TWI799092B - Identification method of identification system - Google Patents

Abstract

The invention relates to an identification method of an identification system. The identification system includes a sensing area and an image sensor. First, the object to be measured is close to the sensing area, so that the image sensor generates a dynamic image. Then, the object under test gradually pressurizes the sensing area. After that, the object to be tested completely covers the sensing area, and the image sensor further produces a perspective image. Finally, the identification module is used to determine whether the dynamic image is biological imaging or not according to the perspective image, and the dynamic image is subtracted, which is used as the basis for identifying whether the system is unlocked. In this way, the identification system of the present invention is combined with the identification method to achieve real-time determination of whether the dynamic image is biological imaging or not. At the same time, through the recognition method, the dynamic image is subtracted, which greatly improves the recognition system's False Acceptance Rate (FAR) and False Rejection Rate (FRR), achieving high accuracy and wide application.

Description

Identification method of the identification system

The present invention relates to an identification system, in particular to an identification system and its method.

Since the development of fingerprint identification technology, it has become the standard equipment of most smart phones. The advantage of fingerprint identification is that fingerprints are unique characteristics of the human body, and the complexity of fingerprints is sufficient for identification. Multiple fingerprints and identify more fingers, up to ten, each unique. Furthermore, the speed of scanning fingerprints is quite fast and easy to use, which is also a major reason why fingerprint recognition technology can occupy most of the market.

However, fingerprint identification is not absolutely safe. People leave their fingerprints in many public places every day. If someone really wants to obtain and copy them, it is very easy to obtain fingerprint information. Once the fingerprint is restored, personal devices and information security may be stolen. At the same time, compared with using a password, even if the password is cracked, it can still be reset, but the fingerprint cannot be reset. In view of this, how to improve the security and identification ability of fingerprint identification is one of the problems that researchers should solve.

In addition, the identification ability index in fingerprint identification technology is an important index. The identification ability index represents an index for evaluating or comparing the performance of biometric security systems, including acceptance error rate (False Acceptance Rate, FAR) and rejection error rate ( False Rejection Rate, FRR). It needs to be further explained that the acceptance error rate indicates the probability that the biometric system mistakenly identifies an illegal user as a legitimate user, that is, the security level of the biometric system; the rejection error rate indicates that the biometric system misjudges a legitimate user as an unauthorized user. The probability of legitimate users, that is, the convenience of the biometric system. Therefore, how to reduce the acceptance error rate and increase the rejection error rate is one of the issues to be solved by the researchers.

Therefore, the inventor of this case came up with the present invention after observing the above-mentioned deficiency.

The purpose of the present invention is to provide an identification method, which uses an image sensor to generate a dynamic image within a time range for the object to be tested. The dynamic image includes a plurality of interval images, wherein the time range includes a plurality of exposure intervals, and when the time The images in these intervals are clearer when the range is further back, so that the image sensor system can further generate a perspective image, and judge whether the dynamic image is a biological image according to whether the dynamic image has a change process and the perspective image. In this way, Effectively prevent others from cracking the identification system with fingerprint images, pictures, or arbitrary models, and greatly increase the security and identification capabilities of the identification system.

Another object of the present invention is to provide an identification method, which uses a processing module to execute an algorithm, which subtracts dynamic images in different exposure intervals from each other, so that the RV value of the dynamic image according to the present invention ( ridge valley value) has been greatly improved, resulting in better fingerprint recognition results. Thereby, the Accepted Error Rate (FAR) and Rejected Error Rate (FRR) of the identification system according to the present invention are greatly improved, achieving the goals of high accuracy and wide applicability.

To achieve the above purpose, the present invention provides an identification method, which is applied to an identification system, the identification system includes a sensing area, an image sensor and an identification module, the identification module is coupled to the sensing area and the image sensor, the identification method includes the following steps: a starting step, when the object to be tested touches the sensing area, the image sensor is activated and generates a dynamic image for the object to be tested; a sensing A detection step, the object under test covers the sensing area from contact, so that the image sensor further generates a perspective image; and an identification step, if the object under test has a change process of the dynamic image and the perspective image, the identification module determines that the dynamic image is a biological image; otherwise, the identification module determines that the dynamic image is a non-biological image.

