A brief analysis of the image acquisition technology principles of the four major fingerprint recognition

After nearly 10 years of slow natural growth, fingerprint recognition technology is about to usher in a golden period of rapid development. Experts conservatively estimate that in the next five years, my country will have a market of nearly 10 billion yuan waiting for enterprises to explore. At present, fingerprint recognition is mainly used in the fields of attendance, access control, safes, etc. It is believed that with the improvement of fingerprint recognition technology, it will be widely used in more fields such as ID cards, motor vehicles, and homes. So what are the main technologies for fingerprint recognition image acquisition?

Fingerprint recognition

Fingerprint recognition technology is a high-tech technology that utilizes the uniqueness and stability of fingerprints and integrates sensor technology, biotechnology, digital image processing, pattern matching, and electronic technology.

Development Trend of Fingerprint Recognition Technology

The huge market prospect of fingerprint recognition technology will have a huge impact on the international and domestic security industry. Smaller companies will face ruthless competition from new entrants of large traditional industry companies. In front of these giants, it is difficult to say that existing small and medium-sized companies have much competitiveness. Industry reshuffle is inevitable, and mergers and exits may become the helpless choice of most small and medium-sized companies. In the end, it may form a situation where companies in traditional industries or large capital dominate the biometric recognition industry in a relatively short period of time. Of course, it is not ruled out that some companies with core competitiveness will grow into emerging large companies in the way of "fast fish eating slow fish". The coexistence of "big fish" and "fast fish" is also the inevitable result of every emerging market. And the result of competition will form a new large industry.

The civilian market opportunities for fingerprint recognition technology already exist, so is it time for large-scale application? This depends on whether the consumer awareness, product stability and price of fingerprint products can be solved. In other words, whether fingerprint recognition technology can be applied on a large scale in the civilian field requires three conditions: sufficient audience awareness, stable product quality and good service and guarantee, and sufficient product profit.

Consumer awareness of fingerprint products mainly refers to people's understanding of the application of fingerprint recognition technology, including manufacturers, distributors, contractors and direct users. An interesting example is that more than 80% of the requirements for access control that require fingerprints are from direct users, while most access control manufacturers, distributors, contractors and other intermediaries do not actively recommend products. This is the biggest problem that fingerprint recognition cannot be applied on a large scale, and this is a reflection of low consumer awareness.

The stability of the product requires fingerprint manufacturers to have a deeper understanding of the security industry. The fingerprint products they develop must comply with usage specifications and meet most requirements as well as service guarantees.

Product price is very important, especially in the product development stage. Some users will say that even if the product is stable, it will not be used if the price is not right! The analysis shows that taking the mid-range route, playing to one's strengths and avoiding weaknesses, and giving full play to the absolute competitive advantage of the product in the market terminal is the fundamental to being invincible in the civilian market.

Fingerprint recognition technology type

There are four main types of fingerprint image acquisition technologies: optical scanning devices (such as micro prism matrix), temperature difference sensing fingerprint sensors, semiconductor fingerprint sensors, and ultrasonic fingerprint scanning.

Optical recognition technology

Fingerprint collection with the help of optical technology is the oldest and most widely used technology. Place your finger on an optical lens, illuminate it with a built-in light source, and use a prism to project it onto a charge-coupled device (CCD), thereby forming a digital multi-grayscale fingerprint image that can be processed by the fingerprint device algorithm, with ridges (lines with a certain width and direction in the fingerprint image) in black and valleys (the concave parts between the lines) in white.

Optical fingerprint collection technology has obvious advantages: it has been tested for a long time, can adapt to temperature changes to a certain extent, can achieve a high resolution of 500DPI, and most importantly, is inexpensive. It also has obvious disadvantages: it requires a sufficiently long optical path, so it requires a sufficiently large size, and overly dry or overly greasy fingers will also deteriorate the effect of optical fingerprint products.

The limitation of optical fingerprint sensing is reflected in the latent fingerprint (latent fingerprint is left after the finger is pressed on the platen), which not only reduces the quality of the fingerprint image, but also may cause two fingerprints to overlap in serious cases. Obviously, it is difficult to meet the needs of practical applications. In addition, the platen coating and CCD array will wear out over time, which may cause the quality of the collected fingerprint image to deteriorate. However, it has the disadvantages of being unable to identify live fingerprints and poor applicability to dry and wet fingers.

Since light cannot penetrate the surface layer of the skin (dead skin layer), the optical fingerprint recognition system can only scan the surface of the finger skin, or scan the dead skin layer, but cannot penetrate into the dermis. In this case, the cleanliness of the finger surface directly affects the recognition effect. If there is a lot of dust on the user's fingers, recognition errors may occur. In addition, if people make a fingerprint hand model according to their fingers, it may also pass the recognition system, which is not very safe and stable for users to use.

For example, in the prison break case in Inner Mongolia at the end of last year, the escaped prisoners cut off the prison guard's finger to verify the optical fingerprint machine to open the prison door. Recently, major media have reported that you can buy a custom silicone fingerprint on Taobao for about 100 yuan, which can be easily verified by the optical fingerprint machine. Office workers use it to check attendance.

In addition, there are prisms in optical sensors, which are relatively large, usually several times or even 10 times the size of semiconductors, so their application in small devices is limited. There is no problem of volume limitation when using them on large devices such as attendance machines and access control, but when using them on USB flash drives, mobile hard drives, and handheld devices, volume becomes the biggest obstacle. Low cost has always been considered the biggest advantage of optical sensors, but because it is difficult to ensure the consistency of their manufacturing process, with the large-scale development of semiconductor sensors represented by capacitive sensors, the cost advantage of optical sensors is no longer obvious. Although most companies are still using optical sensors, the development trend is towards novel, high-quality semiconductor capacitive fingerprint sensors.

