Historically, in law enforcement applications, the acquisition of fingerprint images was performed by using the so-called “ink-technique”: the subject's fingers were smeared with black ink and pressed or rolled on a paper card; the card was then scanned by using a general purpose scanner, producing a digital image. This kind of acquisition process is referred to as offline fingerprint acquisition or off-line sensing and is briefly discussed in Section 2.2. A particular case of off-line sensing is the acquisition of latent fingerprints from crime scenes (Colins, 1992). Nowadays, most civil and criminal AFIS accept live-scan digital images acquired by directly sensing the finger surface with an electronic fingerprint scanner (also called fingerprint reader). No ink is required in this method, and all that a subject has to do is to present his finger to a live-scan scanner. Although AFIS has greatly benefited from the use of live-scan acquisition techniques, this innovation is undoubtedly more important for a broad range of civil and commercial applications where user acceptance and convenience, low-cost, and reliability are necessary and expected. In civil and commercial applications, certainly, an employee cannot be expected to apply ink to his fingertip every time he has to logon to his personal computer or to carry out a financial transaction; neither could we expect a wide adoption of fingerprint-based biometric techniques if the cost of the acquisition devices was too high.
The general structure of a typical fingerprint scanner is shown in Figure 2.1: a sensor reads the ridge pattern on the finger surface and converts the analog reading in the digital form through an A/D (Analog to Digital) converter; an interface module is responsible for communicating (sending images, receiving commands, etc.) with external devices (e.g., a personal computer). Throughout this chapter we use the terms “scanner” and “sensor” with different meanings: with sensor we denote the internal active sensing element of a fingerprint scanner that reads the finger surface. The different technologies the sensors are based on (e.g., optical, solid-state, ultrasound, etc.) are surveyed in Section 2.3. In practice, there exist several variants of the schema in Figure 2.1: for example, often the sensor output is already a digital signal and therefore no separate A/D conversion is necessary; some fingerprint scanners may not have an integrated A/D converter and an external frame grabber would be needed to transform their analog output signal. Furthermore, some embedded System-on-a-Chip devices have been proposed (Anderson et al. (1991); Shigematsu et al. (1999); Jung et al. (1999, 2005)) where, besides the sensor, a processing board is embedded into the chip in order to locally process and/or match the fingerprint data (see Section 9.6.2). The design of secure fingerprint-based biometric systems requires protection/encryption mechanisms to be implemented in the bio-metric scanners. Chapter 9 discusses the techniques used to protect fingerprint scanners against various attacks and to detect fake fingers presented to the sensors.
- Discrete Wavelet Transform
- Modulation Transfer Function
- Iterative Close Point
- Capacitive Sensor
- Fingerprint Image
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.