Preferably, according to the identification method of the present invention, wherein the dynamic image includes a plurality of interval images generated by the object under test within a time range, the identification module judges the Whether the dynamic image has the change process, when one of the sharpness values exceeds the threshold, the identification module determines that the dynamic image has the change process.

Preferably, in the identification method according to the present invention, the sharpness value is calculated through one of the image difference method and the image gradient value.

Preferably, according to the identification method of the present invention, the image sensor further includes: a plurality of photosensors arranged in an array, the photosensors are used to generate complex image intensity information, and The dynamic image is generated by the image intensity information; a plurality of complementary metal-oxide semiconductors (Complementary Metal-Oxide Semiconductor, CMOS), which are coupled to the light sensors, and the complementary metal-oxide semiconductors are used for The output of the image intensity information is controlled, but the present invention is not limited thereto.

Preferably, according to the identification method of the present invention, wherein, the image sensor is arranged below the sensing area, the image sensor has a shutter mechanism, and the shutter mechanism is used to control the exposure interval, but the present invention Not limited to this.

Preferably, according to the identification method of the present invention, the shutter mechanism is a global shutter (Global Shutter, GS), so that the photosensors are simultaneously exposed to generate the image intensity information, but the present invention is not limited to this.

Moreover, in order to achieve the above object, the present invention is based on the above identification system, and further provides an identification method for implementing the above identification system, which includes: an object under test approaches the sensing area, when the object under test touches the sensing area, When sensing the area, the image sensor is activated and generates a dynamic image for the object under test, the dynamic image includes a plurality of interval images; a subtraction step, a processing module executes an algorithm to compare the interval images Perform subtraction operations and generate complex subtraction signals; a signal amplification step, an operation module amplifies the subtraction signals by multiples, so that the peaks and valleys of the amplified subtraction signals are clear and within the signal processing range and an identification step, where an identification module uses the amplified subtraction signals as the basis for whether the identification system is unlocked.

Preferably, according to the identification method of the present invention, wherein, after performing the signal amplification step, the identification method further includes: an averaging step, the calculation module taking an average value of the subtracted signals; wherein, the identification The step further uses the identification module as a basis for unlocking the identification system according to the average value.

Preferably, the identification method according to the present invention further includes a selection step, the processing module uses one of the interval images as a background interval image; wherein, the subtraction step further combines the background interval image with A subtraction operation is performed on the interval images, and the subtraction signals are generated.

To sum up, the identification method provided by the present invention mainly utilizes the image sensor of the present invention to generate a dynamic image for the object under test within a time range, wherein the time range includes multiple exposure intervals, and the longer the exposure interval, the longer the The clarity value of the interval image exceeds the threshold, and the image sensor further generates a perspective image to judge whether the dynamic image is a biological image according to whether the dynamic image has the change process and the perspective image. In this way, it is effective to prevent others from using Fingerprint images, pictures, or arbitrary models crack the identification system, and greatly increase the security and identification capabilities of the identification system. In addition, through the processing module to execute the algorithm, the algorithm subtracts the dynamic images of different exposure intervals from each other and then takes the average value, so that the RV value (ridge valley value) of the dynamic image according to the present invention is greatly improved, resulting in Better fingerprint recognition effect. Thereby, the Accepted Error Rate (FAR) and Rejected Error Rate (FRR) of the identification system according to the present invention are greatly improved, achieving the goals of high accuracy and wide applicability.

In order to enable those skilled in the art to understand the purpose, features and effects of the present invention, the present invention will be described in detail below by means of the following specific embodiments and accompanying drawings.

The inventive concept will now be explained more fully hereinafter with reference to the accompanying drawings, in which exemplary embodiments of the inventive concept are shown. Advantages and features of the inventive concept and methods for achieving it will be apparent below by referring to the exemplary embodiments described in more detail with reference to the accompanying drawings. It should be noted, however, that the inventive concept is not limited to the following exemplary embodiments, but can be implemented in various forms. Therefore, the exemplary embodiments are provided only to disclose the inventive concept and to make one skilled in the art understand the category of the inventive concept. In the drawings, the exemplary embodiments of the inventive concepts are not limited to the specific examples provided herein and are exaggerated for clarity.