Temperature difference sensing recognition technology

Temperature difference sensing recognition technology is based on the principle of temperature sensing. Each pixel is equivalent to a miniaturized charge sensor, which is used to sense the temperature difference between the finger and a certain point in the chip imaging area, and generate an electrical signal representing the image information.

Its advantages are that it can obtain fingerprint images within 0.1 seconds, and the sensor size and area are the smallest, that is, the sliding fingerprint recognition device commonly used at present adopts this technology. The disadvantage is that it is subject to temperature limitations, and after a long time, the finger and the chip are at the same temperature.

Semiconductor silicon sensing technology (capacitive technology)

In the late 1990s, the technology based on the semiconductor silicon capacitance effect matured. The silicon sensor becomes one plate of the capacitor, and the finger is the other plate. The capacitance difference between the ridges and valleys of the fingerprint line relative to the smooth silicon sensor is used to form an 8-bit grayscale image. The capacitive sensor sends out an electronic signal, which will pass through the surface and dead skin layer of the finger, directly reach the living layer (dermis) of the finger skin, and directly read the fingerprint pattern. Because it penetrates deep into the dermis, the sensor can capture more real data, is not easily affected by dust on the finger surface, improves recognition accuracy, and effectively prevents recognition errors.

Semiconductor fingerprint sensors include semiconductor pressure-sensitive sensors, semiconductor temperature-sensitive sensors, etc. Among them, the most widely used is the semiconductor capacitive fingerprint sensor.

The semiconductor capacitive sensor determines which position is the ridge and which position is the valley based on the different capacitance values ​​formed by the ridges and valleys of the fingerprint and the semiconductor capacitive sensing particles. Its working process is to pre-charge the capacitive sensing particles on each pixel to a certain reference voltage. When the finger touches the semiconductor capacitive fingerprint display, because the ridges are raised and the valleys are concave, different capacitance values ​​will be formed at the ridges and valleys according to the relationship between the capacitance value and the distance. Then the discharge current is used for discharge. Because the capacitance values ​​corresponding to the ridges and valleys are different, the discharge speed is also different. The pixels under the ridges (high capacitance) discharge slowly, while the pixels under the valleys (low capacitance) discharge quickly. According to the different discharge rates, the positions of the ridges and valleys can be detected, thereby forming fingerprint image data.

Unlike optical devices that mostly use manual adjustments to improve image quality, capacitive sensors use automatic control technology to adjust fingerprint image pixels and the sensitivity of the local range of the fingerprint, and generate high-quality images in combination with feedback information in different environments. Due to the local adjustment capability, even images with poor contrast (such as areas where the finger is pressed lightly) can be effectively detected, and the sensitivity of these pixels can be increased at the moment of capture to generate high-quality fingerprint images.

The advantages of semiconductor capacitive fingerprint sensors are good image quality, generally no distortion, small size, and easy integration into various devices. The electronic signal it sends will pass through the surface and dead skin layer of the finger, reach the living layer (dermis) of the finger skin, and directly read the fingerprint pattern, thereby greatly improving the security of the system.

The most important advantage of semiconductor silicon sensing technology is that it can achieve live fingerprint recognition. It can also achieve better image quality than optical technology on a smaller surface, and obtain a resolution of 200-300 lines on a surface of 1cm×1.5cm (a smaller surface also leads to a reduction in cost and can be integrated into smaller devices). It is small in size, low in cost, high in imaging accuracy, and consumes very little power, so it is very suitable for use in security and high-end consumer electronic products, and is called the second generation of fingerprint recognition technology after optics.

Semiconductor capacitive fingerprint sensors have complex manufacturing processes, many sensing units per unit area, and high-end IC design technology, large-scale integrated circuit manufacturing technology, IC chip packaging technology, etc. Therefore, capacitive fingerprint sensors are almost all designed and manufactured by countries or regions with advanced IC technology, such as the United States, Europe, Taiwan, etc. At present, only a very small number of manufacturers in China are capable of producing semiconductor fingerprint sensors.

However, semiconductor silicon sensing technology also has disadvantages, that is, it will be affected by static electricity, but this can be solved by grounding during installation. In the past, the cost was relatively expensive, but in recent years, the cost has dropped significantly, and it is getting closer to the cost of optical sensors. It is currently the most ideal fingerprint recognition technology. For example, in the security access control system of high-risk security places such as bank vaults and prisons, fingerprint machines using semiconductor silicon sensing recognition technology are used for live fingerprint recognition at the front end of the access control, replacing traditional passwords, card swiping, and optical fingerprint machines, so as to truly achieve the uniqueness of identity recognition and ensure foolproofness.

Ultrasonic technology

Ultrasonic fingerprint collection is a new technology that uses the ability of ultrasound to penetrate materials and produce echoes of different sizes depending on the material (when ultrasound reaches the surface of different materials, it is absorbed, penetrated and reflected to different degrees). Therefore, by using the difference in acoustic impedance between skin and air, the location of fingerprint ridges and valleys can be distinguished.

The ultrasonic frequency used by ultrasonic technology is 1×104Hz-1×109Hz, and the energy is controlled to a level that is harmless to the human body (the same intensity as medical diagnosis). Ultrasonic technology products can achieve the best accuracy, and they have lower requirements for the cleanliness of fingers and surfaces, but their collection time is significantly longer than the above two types of products, and they are expensive and cannot achieve live fingerprint recognition, so they are rarely used at present.