尚光网 本站首页 本所首页 联系我们
chi20190524上海光机所黄立华557591Trace is a carrier of evidence and information. The information of people, objects and things related to criminal acts carried by traces can provide strong technical support for investigation. Traces include marks left by handwriting, fingerprinting tools, bullets, etc. There are many differences between the image reflected by handwriting and fingerprints, tool marks and bullet marks. Handwriting, as a special kind of trace, reflects not only the static characteristics of the sample surface changed by writing tools, but also the unique writing habits of the writer. Handwriting trace detection is an important source of evidence in criminal investigation. The commonly used methods for detecting the trace on the surface of the dielectric include the side light observation method, the side light infrared image method, the static voltage trace development method, the 3d image reconstruction method and the laser confocal scanning microscope method. But each approach has its limitations. The side light observation method and side light infrared image method are limited by the direction of light source and can’t exclude the interference of ink; Static voltage indentation method of paper humidity requirements are much stringent, and there are other indentation or sweat fingerprints and other interference; The equipment of 3d reconstruction of microimage and confocal scanning laser microscopy is relatively high-end and expensive, and the scanning time is relatively long. Reflectance Transformation Imaging (RTI) was invented by Hewlett-Packard laboratory, and it is called Polynomial Texture Mapping (PTM) technology. RTI is a visual tool that can achieve a clear appearance of surface traces. At present, it has been widely used in the field of forensic science and other fields of cultural relic research. In this paper, surface trace detection technology based on Reflectance Transformation Imaging is studied. This paper mainly studies the trace detection of a special kind of trace – handwriting. The main content is as follows: (1) The characteristics and application of traces were investigated, and the characteristics of traces and its development in detection technology were emphatically described. The Reflectance Transformation Imaging, its application and the development of trace detection technology are summarized. (2) The theory basis of polynomial texture mapping and Reflectance Transformation Imaging and the application of Reflectance Transformation Imaging technology are studied. The experimental equipment of Reflectance Transformation Imaging is described in detail, which lays a foundation for the follow-up work. (3) The image quality evaluation method commonly used at present is described, the structural similarity (SSIM) image quality evaluation algorithm is detailed, and the no-reference image quality evaluation algorithm (NRSS) is studied. The NRSS flowchart used in this paper is given. (4) The experimental device based on RTI principle was set up and the experimental research was carried out. Four different