Whether it is shou machine or door lock field, semiconductor fingerprint identification devices have been widely used. In the shou machine industry, most of the high-, medium-, and low-end shou machine fingerprint identification applications have become very common. In the field of smart door locks, fingerprint recognition also occupies most of the production and sales, semiconductor fingerprint identification devices. It can be described as half of the smart door lock market.
Up to now, the identification of semiconductor applications in the field of smart door locks in China has achieved rapid development in 2016, with an average annual growth of more than 100%. By the end of 2017, the number of semiconductor fingerprint identification devices in the field of smart door locks in China has exceeded 4 million sets. In 2018, the demand will exceed 800 sets, and the total demand in the next five years will exceed 30 million sets. It is the market with the largest growth and retention of semiconductor fingerprint identification devices in the world.
However, most of the current market distributors and consumers, and even some skilled workers in the production of smart door locks, do not understand the working principle and mechanism of semiconductor fingerprint identification devices, and are easily exploited by some bad phenomena. There is a problem with product selection, and even cause consumer distrust.
What is semiconductor fingerprint recognition?
It is necessary to use a potential difference between the semiconductor (capacitance) and the hand-fingerprint road bump to obtain a fingerprint identification image or feature point, and to identify it accordingly. Then it is not difficult to understand that the operation of the semiconductor fingerprint identification device should be divided into two parts, one part is hardware, including semiconductor device (sensor), processing chip, memory chip, part is software part, core is algorithm. The overall four parts are the main ones.
The semiconductor device (sensor) is responsible for sensing the potential difference, the processing chip is responsible for computing and judging the obtained picture or feature, and the memory chip is responsible for storing relevant data and algorithms.
The core part is the sensor, its good and bad directly determine the efficiency and accuracy of the recognition on the hardware level; the second is the processor, which is the central part of the entire semiconductor fingerprint identification device, the processor determines the speed of the algorithm, The efficiency of recognition is also the size of the smart door lock power; the last is the memory.
But the three are co-prosperous and symbiotic, the sensor is no better, if the processor is not working, it will not achieve the application effect, the first two are excellent, but the memory can not store more data, and the overall effect is also dazed.
The software algorithm is in the position of the conductor on the whole semiconductor fingerprint identification device. The hardware is the foundation, and the algorithm is an important embodiment of the function and experience. The so-called fingerprint identification algorithm refers to preprocessing the captured fingerprint image in the process of fingerprint recognition. , data feature extraction, feature matching, and a series of clear instructions to solve the problem.
Among the whole fingerprint identification algorithms, including three parts, the first is pre-processing, mainly to improve the fingerprint input to improve the accuracy of feature extraction; the second is fingerprint feature extraction: refers to a specific area or feature The information is extracted, thereby removing a large number of pseudo features, improving the efficiency and accuracy of the recognition; finally, matching, that is, based on the above two points of retention information, after the fingerprint is entered, the matching of various features and images is performed.
The way the semiconductor fingerprint identification device is connected with the fingerprint lock
Currently, in the field of fingerprint door lock applications, the way in which the semiconductor fingerprint identification device and the fingerprint door lock are connected is mainly divided into the following two types, one is relatively independent in the form of a kit, and the other is related to each component of the electronic solution. integrated.
Relatively independent: the semiconductor fingerprint identification device exists in the form of a kit in the fingerprint door lock. The three main components of the sensor, the memory and the processor are independent of the electronic solution of the fingerprint door lock, only through the data line and the fingerprint door lock. The electronic parts are connected, and the application of fingerprint recognition is relatively independent.
The advantage of this is that it is adaptable, easy to maintain, easy to update, reduces the difficulty of electronic development, and forms a modular application; the disadvantage is that the integration is not high, the volume is relatively large, and the cost is relatively high.
The integration scheme is: the semiconductor fingerprint identification device exists in a relatively integrated manner among the fingerprint door locks, and only the semiconductor sensor is connected to the PCBA main board of the electronic solution by means of plugging or data line connection, that is, integrating the processor, the memory, and the like in the Among the framework of the overall electronic solution, fingerprint recognition has been dissolved in the overall scheme, and the fingerprint identification device is only a part of the electronic scheme.
The advantages of this approach are: product quality is easy to control, high integration, relatively low cost, and easier to update; disadvantages are: increased development difficulty, unable to form modular applications, poor adaptability, the same type of fingerprint identification device Cannot match exactly with other electronic schemes.
Two kinds of connection methods, among which the first type of connection of the fingerprint door lock is loved by the manufacturer, and the integration scheme only accounts for a very small part, but most of the integrated production companies are technically strong or have Very fixed supplier of production companies, and the scale is relatively large.
The security status of fingerprint algorithm in fingerprint lock
The security core of the electronic part of the fingerprint door lock is in the fingerprint identification device. Whether it is the recognition efficiency, the accuracy of the recognition, the accuracy of the recognition, the difficulty of being cracked, etc., the fingerprint lock is determined. The ultimate product quality and safety. In addition to the fingerprint recognition device being good enough in hardware, these require a very high level of algorithmic knowledge.
