Fingerprint scanners detect and identify a person by analyzing the fingerprint patterns collected by the device registration. The technology combines optical, statistical findings, fingerprint algorithms, feature matching and pattern recognition algorithm support.
The development of fingerprint recognition technology in the present day is well known by the public for its maturity and security. Among all the biometric scanning devices, fingerprint recognition has its own side of the world.
The scientific theory behind fingerprint recognition
The important measurement mark of fingerprint recognition system is the recognition rate. It consists of two main components, the rejection rate (FRR) and the false positive rate (FAR). Because of this, authorities consider a 1% false positive rate to be acceptable in applications.FRR is actually an important indicator of the ease of use of the system as well. Since FRR and FAR are contradictory, this makes it important to weigh ease of use and security in the design of the application system. The following is a collection of what principle fingerprint recognition uses, hope it will be helpful to you.
Step 1:
A fingerprint is a pattern produced by the bumpy and uneven skin on the front of the end of the finger.
Although the fingerprint is only a small part of the human skin, it contains a large amount of information. Fingerprint characteristics can be divided into two categories: general characteristics and local characteristics. The general features are those that can be directly observed by the human eye, including the basic pattern, pattern areas, core points, triangular points, pattern lines and lines, etc. The basic pattern includes rings, arches, and spirals. Local features are the characteristics of nodes on a fingerprint. These nodes with certain characteristics are called feature points. Two fingerprints will often have the same general characteristics, but their local characteristics, the feature points, are not likely to be identical. Feature points on a fingerprint are the end points, bifurcation points and turning points on the fingerprint pattern.
Step 2:
Fingerprint recognition technology usually uses the overall features of the fingerprint such as pattern shape and triangle points for classification, and then uses local features such as position and orientation for user identification.
Usually, the "minutiae" are first found on the acquired fingerprint image, and then a digital representation of the user's live fingerprint is created based on the characteristics of the minutiae - a fingerprint signature data (a one-way conversion from fingerprint image to signature data but not from signature data to fingerprint image). (a one-way conversion from fingerprint image to feature data but not from feature data to fingerprint image). Since two different fingerprints do not produce the same feature data, the fingerprint image is captured by. The pattern matching of the collected fingerprint image and the fingerprint feature data stored in the database is performed to calculate their similarity, and finally the matching result of the two fingerprints is obtained, and the user's identity is identified based on the matching result. Since the fingerprints of each person are different, even among the ten fingers of the same person, there are obvious differences, so fingerprints can be used for identification.
First, the human fingerprint image is read by a fingerprint reading device, and after the fingerprint image is taken, the original image is pre-processed.
Second, the fingerprint recognition software creates a digital representation of the fingerprint data, which is a unidirectional conversion from fingerprint to feature data but not from feature data to fingerprint, and two different fingerprints do not produce the same feature data. The software finds the data points called "nodes" on the fingerprint, which are the coordinates of the bifurcations, terminations or circles of the fingerprint pattern that have more than seven unique characteristics at the same time.
Third, the average fingerprint has 70 nodes.
The average finger has 70 nodes, so this method produces about 500 data points. Some algorithms produce more data by combining node and orientation information that indicates the relationship between the individual nodes, and some algorithms also process the entire fingerprint image. In summary, this data, often called templates, is saved as 1KB size records.
Finally, the templates of the two fingerprints are compared by means of a computerized fuzzy comparison, which calculates their degree of similarity and finally gives the result of matching the two fingerprints.
Airport security and fingerprint identification
Fingerprint recognition scanning systems have been put into use in airports around the world, and airports also have military priorities where the use of biometrics can provide fast, reliable identification and ensure data security. The current fingerprint recognition technology is a relatively mature product in the development of biometric identification, if still using the traditional manual identification, how many security risks there will be
How fingerprint recognition works
1、Fingerprint image acquisition
Through the special fingerprint acquisition instrument can collect fingerprint image. The fingerprint sensor used in the fingerprint collector is mainly divided into two types according to the collection method: scratch type and press type, and there are optical type, pressure sensitive type, capacitive type, inductive type, thermal type and ultrasonic type according to the signal collection principle. In addition, fingerprint hydrogen image can also be obtained through scanners, digital cameras, etc. The public security industry generally uses rolling fingerprint.
2、Fingerprint image processing
(1) Fingerprint image compression: large-capacity fingerprint database must be stored after compression to reduce storage space. The main methods include JPEG, WSQ, EZW, etc.
(2) Fingerprint image processing: including fingerprint region detection, image quality judgment, direction map and frequency estimation, image enhancement, fingerprint image binarization and refinement, etc. Pre-processing refers to the processing of fingerprint images containing noise and pseudo-features using certain algorithms to make the structure of lines clear and feature information prominent. The purpose is to improve the quality of the fingerprint image and the accuracy of feature extraction. Usually, the pre-processing process includes normalization, image segmentation, enhancement, binarization and refinement, but the steps of pre-processing vary according to the specific situation.
