基于高光谱成像技术的涂抹掩盖字迹识别方法研究

王鸣久; 代雪晶; 汤澄清; 卢兆一

基于高光谱成像技术的涂抹掩盖字迹识别方法研究

1.
中国刑事警察学院公安信息技术与情报学院, 辽宁沈阳 110035
2.
中国刑事警察学院刑事技术学院, 辽宁沈阳 110035

基金项目:

公安部科技强警基础工作专项项目 2016GABJC06

详细信息

作者简介:
王鸣久（1983-），男，讲师，主要研究方向为公安图像技术，E-mail：396105061@qq.com

通讯作者:
代雪晶（1970-），女，教授，主要研究方向为公安图像技术，E-mail：1210724331@qq.com

中图分类号: DF794.2
计量
- 文章访问数: 97
- HTML全文浏览量: 33
- PDF下载量: 26
- 被引次数: 0
出版历程
- 收稿日期: 2022-01-11
- 修回日期: 2022-03-18
- 刊出日期: 2023-01-20

Research on Identification Method of Covered Handwriting Based on Hyperspectral Imaging Technology

1.
College of Public Security Information Technology, Criminal Investigation Police University of China, Shenyang 110035, China
2.
College of Forensic Sciences, Criminal Investigation Police University of China, Shenyang 110035, China

摘要

摘要: 在刑事案件和民事案件中，许多重要文件中的签名、日期、数字等字迹常常被有意涂抹而掩盖其真实信息，因而无法作为证据使用。因此，开展高光谱成像技术（450~950 nm）快速无损显现涂抹掩盖字迹十分必要。涂抹掩盖字迹中，黑色签字笔作为书写及涂抹工具占多数，因此从21种黑色签字笔中随机选取两种进行相互涂抹作为样本。同时，选取墨水中含有碳元素和不含有碳元素的黑色签字笔各两支，用同一支笔书写字迹并进行涂抹掩盖作为样本。实验结果表明，用高光谱成像技术在450~950 nm波长范围内进行显现肉眼无法辨别的涂抹掩盖字迹，大部分被涂抹掩盖的字迹能够得到完整、清晰的显现，反映出被掩盖字迹的真实形态与信息，具有良好的显现效果。部分显现效果不佳，原书写字迹细节特征残缺，但能识别出原字迹，还有少部分较为模糊，无法达到检验要求。而使用传统红外显现方法时，大部分涂抹掩盖字迹难以显现。
- 高光谱成像 /
- 掩盖字迹 /
- 无损检验 /
- 黑色签字笔
Abstract: In criminal and civil cases, signatures, dates, numbers, and other handwritings of several important documents are deliberately covered to conceal the real information such that they cannot be used as evidence. Therefore, it is necessary to study the feasibility of hyperspectral imaging technology (approximately 450–950 nm) for rapid and non-destructive identification of covered handwriting. Black handwriting pens are mostly used such that two pens are randomly selected from 21 kinds of black handwriting pens to cover each other as samples. Simultaneously, two kinds of black handwriting pens with carbon in the ink and two kinds of black handwriting pens without carbon in the ink are selected, and the same pen is used to write and cover by itself as a sample. The experimental results show that most of the covered handwriting can be completely and clearly identified using hyperspectral imaging technology in the band of approximately 450–950 nm with the real shape and information, which has a good effect. Part of the identification effect is not good, and the details of the original handwriting are incomplete, but the original handwriting can be recognized. In addition, few parts are blurred, which cannot meet the identification requirements. However, most of the covered handwriting is difficult to develop using the traditional infrared method.
- hyperspectral imaging /
- covered handwriting /
- non-destructive identification /
- black handwriting pen

HTML全文

图 1 光谱成像原理示意图

Figure 1. Diagram of spectral imaging principle

下载: 全尺寸图片幻灯片

图 2 光谱成像系统

Figure 2. Device of spectral imaging technology

下载: 全尺寸图片幻灯片

图 3 1#签字笔掩盖报纸上打印字迹

Figure 3. Printed handwriting in newspaper covered with 1#

下载: 全尺寸图片幻灯片

图 4 2#签字笔掩盖报纸上打印字迹

Figure 4. Printed handwriting in newspaper covered with 2#

下载: 全尺寸图片幻灯片

图 5 3#签字笔掩盖报纸上打印字迹

Figure 5. Printed handwriting in newspaper covered with 3#

下载: 全尺寸图片幻灯片

图 6 4#签字笔掩盖报纸上打印字迹

Figure 6. Printed handwriting in newspaper covered with 4#

下载: 全尺寸图片幻灯片

图 7 2#签字笔掩盖复印纸上打印字迹

Figure 7. Printed handwriting on copy paper covered with 2#

下载: 全尺寸图片幻灯片

图 8 3#签字笔掩盖复印纸上打印字迹

Figure 8. Printed handwriting on copy paper covered with 3#

下载: 全尺寸图片幻灯片

图 9 4#签字笔掩盖复印纸打印字迹

Figure 9. Printed handwriting on copy paper covered with 4#

下载: 全尺寸图片幻灯片

图 10 5#签字笔掩盖复印纸上打印字迹

Figure 10. Printed handwriting on copy paper covered with 5#

下载: 全尺寸图片幻灯片

图 11 6#签字笔涂抹掩盖5#签字笔

Figure 11. 5# handwriting covered with 6#

下载: 全尺寸图片幻灯片

图 12 5#签字笔涂抹掩盖6#签字笔

Figure 12. 6# handwriting covered with 5#

下载: 全尺寸图片幻灯片

图 13 8#签字笔涂抹掩盖1#签字笔

Figure 13. 1# handwriting covered with 8#

下载: 全尺寸图片幻灯片

图 14 1#签字笔涂抹掩盖8#签字笔

Figure 14. 8# handwriting covered with 1#

下载: 全尺寸图片幻灯片

图 15 1#签字笔书写并涂抹掩盖

Figure 15. 1# handwriting covered by itself

下载: 全尺寸图片幻灯片

图 16 4#签字笔书写并涂抹掩盖

Figure 16. 4# handwriting covered by itself

下载: 全尺寸图片幻灯片

图 17 9#签字笔书写并涂抹掩盖

Figure 17. 9# handwriting covered by itself

下载: 全尺寸图片幻灯片

图 18 10#签字笔书写并涂抹掩盖

Figure 18. 10# handwriting covered by itself

下载: 全尺寸图片幻灯片

图 19 传统红外显现掩盖打印字迹效果

Figure 19. The covered printed handwriting developed by traditional infrared method

