Suppr超能文献

在法医成像中使用光谱信息。

Using spectral information in forensic imaging.

作者信息

Miskelly Gordon M, Wagner John H

机构信息

Department of Chemistry, Forensic Science Programme, The University of Auckland, Private Bag 92019, Auckland, New Zealand.

出版信息

Forensic Sci Int. 2005 Dec 20;155(2-3):112-8. doi: 10.1016/j.forsciint.2004.11.005. Epub 2004 Dec 30.

Abstract

Improved detection of forensic evidence by combining narrow band photographic images taken at a range of wavelengths is dependent on the substance of interest having a significantly different spectrum from the underlying substrate. While some natural substances such as blood have distinctive spectral features which are readily distinguished from common colorants, this is not true for visualization agents commonly used in forensic science. We now show that it is possible to select reagents with narrow spectral features that lead to increased visibility using digital cameras and computer image enhancement programs even if their coloration is much less intense to the unaided eye than traditional reagents. The concept is illustrated by visualising latent fingermarks on paper with the zinc complex of Ruhemann's Purple, cyanoacrylate-fumed fingerprints with Eu(tta)(3)(phen), and soil prints with 2,6-bis(benzimidazol-2-yl)-4-[4'-(dimethylamino)phenyl]pyridine [BBIDMAPP]. In each case background correction is performed at one or two wavelengths bracketing the narrow absorption or emission band of these compounds. However, compounds with sharp spectral features would also lead to improved detection using more advanced algorithms such as principal component analysis.

摘要

通过组合在一系列波长下拍摄的窄带摄影图像来改进法医证据的检测,取决于感兴趣的物质与底层基质具有显著不同的光谱。虽然一些天然物质(如血液)具有独特的光谱特征,很容易与常见的着色剂区分开来,但法医学中常用的可视化剂并非如此。我们现在表明,即使某些试剂的着色肉眼看起来比传统试剂淡得多,也可以选择具有窄光谱特征的试剂,通过数码相机和计算机图像增强程序来提高其可见性。用鲁赫曼紫的锌配合物可视化纸上的潜在指纹、用Eu(tta)(3)(phen)可视化氰基丙烯酸酯熏显的指纹以及用2,6-双(苯并咪唑-2-基)-4-[4'-(二甲氨基)苯基]吡啶[BBIDMAPP]可视化土壤印记,说明了这一概念。在每种情况下,在这些化合物窄吸收或发射带的一个或两个波长处进行背景校正。然而,具有尖锐光谱特征的化合物使用更先进的算法(如主成分分析)也会提高检测效果。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验