The types of fingerprint algorithms, the current algorithms for fingerprint locks in China mainly include three types:
One type is an image-based feature recognition algorithm. The main principle is that after the pre-processing formed by the fingerprint on the sensor, the corresponding overall image is extracted, and the overall/partial image comparison is consistent/inconsistent in recognition. The type of algorithm to make the decision, but because the image will change, or the image is difficult to judge when the face is too small, generally such algorithms have self-learning functions to solve the drawback of accepting too small, so such algorithms Applicable to small area sensors, derived from the shou machine fingerprint recognition algorithm, so the overall cost is lower.
The other is a feature recognition algorithm based on fingerprint feature points. The principle is that, after the fingerprint image is preprocessed by the sensor, the corresponding feature points on the extracted fingerprint image are compared, and the features on the image are compared during recognition. The type of algorithm that makes a consistent/inconsistent decision. This algorithm does not recognize the image because it changes. As long as the feature exists, it can be identified. However, this algorithm is easy to recognize when the sensor area is too small. The reason for the low level is therefore not applicable to products with too small sensor area, so the overall cost is high.
The third type of algorithm based on the two types of images and feature points developed by recent companies focused on fingerprint recognition algorithms is based on the combination of the advantages of the two, using both features and images. The advantages are identified and the overall cost is in the middle position.
Security, from the working principle, we can easily analyze that the security based on image recognition is relatively poor, because the image is easy to be copied, and since most of the image recognition based sensor area is too small, in order to experience better, most Has a self-learning function. This also led to the appearance of the fingerprint lock of the orange peel solution of the product in the previous stage.
It is the deficiencies of the image algorithm, the large-scale image comparison, and self-learning function, the "orange peel" can also be "fingerprinted" through the conductive pen and conductive paper, but because the fingerprint recognition area has been Small, the experience is not good, and the fingerprint door lock application of image algorithms is not many companies, only a very small number of low-priced brands have been affected, but this is also to make us wary, can not sacrifice security because of cost and experience.
Feature-based security is very good because it is based on feature information for identification, and feature information is generally not easy to change, and from the principle we also conclude that the transmitter area is large, making the fingerprint The difficulty of the geometrical growth is greatly increased, which greatly improves the security performance of the identification. It is basically impossible to realize the so-called "orange peel" unlocking, which is also the mainstream algorithm used in the current fingerprint lock production enterprises. Most of the branded fingerprint locks were not affected by this incident.
The image + feature recognition algorithm based on the advantages of both is currently the most promising prospect. The former is low in price, but the security is worrying. Second, security is no problem, but the overall cost is high, with the market. And the development of the industry, the future will be more focused on high-end products, and the algorithm based on the advantages of the two, on the one hand, takes advantage of the former's speed, efficiency and cost reduction, on the other hand, it also has the security of feature recognition. The most important thing is that the overall cost will also fall reasonably, and it may become one of the development trends in the future.
The relationship between fingerprint door lock product experience and security of fingerprint identification device
Fingerprint door lock is a product with heavy experience. It is a high-frequency use, high-frequency experience product. Basically, at least ten times per day, zui is even higher than 30 times, so the use of fingerprint door locks The experience determines the desire of the consumer to buy.
The core and most critical part of the fingerprint door lock experience is on the fingerprint recognition device. This is the place where the most frequent contact and the experience of the consumer are used during the use of the fingerprint. Therefore, the experience of the fingerprint recognition device basically determines the fingerprint door. The sense of experience of the lock product.
The experience and security of the fingerprint identification device is a dialectical relationship between hardware and software. The experience is good and the security performance is good. The general cost will be extremely high, because the sensing area is required to be larger, the algorithm is higher, and the memory is larger. Some, faster, faster processor speed, so the efficiency experience is good, but it will cause a sharp rise in cost and power consumption, rising to the point where consumers can't accept it. Imagine fingerprint locks changing battery consumers one or two weeks. Can you accept? Or in order to reduce power consumption, the algorithm is easier to pass, and other performances remain certain. Although the experience is better, it can reach 0.x seconds to open the door, but the security performance is also reduced a lot, and the security fingerprint lock cannot be guaranteed. Can it be a safe lock?
Therefore, in general, all major fingerprint identification device developers have adopted the third method, using the dialectical relationship of the three, while improving the level of algorithms, sensors, processors, and memories under the premise of setting power consumption. In order to achieve the ultimate goal of the use of fingerprint door lock products, and ultimately reach the level of consumer satisfaction.
Yanmar Cylinder Head
Cylinder Head Porting,Yanmar Cylinder Head,3Tne68 3Tna68 Full Gasket Set,2Tn68 3Tn66 Full Gasket Set
Foshan Evictory Diesel Spare Parts Co., Ltd. , https://www.evictoryparts.com