chrispeng [ad_1]
Development background:
In order to strengthen management, a company hopes to use fingerprint recognition technology for employee attendance in its factory. Among the many biometric technologies used for identity verification, fingerprint recognition technology is currently the most convenient, reliable, non-invasive and inexpensive solution. Fingerprint recognition has a long history as an identification technology. This technology analyzes the global and local features of the fingerprint, and the feature values extracted from the fingerprint can be very detailed in order to reliably confirm the identity of a person through the fingerprint. On average, each fingerprint has several unique and measurable feature points, and each feature point has about seven features. Our ten fingers produce at least 4900 independently measurable features-this is enough to confirm whether fingerprint recognition is A more reliable way of identification. Fingerprints are unique features of the human body, and their complexity is sufficient to provide sufficient features for identification; if we want to increase reliability, we only need to register more fingerprints and identify more fingers, up to ten , And each fingerprint is unique; scanning fingerprints is fast and easy to use; when reading fingerprints, the user must touch the fingerprint collection head with each other, and direct contact with the fingerprint collection head is the most important way to read the biological characteristics of the human body. Reliable method. This is also one of the main reasons that fingerprint recognition technology can occupy most of the market. The estimated number of employees in the company’s factory is 3-5 thousand. There are 3 gates and 10 small doors in the factory. The fingerprint attendance points of employees are distributed at the entrance of each small door, and the number of attendance at each small door is about 400. That is, this system solution needs to complete the daily fingerprint attendance management of 3-5 thousand employees.
Main technical difficulties and their solutions:
1. Large number of people
1. Fingerprint recognition algorithm
Although Biokey is currently the fastest fingerprint recognition algorithm, if 4,000 people register 2 fingerprints per person, the number of fingerprints that need to be registered in the entire system is 8,000. The average verification time for attendance on a Pentium Ⅲ 800 computer It reached 3 seconds, and the verification time of attendance failure reached 5 seconds, which is obviously unacceptable, so necessary measures must be taken to solve this problem. When checking attendance, people can accept a verification time of up to 3 seconds, so the search range of a verification terminal when verifying fingerprints cannot exceed 4000 fingerprints.
Method 1: In a certain verification terminal, limit the number of searchers to an acceptable range. For example, only employees of a certain branch are allowed to sign in with their fingerprints on the attendance machine in front of the factory, but they cannot be recognized when signing in elsewhere. Under normal circumstances, employees only need to check attendance in their own branch, and can only enter and leave their own branch. Therefore, this method can be applied to a branch with less than 2,000 employees, and it is feasible to check attendance. However, for the attendance management at the gate, every branch employee may use it, so this method cannot be used.
Method 2: Enter the user ID before comparing fingerprints. In this way, the system only needs to compare the registered fingerprint with the job number to verify the identity of the employee, which is a one-to-one comparison, so there is no limit to the number of employees that can be verified here. The disadvantage of this method is that the employee’s work number must be entered, which will increase the employee’s memory burden and operating time during use.
2. Staff operation
Since the commuting time is relatively concentrated, many employees will be gathered for attendance when commuting, and it must be ensured that they can complete the attendance in a short time, and there should be no long queues. Everyone’s attendance time cannot be simply calculated as the time for the computer to compare fingerprints, because employees also need to go to the attendance machine and press their fingerprints. In addition, fingerprints may not be recognized at one time, or employees may not perform improper operations, resulting in identification failures, etc. According to actual experience, the average passing time of each employee is 3 seconds. Taking 20 minutes as an estimate of the attendance time, an attendance point can accommodate 450 people to attend attendance at the same time. Taking into account the existence of emergencies, each attendance point can only be counted as 400 people.
For branch factories with more than 400 people, it is recommended to set up the number of attendance points according to the actual number of people. For example, a branch factory with 1000 people commuting at the same time should set up 3 attendance points, or a computer connected to 3 attendance terminals. A branch factory with less than 400 employees can set up only one attendance point.
2. Unreliability of fingerprint recognition technology
Fingerprint recognition technology is not 100% reliable. There are two pointers “rejection rate” and “false recognition rate” to characterize the reliability of a fingerprint recognition system:
Rejection rate: refers to the percentage of probability that the system will not recognize a registered fingerprint.
False recognition rate: refers to the percentage of the probability that the system mistakenly recognizes a fingerprint as other fingerprints.
The Biokey fingerprint recognition technology of Supcon Technology is one of the most reliable fingerprint recognition technologies at present, with a nominal rejection rate of 1.0% and a false recognition rate of 0.001%, which means that among the 4000 employees, there are about 50 employees Fingerprint recognition is difficult to pass. At the same time, if a branch factory with a maximum of 4,000 employees is used, a misjudgment may occur once a month.
The rejection rate and the false admission rate are two indicators that restrict each other. If one is lowered, the other will rise. Since the false recognition rate is a very critical indicator, we must ensure that there will not be too many misjudgments for at least one month. When a misjudgment occurs, the administrator must modify the record in the management program to ensure the correctness of attendance.
Entering the number and using the card in combination is the most effective way to solve the system’s identification performance. When comparing fingerprints, it only compares with your own fingerprints, so there is no possibility of false recognition. In order to prevent attendance on behalf of others, the false recognition rate can be adjusted to about 0.1%, and the rejection rate at this time is 0.6%. The number of people in difficulty was reduced to 24.
Three, summary
If there is a dedicated person in charge of attendance statistics, then only fingerprints can be used for attendance. For large factories and enterprises, this is the preferred solution. If misjudgment is not desired, the ID+fingerprint method can be used for attendance.
[ad_2]