experiments were conducted to verify the application effect of RTI in handwriting trace detection. Furthermore, the acquisition and analysis of handwriting trace characteristics are carried out. Finally, NRSS method was used to objectively evaluate the rendering effect of RTI on colored and colorless handwriting traces. In this paper, an experimental study on the detection of colored handwriting traces and colorless handwriting traces based on Reflectance Transformation Imaging technology is carried out. The test results show that the information such as the writing content, writing characteristics and the person's movement characteristics of colored and colorless handwriting can be obtained, and the clear appearance of colorless handwriting with a depth of more than ten microns can be realized. This paper proposes an NRSS reference-free image quality evaluation algorithm for RTI detection technology, which can objectively evaluate the colored and colorless handwriting trace detection results and eliminate the subjectivity of artificial evaluation.2019atalunwen21961323921868Reflectance Transformation Imaging; Polynomial Texture Mapping; handwriting trace detection; objective image quality evaluation; no referenceApplication of Reflectance Transformation Imaging for display of traces基于反射变换成像的表面痕迹检测技术研究痕迹是一种物证以及信息的载体。痕迹所携带的和犯罪行为有关联的人、物、事的相关信息可以为侦查提供有力的技术支撑。痕迹包括笔迹、指印、工具遗留痕迹、枪弹遗留痕迹等。笔迹与指印、工具遗留痕迹、枪弹的遗留痕迹等所反映出的形象有很多的不同之处。笔迹作为一类特殊的痕迹,其所反映出的不仅仅是样本表面被书写工具所改变的静态特征,还有书写人本身所独特的书写习惯等特征。笔迹痕迹检测是刑事侦查领域的重要证据来源。常用的介质表面痕迹检测方法有显微镜侧光观察法、侧光红外影像法、静电压痕显现法、显微图像三维重建法和激光共聚焦扫描显微镜法。但每种方法都有各自的局限。显微镜侧光观察法和侧光红外影像法受限于光源方向且不能排除墨迹干扰;静电压痕显现法对纸张的湿度要求较为苛刻且存在其他压痕或汗液指纹等干扰;显微图像三维重建法和激光共聚焦扫描显微镜法仪器设备较为高端昂贵,扫描时间较长。 反射变换成像技术(Reflectance Transformation Imaging,简称RTI)由惠普(Hewlett-Packard)实验室发明,并将此方法命名为多项式纹理映射(Polynamial Texture Mapping,简称PTM)技术。RTI是一种视觉工具,能实现表面痕迹的清晰显现,目前已广泛应用于文物研究、法庭科学领域等领域。 本文开展了基于反射变换成像的表面痕迹检测技术研究,主要针对特殊的一类痕迹—笔迹,进行痕迹检测研究,主要内容如下: (1)调研了痕迹的特性及应用,着重叙述了笔迹痕迹的特点及其在检测技术发展;概述了反射变换成像技术及应用,以及痕迹检测技术发展的现状。 (2)研究了多项式纹理映射技术、反射变换成像技术的理论基础以及反射变换成像技术的应用,详述了反射成像技术的实验装置,为后续工作的开展奠定了基础。 (3) 叙述了目前普遍应用的图像质量评价方法,详述了结构相似性(SSIM)图像质量评估算法,研究了无参考图像质量评价算法(NRSS)。结合RTI痕迹显现的特点,给出了本文所使用的NRSS流程图。 (4)搭建了基于RTI原理的实验装置,开展了实验研究。通过4次不同的实验分别了验证RTI在笔迹痕迹检测方面的应用效果。进而,开展了深入的笔迹痕迹特征的获取与分析,通过镜面增强模式,可获得笔迹的运笔、顿笔等特征。最后,采用NRSS方法对RTI有色、无色笔迹痕迹显现效果进行了客观评价。 本文开展了基于反射变换成像技术的有色笔迹痕迹和无色笔迹痕迹检测的实验研究。检测结果表明,能获得有色笔迹、无色笔迹的书写内容、书写特征以及书写人的动作特征等信息,且实现了对十几微米深度的无色笔迹的清晰显现。本文提出了适用于RTI检测技术的NRSS无参考图像质量评价算法,可以对有色的以及无色的笔迹痕迹检测结果进行客观评价,消除了人为评价的主观性。反射变换成像技术;多项式纹理映射;笔迹痕迹检测;客观图像质量评价;无参考中国科学院上海光学精密机械研究所魏巍光学工程硕士
中文题目: 基于反射变换成像的表面痕迹检测技术研究
外文题目: Application of Reflectance Transformation Imaging for display of traces
作者: 魏巍
导师姓名: 黄立华
学位授予机构: 中国科学院上海光学精密机械研究所
答辩时间: 20190524
中文关键词:
反射变换成像技术;多项式纹理映射;笔迹痕迹检测;客观图像质量评价;无参考
英文关键词:
Reflectance Transformation Imaging; Polynomial Texture Mapping; handwriting trace detection; objective image quality evaluation; no reference
中文摘要:
英文摘要:
文献类型:学位论文
学位级别: 硕士
正文语种: chi
全文传递服务
clickdetails
页面点击量: 43
文章下载量: 1
visitlog
友情链接:
  上海大恒公司
  南京先进激光技术院
  光学产品库
版权所有 © 2009 中国科学院上海光学精密机械研究所 沪ICP备05015387号
主办:中国科学院上海光学精密机械研究所 上海市嘉定区清河路390号(201800)