The terminology used herein is for describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the terms "a", "an" and "the" in the singular are intended to include the plural forms as well, unless the context clearly dictates otherwise. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may be present.

Similarly, it will be understood that when an element such as a layer, region or substrate is referred to as being "on" another element, it can be directly on the other element or intervening elements may be present. In contrast, the term "directly" means that there are no intervening elements. It should be further understood that when the words "comprising" and "comprising" are used herein, it indicates the existence of stated features, integers, steps, operations, elements, and/or components, but does not exclude one or more other features. , integers, steps, operations, elements, components, and/or the presence or addition of groups thereof.

Furthermore, exemplary embodiments in the detailed description will be explained by means of cross-sectional views that are idealized exemplary views of the inventive concept. Accordingly, the shapes of the exemplary figures may be modified according to manufacturing techniques and/or allowable errors. Accordingly, exemplary embodiments of the inventive concepts are not limited to the specific shapes shown in the exemplary figures, but may include other shapes that may be produced according to manufacturing processes. Regions illustrated in the drawings have general characteristics and are used to illustrate specific shapes of elements. Accordingly, this should not be seen as limiting the scope of the inventive concept.

It should also be understood that although the terms “first”, “second”, “third” etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish various elements. Thus, a first element in some embodiments could be termed a second element in other embodiments without departing from the teachings of the present invention. Exemplary embodiments of aspects of the inventive concept illustrated and illustrated herein include their complementary counterparts. Throughout this specification, the same reference number or the same designator designates the same element.

Additionally, exemplary embodiments are described herein with reference to cross-sectional and/or plan views, which are idealized exemplary illustrations. Accordingly, deviations from the illustrated shapes as a result, for example, of manufacturing techniques and/or tolerances are to be expected. Thus, example embodiments should not be construed as limited to the shapes of regions illustrated herein but are to include deviations in shapes that result, for example, from manufacturing. Thus, the regions shown in the figures are schematic and their shapes are not intended to illustrate the actual shape of a region of a device and are not intended to limit the scope of example embodiments.

Please refer to Fig. 1 to Fig. 4, Fig. 1 is a schematic diagram according to the identification system of the present invention; Fig. 2 is a block diagram illustrating the steps of implementing the identification method of the present invention; Fig. 3 is an illustration of the actual execution process of the identification method according to the present invention The flow chart of the steps. As shown in FIG. 1 , an identification system 100 according to the present invention includes: a sensing area 11 , an image sensor 12 , and an identification module 13 .

Specifically, according to the sensing region 11 of the present invention, it is used for the object under test 200 to approach to perform sensing. In some embodiments, the sensing region 11 can serve as an isolation structure of the identification system 100 . In the present invention, the term "isolation" covers both electrical isolation and physical isolation. The identification system 100 can be a single layer of inorganic packaging material, a multi-layer stack of inorganic packaging material, or a stack of paired inorganic packaging material and organic packaging material. The inorganic packaging material used is, for example but not limited to, silicon nitride (SiNx), silicon oxide (SiOx), silicon oxynitride (SiONx), aluminum oxide (AlOx), or titanium oxide (TiOx), but the present invention is not limited thereto .

Specifically, the image sensor 12 according to the present invention is disposed below the sensing region 11, the image sensor 12 has a shutter mechanism (not shown), and the image sensor 12 is used for targeting the object under test 200 A dynamic image 21 is generated within a time range 31, wherein the dynamic image 21 includes a plurality of interval images 23 (23-1 to 23-8 shown in FIG. 4 ), the time range 31 includes a plurality of exposure intervals 32, and the shutter mechanism uses The exposure interval 32 is controlled, and the exposure interval 32 and the interval image 23 correspond to each other. In some embodiments, the image sensor 12 according to the present invention may be a complementary metal oxide semiconductor (Complementary Metal-Oxide Semiconductor, CMOS) image sensor, and may be selected from a back-illuminated CMOS image sensor or One of the front-illuminated CMOS image sensors, but the present invention is not limited thereto.

