Восстановить пароль
FAQ по входу

Bouridane A. Imaging for Forensics and Security. From Theory to Practice

  • Файл формата pdf
  • размером 8,37 МБ
  • Добавлен пользователем
  • Отредактирован
Bouridane A. Imaging for Forensics and Security. From Theory to Practice
Springer, 2009. — 221 p.
The field of security has witnessed an explosive growth during the last years, as phenomenal advances in both research and applications have been made. Biometric and forensic imaging applications often involve photographs, videos and other image impressions that are fragile and include subtle details that are difficult to see. As a developer, one needs to be able to quickly develop sophisticated imaging applications that allow for an accurate extraction of precious information from image data for identification and recognition purposes. This is true for any type of biometric and forensic image data.
The applications covered in this book relate to Biometrics, Watermarking and Shoeprint recognition for forensic science. Image processing transforms using Discrete Fourier Transform, Discrete Wavelet Transforms Gabor Wavelets, Complex Wavelets, Scale Invariant Feature Transforms and Directional Filter banks are used in data modelling process for either feature extraction or data hiding tasks. The emphasis is on the methods and the analysis of data sets including comparative studies against existing and similar techniques. To make the underlying methods accessible to a wider audience, we have stated some of the key mathematical results given in a logical structure of the development.
For example, biometric based methods are emerging as the most reliable solutions for authentication and identification applications where traditional passwords (knowledge-based security) and ID cards (token-based security) have been used so far to access restricted systems. Automated biometrics deal with physiological or behavioural characteristics such as fingerprints, iris, voice and face that can be used to authenticate a person’s identity or establish an identity within a database. With rapid progress in electronic and Internet commerce, there is also a growing need to authenticate the identity of a person for secure transaction processing. Current biometric systems make use of fingerprints, hand geometry, iris, retina, face, facial thermograms, signature gait, and voiceprint to establish a person’s identity. While biometric systems have their limitations they have an edge over traditional security methods in that they cannot be easily stolen or shared. Besides bolstering security, biometric systems also enhance user convenience by alleviating the need to design and remember passwords.
Driven by the urgent need to protect digital media content that is being widely and wildly distributed and shared through the Internet by an ever-increasing number of users, the field of digital watermarking has witnessed an extremely fast-growing development since its inception almost a decade ago. The main purpose of digital watermarking, information embedding, and data hiding systems is to embed auxiliary information, usually called digital watermarks, inside a host signal (audio, image, and video) by introducing small and minor perturbations into the host signal. The quality of the host signal should not be degraded unacceptably and the introduced changes lie below the minimum perception threshold of the intended recipient. Watermark detection and extraction from the composite host signal should be possible in the presence of a variety of intentional and unintentional manipulations and attacks. It is obvious that these attacks and manipulations do not corrupt the composite host signal at an unacceptable level.
Watermarking systems are expected to play an important role in meeting at least two major challenges that resulted from the widespread use of Internet for the distribution and exchange of digital media: (i) error-free perfect copies of digital multimedia and (ii) availability of free and affordable tools for the manipulation and alteration of digital content. The first challenge was the driving force that led the combined efforts of academic and industrial research to produce first-generation watermarking algorithms. These algorithms were mainly concerned with the copyright protection of the digital content. For instance, illegal distribution and copying of digital music is causing the music industry massive gain losses. The second challenge has guided the research efforts to develop what are so-called tamper-proof or fragile watermarking algorithms. This class of watermarking schemes aims at detecting any intentional manipulation or corruption of the media.
Following the emergence and success of forensic science as a powerful and irrefutable tool for solving many enigmatic crime puzzles, images collected from crime scenes are abounding and, therefore, large image collections are being created. Shoeprint images are no exception and it has been indicated, recently, that shoeprint evidence, at crime scenes, is more frequently present than fingerprints. Very recently, it has been suggested that shoeprint evidence should be made comparable to that of fingerprint and DNA evidence. It is also true that shoeprint intelligence remains an untapped potential forensic source (usually overshadowed by the accepted fingerprint and DNA evidence). However, there is no practical technology to efficiently search shoeprint on large databases. Existing commercial systems still require manual involvement (manual annotation of both the impression under investigation and the primary database). The task of automated scenemark matching is a tedious one and researching the use of existing image processing and pattern recognition techniques is desired before an underpinning technology is developed.
One of the most distinctive features of the book is that it covers in detail a number of imaging applications and their deployment in security problems. In addition, the book appeals to both undergraduate and postgraduate students since each application problem includes a detailed description of the mathematical background and its implementation.
Introduction and Preliminaries on Biometrics and Forensics Systems
Data Representation and Analysis
Improving Face Recognition Using Directional Faces
Recent Advances in Iris Recognition: A Multiscale Approach
Spread TransformWatermarking Using ComplexWavelets
Protection of Fingerprint Data Using Watermarking
Shoemark Recognition for Forensic Science: An Emerging Technology
Techniques for Automatic Shoeprint Classification
Automatic Shoeprint Image Retrieval Using Local Features
  • Чтобы скачать этот файл зарегистрируйтесь и/или войдите на сайт используя форму сверху.
  • Регистрация