Fingerprint scanners are well known and used because of their accuracy and uniqueness. We all know that fingerprints are a unique feature of the human body and are complex enough to provide enough features for identification. To increase reliability, it is only necessary to register more fingerprints and identify more fingers, up to 10, and each fingerprint is unique. To register a 10-finger fingerprint device, we usually choose the bio7 plus 10-finger fingerprint device.

This fingerprint device HFSecurity Bio7 Ten Fingerprint Scanner Device, support 4-4-3 fingerprint collect

You can click here for more information

 

Fingerprint scanning is fast and very easy to use, usually within 0.1S to complete registration, identification, and comparison. 

 

Because fingerprint accuracy and speed are the key reasons why fingerprint scanners are used and recognized, if the fingerprint scanner does not meet these two requirements, then the existence of the fingerprint scanner will be a problem rather than a solution.

 

Here we will go one step deeper to understand how accurate fingerprint scanners really are?

First let's understand a term that is exclusive to fingerprint scanners

Fingerprint matching

 

In fingerprint matching, accuracy is the core key

 

The streaks that rise upward in a fingerprint are called "augmentation lines". Some of these lines are broken and divergent. The break is called an "endpoint" and the bifurcation is called a "bifurcation point". These two points are known as "characteristic points". In general, the center of a fingerprint contains about 50 points. The "position" and "direction" obtained from these feature points are the basic data for comparison.

It is possible that only the position and direction of the feature points are matched, but it is possible that they may coincidentally match.

 

1:N Matching Use the "input fingerprint" to match all the registered fingerprint data to find a consistent fingerprint pattern.

1:1 Matching The ID number must be entered first in order to match individual fingerprints among all the registered fingerprints.

The security and accuracy of the fingerprint scan may cause you to fail even if you are in person. This is because the fingerprint recognition program did not find enough details to recognize your fingerprint information, because they will ask you to rescan.

 

The fingerprint is too wet, if it is too wet you need to wipe off the water agent on the fingerprint with a paper towel

If the fingerprint is too dry, you can apply a light hand cream to make your finger moist.

 

Whether the fingerprint is cracked or there is a wound

 

If none of the above, you can try several times. In winter, the fingerprint recognition program will be a little slower to start and needs to be heated to start the program.

Accuracy consists of many factors, such as the location of the scanner, the lifetime of the scanner, the software the scanner can run and the quality of the user's own fingerprints. All of these factors affect the quality of the fingerprint capture and therefore the accuracy of the quality. So when purchasing a fingerprint scanner, we need to be very clear about the environment we are using it in, and the applicable user groups. Convenient, or rechargeable, only when we have determined our needs, we can choose the right fingerprint scanner.

The following table is the navigation chart of Huifan's fingerprint scanner. Huifan Technology has 16 years of biometric manufacturing experience, manufacturing optical fingerprint scanner, capacitive fingerprint instrument, ten fingerprint scanner

 

Click here for more samples

Fingerprint recognition error

 

From the user level, the device I registered with my left finger, the right finger will not be recognized because each fingerprint is different, so this is the goal we need to achieve

But in practice, we may find that a finger that is not registered will be registered successfully and a finger that is registered will be recognized instead. 

False acceptance rate (FAR)

False rejection rate (FRR)

False acceptance rate (FAR) is the percentage of different fingerprints that match with a score greater than a given threshold when the fingerprint recognition algorithm is tested on a standard fingerprint database and are thus considered to be the same fingerprint.

It is simply the percentage of fingerprints that are considered to be the same fingerprint when tested on a standard fingerprint database.

 

The rejection rate (FRR) is the percentage of identical fingerprints with a match score below a given threshold when testing the fingerprint recognition algorithm on a standard fingerprint database, and thus are considered to be different.

It is simply the percentage of fingerprints that are considered as unmatched when they should match each other successfully.

 

 As an example. 

 

Suppose there are 110 people and 8 fingerprint images of each person's thumb for a total of 110*8 = 880 fingerprints in the database, i.e. 110 categories with 8 images per category. Of course, we want

We want any two images within a class to match successfully, and any image between classes to fail. Now we let each image in the library match with all other

Now we let each image in the library match all images except itself, and calculate the false recognition rate and rejection rate respectively.

False recognition rate (FAR):

assume that due to the performance of the fingerprint recognition algorithm, a match that should have failed is judged to be a successful match, if we assume that the number of such errors is 10

00 times. In the theoretical case, the images from the same fingerprint are successfully matched, the number of times is 7*8*110=6160 times, and the total number of matches, i.e. 880×(880-1)=7

73520 times. The number of failed matches should be 773,520 - 6160 = 767,360 times. Then the false recognition rate FAR is 1000/767360*100%=0.13%.

False recognition rate and rejection rate of each fingerprint identification scanner will exist this situation, of course in the industry will have a unified a standard, FAR less than 0.0001%, FRR less than 1%

So, how accurate are today's fingerprint scanners?

Let us have look case

 

Fingerprint Scanner EVOTING

SIM Card Registration

Biometric Banking

GHANA Household Registraction

Farming Management