It should be further explained that, in some embodiments, the image sensor 12 according to the present invention may have one of a rolling shutter (Rolling Shutter) mechanism and a global shutter (Global Shutter) mechanism. During the shutter mechanism, due to the exposure time difference of the rolling shutter mechanism, when the image sensor 12 shoots a dynamic image, the exposure time of the upper and lower half of the interval image 23 is different, which may cause the upper half of the interval image 23 to appear first. However, there is a time gap that has not yet appeared at the lower end of the section image 23 , which causes the image distortion of the section image 23 . In a preferred embodiment of the present invention, since the present invention is mainly aimed at generating the dynamic image 21 within the time range 31 of the object under test 200, in order to avoid the jello effect (Jello Effect) in the interval image 23 of the dynamic image 21, the global method is adopted The shutter mechanism, that is, each pixel of all the pixel arrays on the image sensor 12 will simultaneously obtain the interval image 23 in the same exposure interval 32 , but the present invention is not limited thereto.

Specifically, the recognition module 13 according to the present invention is coupled to the image sensor 12 , and the recognition module 13 judges whether the dynamic image 21 is a biological image according to the perspective image 22 . It should be further explained that, in some embodiments, the object to be tested 200 can be a fingerprint of a human body, and the perspective image 22 can be a vein of the human body or be associated with the vein of the human body, and the identification module 13 can be The perspective image 22 of finger veins is captured by the image sensor 12, and the perspective image 22 is used as a basis for judging whether the dynamic image 21 is a biological image, but the present invention is not limited thereto. In the present invention, the term "biological image" means preventing others from deciphering the identification system 100 with fingerprint images, pictures, or arbitrary models.

It is worth mentioning that, in some embodiments, the identification system 100 according to the present invention can execute an algorithm (not shown in the figure) to perform a subtraction operation on the interval images 23 of different exposure intervals 32, so that according to the present invention The RV value (ridge valley value) of the dynamic image 21 is greatly improved, resulting in a better fingerprint recognition effect. In addition, the subtraction operation described in this article can refer to subtracting one of the interval images 23 One is the calculation method of reducing the noise value. In a preferred embodiment of the present invention, the subtraction operation is to subtract the interval image 23 of the previous exposure interval 32 from the interval image 23 of the subsequent exposure interval 32, and the obtained values are averaged after the subtraction operation is completed. operation. It can be understood that, in the conventional fingerprint recognition system, the interval images 23 of different exposure intervals 32 are only added to each other to obtain an average value, or the interval images 23 are compared after only subtracting the empty background value. After the addition operation, the average value is taken. However, due to the large gap between the deducted background value and the interval image 23, the noise in the interval image 23 cannot be effectively deducted. As a result, when the image in the interval image 23 is enlarged, the noise will be are amplified simultaneously. To sum up, the identification system according to the present invention executes an algorithm to subtract the interval images 23 of different exposure intervals from each other, which greatly improves the acceptance error rate (FAR) and rejection error rate of the identification system according to the present invention (FRR) to achieve high accuracy and wide applicability.

In order to further understand the structural features, technical means and expected effects of the present invention, the use of the present invention will be described hereby. It is believed that the present invention can be understood more deeply and specifically, as follows:

Please refer to FIG. 2, and as shown in FIG. 4, the present invention is based on the above-mentioned identification system 100, and further provides an identification method of the identification system 100, which includes the following steps:

In the starting step S11 , when the object under test 200 touches the sensing area 11 , the image sensor 12 is activated and generates a dynamic image 21 for the object under test 200 , the dynamic image 21 includes a plurality of interval images 23 , and then the sensing step S12 is executed.

In the sensing step S12 , the object under test 200 completely covers the sensing area 11 from being in contact, so that the image sensor 12 further generates a perspective image 22 , and then the identification step S13 is performed.