下载: 全尺寸图片幻灯片

图 20 传统红外显现掩盖黑色签字笔字迹效果

Figure 20. The covered black pen handwriting developed by traditional infrared method

下载: 全尺寸图片幻灯片

参考文献(15)

[1]	许小京, 黄威. 光谱成像技术在物证鉴定领域的应用[J]. 红外与激光工程, 2012, 41(12): 3280-3284. doi: 10.3969/j.issn.1007-2276.2012.12.027 XU Xiaojing, HUANG Wei. Application of spectral imaging in forensic science[J]. Infrared and Laser Engineering, 2012, 41(12): 3280-3284. doi: 10.3969/j.issn.1007-2276.2012.12.027
[2]	Edelman G J, Gaston E, Leeuwen T G van, et al. Hyperspectral imaging for non-contact analysis of forensic traces[J]. Forensic Science International, 2012, 223: 28-39. doi: 10.1016/j.forsciint.2012.09.012
[3]	Panagou E Z, Papadopoulou O, Carstensen J M, et al. Potential of multispectral imaging technology for rapid and non-destructive determination of the microbiological quality of beef filets during aerobic storage[J]. International Journal of Food Microbiology, 2014, 174: 1-11. doi: 10.1016/j.ijfoodmicro.2013.12.026
[4]	Belinda Bastide, Glenn Porter, Adrian Renshaw. Detection of Latent Bloodstains at Fire Scenes Using Reflected Infrared Photography[J]. Forensic Science International, 2019, 302: 109874 doi: 10.1016/j.forsciint.2019.109874
[5]	Zapata F, García-Ruiz C. Trac. Emerging spectrometric techniques for the forensic analysis of body fluids[J]. Trends in Analytical Chemistry, 2015, 64: 53-63.
[6]	张华锋, 王武, 白玉荣, 等. 多光谱成像无损识别冻融猪肉中危害级碎骨[J]. 光谱学与光谱分析, 2021, 41(9): 2892-2897. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN202109046.htm ZHANG Huafeng, WANG Wu, BAI Yurong, et al. Non-destructive identification of hazardousbone fragments embedded in the frozen-thawed pork based on multispectral imaging[J]. Spectroscopy and Spectral Analysis, 2021, 41(9): 2892-2897. https://www.cnki.com.cn/Article/CJFDTOTAL-GUAN202109046.htm
[7]	Alsberg B K, Loke T, Baarstad I, et al. PryJector: a device for in situ visualization of chemical and physical property distributions on surfaces using projection and hyperspectral imaging[J]. Journal of Forensic Sciences, 2011, 56(4): 976-983. doi: 10.1111/j.1556-4029.2011.01747.x
[8]	Schuler R L, Kish P E, Plese C A. Preliminary observations on the ability of hyperspectral imaging to provide detection and visualization of bloodstain patterns on black fabrics[J]. Journal of Forensic Sciences, 2012, 57(6): 1562-1569. doi: 10.1111/j.1556-4029.2012.02171.x
[9]	Tahtouh M, Scott S A, Kalman J R, et al. Four novel alkyl 2-cyanoacylate monomers and their use in latent fingermark detection by mid-infrared spectral imaging[J]. Forensic Science International, 2011, 207(1-3): 223-238 doi: 10.1016/j.forsciint.2010.10.012
[10]	Joong Lee, Seong G Kong, Tae-Yi Kang, et al. Invisible ink mark detection in the visible spectrum using absorption difference[J]. Forensic Science International, 2014, 236: 77–83. doi: 10.1016/j.forsciint.2013.12.024
[11]	Edelman G, Manti V, van Ruth S M, et al. Identification and age estimation of blood stains on colored backgrounds by near infrared spectroscopy[J]. Forensic Science International, 2012, 220: 239-244. doi: 10.1016/j.forsciint.2012.03.009
[12]	Edelman G, van Leeuwen T G, Aalders M C. Hyperspectral imaging for non-contact analysis of forensic traces[J]. Forensic Science International, 2012, 223: 1-3. doi: 10.1016/j.forsciint.2012.04.012
[13]	Binu Melit Devassy, Sony George. Dimensionality reduction and visualisation of hyperspectral ink data using t-SNE[J]. Forensic Science International, 2020, 311: 109874.
[14]	Lívia Rodrigues e Brito, André Braz, Ricardo Saldanha Honorato, et al. Evaluating the potential of near infrared hyperspectral imaging associated with multivariate data analysis for examining crossing ink lines[J]. Forensic Science International, 2019, 298: 169-176. doi: 10.1016/j.forsciint.2019.02.043
[15]	Estelles-Lopez L, Ropodi A, Pavlidis D, et al. An automated ranking platform for machine learning regression models for meat spoilage prediction using multi-spectral imaging and metabolic profiling[J]. Food Research International, 2017, 99: 206. doi: 10.1016/j.foodres.2017.05.013