Background knowledge and system structure of fingerprint recognition attendance machine
I. Background knowledge of fingerprint recognition
We The unevenness of the skin on the palms and the inner surfaces of the fingers, feet, and toes creates a variety of patterns. The presence of these lines increases the friction on the skin’s surface, allowing us to pick up heavy objects with our hands. It was also noted that the lines of these skins, including fingerprints, varied in pattern, breakpoints, and intersections, that is, unique. Relying on this uniqueness, we can associate a person with his fingerprints, and verify his true identity by comparing his fingerprints with pre-stored fingerprints. This technology that relies on the physical characteristics of the human body for identity verification is called biometric technology, and fingerprint recognition is a type of biometric technology.
At present, from a practical point of view, fingerprint identification technology is an identification method that is superior to other biometric identification technologies. This is because it has been recognized that fingerprints are different and basically unchanged for life.
The earliest application of fingerprint identification system and the detection of criminal suspects by the police has a history of more than 30 years, which has laid a good technical foundation for the research and practice of fingerprint identification. In particular, the current fingerprint identification system has reached the stage of convenient operation, accurate and reliable, and moderate price, and is gradually being applied to the civilian market.
Fingerprint identification system collects, analyzes and compares living fingerprints through special photoelectric conversion equipment and computer image processing technology, and can quickly and accurately identify personal identity. The system generally includes the process of fingerprint image acquisition, fingerprint image processing, feature extraction, feature value comparison and matching. Modern electronic integrated manufacturing technology makes fingerprint image reading and processing equipment miniaturized. At the same time, the rapid development of personal computer computing speed provides the possibility of fingerprint comparison operations on microcomputers and even single-chip microcomputers. Excellent fingerprint processing and comparison algorithms guarantee the accuracy of the recognition results.
Automatic fingerprint recognition technology is coming into our real life from science fiction and Hollywood movies. Maybe one day, you don’t have to carry that bunch of keys with you, just press your finger and the door will open; You don’t have to remember that annoying password, you can use your fingerprint to withdraw money and log in to the computer. Believe that this day will not be too far away.
2. System composition
1. Fingerprint collection device
Currently the most commonly used image acquisition Devices fall into two categories: optical and crystal sensors.
Optical imaging equipment has the longest history, dating back to the 1970s. Optical imaging devices are based on the principle of total reflection of light (FTIR). The light hits the glass surface with the fingerprint pressed, and the reflected light is obtained by the CCD. The amount of reflected light depends on the depth of the fingerprint ridges and valleys on the glass surface and the oil and moisture between the skin and the glass. After the light hits the valley through the glass, total reflection occurs at the interface between the glass and the air, and the light is reflected to the CCD, while the light directed to the ridge does not undergo total reflection, but is absorbed by the contact surface between the ridge and the glass or diffusely reflected to other places. place, so that the image of the fingerprint is formed on the CCD.
Due to recent innovations in optical equipment, the size of the equipment has been greatly reduced. As recently as the mid-90s, sensors could fit in a 6x3x6 inch box, and in the near future smaller devices are 3x1x1 inches. These advances depend on the development of multiple optical technologies rather than the development of FTIR. For example, a fiber beam can be used to acquire fingerprint images. The fiber beam hits the surface of the fingerprint perpendicularly, illuminating the fingerprint and detecting the reflected light. Another solution is to mount a surface containing a matrix of micro-triangular prisms on an elastic plane. When a finger is pressed on this surface, the surface of the micro-triangular prisms is changed due to the different pressures of the ridges and valleys. These changes are reflected by the triangular prism light. and reflected.
Crystal sensors are a relatively recent addition to the market, although they have been around for nearly 20 years in technical introductory articles. These planes, which contain tiny crystals, are used to create images of fingerprints using a variety of techniques. The most common silicon capacitive sensors are designed to capture fingerprints by electronic metrology. About 100,000 capacitive sensors can be combined on a semiconductor metal array with an insulating surface on which the skin forms the other side of the capacitive array when the user’s finger is placed on it. The capacitance value of a capacitor decreases due to the distance between the conductors, which here refers to the distance between the ridge (near) and the valley (far) relative to the other pole. Another type of crystal sensor is pressure-sensitive, and the top layer of its surface is an elastic pressure-sensitive dielectric material, which is converted into corresponding electronic signals according to the surface topography (concave and convex) of the fingerprint. Other crystal sensors are temperature-sensing sensors, which can obtain fingerprint images by sensing the difference in temperature between the ridges pressing on the device and the valleys away from the device.
2. Fingerprint comparison algorithm (including fingerprint image processing, special positive extraction, comparison and matching)
In terms of fingerprint comparison algorithm, There are two main concepts, verification and identification; verification is what we often call the 1:1 algorithm, and identification is the 1:N algorithm. At the same time, there are two important parameters for the fingerprint comparison algorithm: the false recognition rate and the false rejection rate.
2.1 Verification
Verification is the one-to-one matching of a fingerprint collected on-site with a registered fingerprint to confirm process of identity. As a prerequisite for verification, his or her fingerprint must have been registered in the fingerprint database. Fingerprints are stored in a certain compressed format and associated with their names or their identifications (ID, PIN). Then at the comparison site, first verify its identity, and then use the fingerprint of the system to compare with the fingerprint collected on site to prove that its identity is legal. Verification actually answers the question: “Is he who he claims to be?” This is a more commonly used method in application systems.
2.2 Identification
Identification is to compare the fingerprints collected on the spot with the fingerprints in the fingerprint data store one by one, and find out the fingerprints that match the on-site fingerprints. This is also called “one-to-many matching”. Verification actually answers the question: “Who is he?” Identification is mainly used in the traditional field of criminal fingerprint matching. The fingerprints of an unidentified person are compared with the fingerprints of someone with a criminal record in the fingerprint database to determine whether the person has a criminal record.
2.3 False Recognition Rate and Rejection Rate
Because when the computer processes fingerprints, it only involves some limited information of the fingerprints, and the comparison algorithm is not an exact match, and the results are also It cannot be guaranteed to be 100% accurate. An important measure of a specific application of a fingerprint recognition system is the recognition rate. It is mainly composed of two parts, the true rejection rate (FRR) and the false recognition rate (FAR). We can adjust these two values according to different uses. FRR and FAR are inversely proportional. Express this number as 0-1.0 or as a percentage.