In the identification step S13 , if the object 200 has a dynamic image change process and a perspective image 22 , the identification module 13 determines that the dynamic image 21 of the object 200 is a biological image.

It is worth mentioning that, as shown in FIG. 3 , the identification system 100 according to the present invention can use the interval images 23 of different exposure intervals 32 within the time range 31 to conduct dynamic images 21 generated by the object under test 200 within the time range 31 To identify authenticity, for example, a fake fingerprint may not be able to gradually enlarge and clear the dynamic image 21 within the time range 31 to produce a perspective image 22 like the fingerprint of a biological image. Therefore, if the dynamic image 21 is within the time range 31 If the change process is abnormal, the identification system 100 according to the present invention can confirm that the dynamic image 21 is not a living thing, thus further improving the recognition accuracy of the present invention.

It should be further explained that, in some embodiments, the identification module 13 further determines whether the clear value (not shown) of the interval image 23 of the dynamic image 21 in the changing process exceeds a threshold (not shown). Whether the dynamic image 21 is a biological image, the clarity value is calculated through one of the image difference method and the image gradient value, and the threshold can be set by the user, or through various algorithms (for example: average value calculation ) and the sharpness value of the past interval image 23 are calculated and generated. Among them, the image difference value method mainly uses the average value of the entire interval image 23, and obtains the difference value between adjacent pixels by subtracting the image value of each point of the entire interval image 23 from the average value to obtain the absolute value, and finally calculates the average The larger the clear value, the clearer the image; in addition, the gradient value of the image mainly performs vertical and horizontal convolution operations on each image value of the interval image 23 through the discrete differential operator (discrete differentiation operator), and the obtained image The gradient value is used as the sharpness value, and the larger the sharpness value is, the clearer the image is, but the present invention is not limited thereto.

Thus, the identification system 100 according to the present invention uses the image sensor 12 to generate a dynamic image 21 for the object under test 200 within a time range 31, the dynamic image 21 includes a plurality of interval images 23, wherein the time range 31 includes a plurality of exposure intervals 32, and when the time range 31 is later, the interval image 23 is clearer, so that the image sensor 12 further generates a perspective image 22, and according to the change process of the dynamic image 21 (not shown) and the perspective image 22 to confirm Whether the dynamic image 21 is a biological image or not, in this way, it can effectively prevent others from cracking the identification system with fingerprint images, pictures, or arbitrary models, and greatly increase the security and identification capabilities of the identification system 100 .

Please refer to FIGS. 4-8. FIG. 4 is a schematic diagram of an identification system according to a first embodiment of the present invention; FIG. 5 is a schematic diagram of an image sensor according to a first embodiment of the present invention; FIG. 6 is an illustration according to the present invention The step block diagram of the identification method of the first embodiment; FIG. 7 is another step block diagram illustrating the implementation of the identification method of the first embodiment of the present invention; FIG. 8 illustrates the actual execution process of the identification method according to the first embodiment of the present invention The flow chart of the steps. As shown in FIG. 4 , the identification system 100 according to the present invention includes: a sensing area 11 , an image sensor 12 , an identification module 13 , and a processing module 14 .

Specifically, as shown in FIG. 4, the identification system 100 according to the first embodiment of the present invention further includes a processing module 14, which is coupled to the identification module 13, and the processing module 14 is used to execute Algorithm, the algorithm performs subtraction operation on the interval images 23 and generates a complex subtraction signal (not shown in the figure), but the present invention is not limited thereto.

Specifically, as shown in FIG. 5, the image sensor 12 according to the first embodiment of the present invention further includes a photosensor 121 and a complementary metal oxide semiconductor 122, wherein the photosensor 121 is Arranged in an array, the light sensor 121 is used to generate image intensity information (not shown in the figure), and the dynamic image 21 is generated by the image intensity information; the complementary metal oxide semiconductor 122 is coupled to the light sensor, The CMOS 122 is used to control the output of the image intensity information. It should be further explained that the image sensor 12 according to the first embodiment of the present invention can use a color filter array (color filter array, CFA) to convert the image received by the image sensor 12 into red light , green light, and blue light, and generate image intensity information through the corresponding light sensor 121 to obtain a stable quality of the dynamic image 21, but the present invention is not limited thereto.