3、Fingerprint feature extraction
The pattern type is the basic classification of fingerprints, which is divided by the basic shape of the central pattern and triangle. The pattern shape is subordinate to the type, and is named by the shape of the central line. Our ten fingerprint analysis method divides fingerprints into three major types and nine patterns. Generally, the fingerprint automatic identification system divides fingerprints into arch-shaped patterns (arc-shaped patterns, tent-shaped patterns), skip-shaped patterns (left skip, right skip), bucket-shaped patterns and miscellaneous patterns.
The fingerprint morphological features include the center (top and bottom) and triangle points (left and right), etc. The detail feature points of fingerprints mainly include the starting point, end point, union point and bifurcation point of the lines. The fingerprint feature point information (end point, bifurcation point...) is extracted from the pre-processed image. ....) The information mainly includes parameters such as type, coordinates, and direction. The detailed features in fingerprints usually include endpoints, bifurcation points, isolated points, short bifurcations, rings, etc. The endpoints and bifurcation points are the most frequent, stable and easily harnessed in fingerprints. These two types of feature points can be matched to the fingerprint features: calculate the similarity between the feature extraction result and the stored feature template.
4、Fingerprint matching
Fingerprint matching is to compare the fingerprint features collected in the field with the fingerprint features stored in the fingerprint database to determine whether they belong to the same fingerprint. A coarse match can be made based on the pattern shape of the fingerprint, and then an exact match can be made using the fingerprint morphology and detail features to give the similarity score of the two fingerprints. Depending on the application, the similarity score is ranked or a verdict is given as to whether the fingerprints are the same
There are two ways of fingerprint comparison:
(1) One-to-one comparison: the user fingerprint to be compared is retrieved from the fingerprint database based on the user ID and then compared with the newly collected fingerprint;
(2)-to-many comparison: The newly collected fingerprints are compared with all fingerprints in the fingerprint database one by one.
The fingerprint image can be collected by a special fingerprint collector. The fingerprint sensor used in fingerprint capture device is mainly divided into two types according to the acquisition method: scratch type and press type, and there are currently optical, pressure sensitive, capacitive, inductive, thermal and ultrasonic types according to the signal acquisition principle. In addition, fingerprint hydrogen image can also be obtained through scanners, digital cameras, etc. The public security industry generally uses rolling fingerprint.
2、Fingerprint image processing
(1) Fingerprint image compression: large-capacity fingerprint database must be stored after compression to reduce storage space. The main methods include JPEG, WSQ, EZW, etc.
(2) Fingerprint image processing: including fingerprint region detection, image quality judgment, direction map and frequency estimation, image enhancement, fingerprint image binarization and refinement, etc. Pre-processing refers to the processing of fingerprint images containing noise and pseudo-features using certain algorithms to make the structure of lines clear and feature information prominent. The purpose is to improve the quality of the fingerprint image and the accuracy of feature extraction. Usually, the pre-processing process includes normalization, image segmentation, enhancement, binarization and refinement, but the steps of pre-processing vary according to the specific situation.
3、Fingerprint feature extraction
The pattern type is the basic classification of fingerprints, which is divided by the basic shape of the central pattern and triangle. The pattern shape is subordinate to the type, and is named by the shape of the central line. Our ten fingerprint analysis method divides fingerprints into three major types and nine patterns. Generally, the fingerprint automatic identification system divides fingerprints into arch-shaped patterns (arc-shaped patterns, tent-shaped patterns), skip-shaped patterns (left skip, right skip), bucket-shaped patterns and miscellaneous patterns.
The fingerprint morphological features include the center (top and bottom) and triangle points (left and right), etc. The detail feature points of fingerprints mainly include the starting point, end point, union point and bifurcation point of the lines. The fingerprint feature point information (end point, bifurcation point...) is extracted from the pre-processed image. ....) The information mainly includes parameters such as type, coordinates, and direction. The detailed features in fingerprints usually include endpoints, bifurcation points, isolated points, short bifurcations, rings, etc. The endpoints and bifurcation points are the most frequent, stable and easily harnessed in fingerprints. These two types of feature points can be matched to the fingerprint features: calculate the similarity between the feature extraction result and the stored feature template.
4、Fingerprint matching
Fingerprint matching is to compare the fingerprint features collected in the field with the fingerprint features stored in the fingerprint database to determine whether they belong to the same fingerprint. A coarse match can be made based on the pattern shape of the fingerprint, and then an exact match can be made using the fingerprint morphology and detail features to give the similarity score of the two fingerprints. Depending on the application, the similarity score is ranked or a verdict is given as to whether the fingerprints are the same
There are two ways of fingerprint comparison:
(1) one-to-one comparison: the user fingerprint to be compared is retrieved from the fingerprint database based on the user ID and then compared with the newly collected fingerprint;
General use of fingerprint scanners in airports
Fingerprint scanners are also used in airports today for a variety of purposes, such as notifying fingerprint permissions to give staff access to the airport, replacing traditional manual checks, using fingerprints to authenticate personnel, using fingerprint logins to log into computers, and querying database information. Performing entry management
Although fingerprint identification technology is mature, it is also constantly innovating, and the concept and industry of fingerprint identification scanning technology is still evolving.
