in the parentheses as shown at the end of this sentence [1]. Number the citations consecutively. The sentence punctuation follows the parentheses. Do not use “Ref.[2]” or “reference[2]” except at the beginning of a sentence.
B. Footnotes Number the footnotes separately in superscripts. Place the actual footnote at the bottom of the column in which it is cited. Do not put footnotes in the reference list.
C. Authors names Give all authors' names; do not use “et al” unless there are six authors or more. Papers that have not been published, even if they have been submitted for publication, should be cited as “unpublished”[3]. Papers that have been accepted for publication should be cited as “in press”[4]. Capitalize only the first word in a paper title, except for proper nouns and element symbols. For papers published in translation journals, please give the English citation first, followed by the original foreign language citations[5].
This template can be downloaded through the ACISC 2007 web site (http://www.acisc.org). If you have any problem, please contact Michael Orshansky
[email protected].
References [1] I. M. Author, “Some related article I wrote,” Some Fine Journal, Vol. 17, pp. 1-100, 1987. [2] A. N. Expert, A Book He Wrote, His Publisher, 1989. [3] M. Smith, “Title of paper optional here,” unpublished. [4] K. Rose, “Title of paper with only first word capitalized,” in press. [5] T. Murayama, “Title of paper published in translation journals,” Some English Journal, Vol. 17, pp. 1-100, 1995.