Please refer to FIG. 6 and FIG. 8 , according to the present invention, based on the identification system 100 of the first embodiment, an identification method for implementing the identification system 100 of the first embodiment is further provided, which includes the following steps:

Start step S11', when the object under test 200 contacts the sensing area 11, the image sensor 12 is activated and generates a dynamic image 21 for the object under test 200, the dynamic image 21 includes a plurality of interval images 23, and then the sensing step S12 is executed '.

In the sensing step S12', the object under test 200 touches and covers the sensing area 11 so that the image sensor 12 further generates a perspective image 22, and then the operation step S13' is executed.

In the operation step S13', the processing module 14 calculates the sharpness value of the interval image 23 within the time range 31 through one of the image difference method and the image gradient value, and then executes the identification step S14'.

In the identification step S14', the identification module 13 further judges whether the dynamic image 21 has a change process according to whether the clear value of the interval image 23 exceeds the threshold value. If the object 200 has a change process and has a perspective image 22, the identification module 13 It is determined that the dynamic image 21 is a biological image, otherwise, the recognition module 13 determines that the dynamic image 21 is not a biological image.

In order to further understand the structural features, technical means and expected effects of the present invention, the actual implementation process of the first embodiment of the present invention will be described. It is believed that the present invention can be understood more deeply and specifically as follows. :

Please refer to Figure 8 and match it with Figures 4 to 6. The actual execution process of the identification system 100 according to the present invention is described as follows: First, the start-up step S11' is executed. When the exposure interval 32 is 1, the object to be tested 200 touches the sensing area 11, so that the image sensor 12 starts and targets the object to be tested. The object 200 generates an interval image 23-1 corresponding to the exposure interval 32 being 1; then, the sensing step S12' is executed, and when the exposure interval 32 is 3 and 4, the object 200 to be tested touches to completely cover the sensing area 11 to generate an interval images 23-3 and 23-4, and the image sensor 12 further generates a perspective image 22; then, the operation step S13' is executed, and the processing module 14 calculates the time range through one of the image difference method and the image gradient value 31, the clear value of the interval image 23; finally, execute the identification step S14', through the identification module 13, further judge whether the dynamic image 21 has a change process according to whether the clear value of the interval image 23 exceeds the threshold value, if the object under test 200 If there is a change process and the perspective image 22, the identification module 13 determines that the dynamic image 21 is a biological image; otherwise, the identification module 13 determines that the dynamic image 21 is not a biological image.

Therefore, it can be seen from the above description that according to the identification system 100 of the first embodiment of the present invention, the identification module 13 can be used to determine whether the dynamic image 21 is a biological image, which is different from the prior art that uses blood vessel and vein identification as an identification method. The identification module 13 judges whether the dynamic image 21 is a biological image according to whether the clear value of the interval image 23 from the starting step S11' to the sensing step S12' of the dynamic image 21 exceeds the threshold, and at the same time cooperates with the perspective image 22 to effectively prevent others from Crack the identification system 100 with fingerprint images, pictures, or arbitrary models. At the same time, the identification results do not need to go through a complicated machine learning mechanism and accumulate a large number of identification features, which greatly improves the feasibility of offline identification, and has wide applicability and high security. .

Please refer to FIG. 7, and as shown in FIG. 8, in this embodiment, the present invention is further based on the identification system 100 of the above-mentioned first embodiment, and provides an identification method of the identification system 100, which includes the following steps:

In the starting step S21 , when the object under test 200 touches the sensing area 11 , the image sensor 12 starts to generate a dynamic image 21 , the dynamic image 21 includes a plurality of interval images 23 , and then the subtraction step S22 is performed.

In the subtraction step S22, the processing module 14 executes an algorithm, which performs a subtraction operation on the interval image 23 and generates a subtraction signal (not shown in the figure), and then performs the identification step S23.

In the identification step S23, the identification module 13 uses the subtraction signal as the basis for whether the identification system is unlocked.

In order to further understand the structural features, technical means and expected effects of the present invention, the actual implementation process of the present invention will be described. It is believed that the present invention can be understood more deeply and specifically, as follows:

Please refer to Figure 8 and match it with Figures 4 to 7. The actual execution process of the identification system 100 according to the first embodiment of the present invention is described as follows: First, the start-up step S11 is executed. When the exposure interval 32 is 1, the object under test 200 contacts the sensing area 11, so that the image sensor 12 is started and Generate the interval images 23-5 to 23-8 corresponding to the exposure interval 32 being 5 to 8; then, perform the subtraction step S22, and the interval image 23-8 is increasingly larger than the previous interval images 23-5 to 23-7 It is clear and performs an algorithm on the interval image 23. The algorithm performs a subtraction operation on the interval images 23 and generates a subtraction signal; finally, the identification module 13 uses the subtraction signal as the basis for unlocking the identification system.

It should be further explained that the above interval image 23 may include a plurality of single-frame images generated in the exposure interval 32, and the subtraction signal generated after the interval images 23 perform subtraction operations may have a plurality of not limited to a single phase The subtraction signal, the present invention can use any one of the subtraction signals as the basis for identifying whether the system 100 is unlocked. In some embodiments, when there are a plurality of subtraction signals, the average value of these subtraction signals can be used as the basis for identifying whether the system 100 is unlocked, but the present invention is not limited thereto.

Therefore, it can be known from the above description that the identification system 100 according to the first embodiment of the present invention further uses the processing module 14 to execute an algorithm, which subtracts the interval images 23 of different exposure intervals 32 from each other, so that The RV value (ridge valley value) of the dynamic image according to the present invention is greatly improved, resulting in a better fingerprint recognition effect. Thereby, the Accepted Error Rate (FAR) and Rejected Error Rate (FRR) of the identification system 100 according to the present invention are greatly improved, achieving the goals of high accuracy and wide applicability.

Hereinafter, with reference to the drawings, the embodiment of the first embodiment of the identification system 100 of the present invention will be described, so that those skilled in the art of the present invention can understand possible changes more clearly. Components denoted by the same reference numerals as above are substantially the same as those described above with reference to FIG. 1 . The same elements, features, and advantages as those of the identification system 100 will not be repeated.

Other examples of the identification system 100 are provided below, so that those skilled in the art to which the present invention pertains can understand possible variations more clearly. Components denoted by the same reference numerals as in the above embodiment are substantially the same as those described above with reference to FIG. 1 and FIG. 2 . The same elements, features, and advantages as those of the identification system 100 will not be repeated.

Please refer to Figures 9-11B, Figure 9 is a schematic diagram of an identification system according to a second embodiment of the present invention; Figure 10 is a block diagram illustrating the steps of an identification method according to a second embodiment of the present invention; Figure 11A is an exemplary illustration A schematic diagram of an interval image according to the second embodiment of the present invention; FIG. 11B is a schematic diagram illustrating an algorithm performed on an interval image according to the second embodiment of the present invention. As shown in FIG. 9 , the recognition system 100 according to the present invention includes: a sensing area 11 , an image sensor 12 , a recognition module 13 , a processing module 14 , and a computing module 15 .

Specifically, the identification system 100 according to the second embodiment of the present invention further includes a computing module 15, and the computing module 15 according to the second embodiment of the present invention performs multiple amplification on the above-mentioned subtraction signal, To ensure that the peaks and valleys of the amplified subtraction signal are clear and within the range of signal processing, the computing module 15 can be either hardware or software with computing functions, but the present invention is not limited thereto.

It should be further explained that, according to the processing module 14 of the second embodiment of the present invention, it executes an algorithm to subtract the interval images 23 of different exposure intervals 32 within the time range 31 from each other. In the range 31, the interval image 23 of the next exposure interval 32 is subtracted from the dynamic image of the previous exposure interval 31 to take the average value. In this way, it is different from the conventional technology that only compares the interval images 23 of different exposure intervals 32 to each other. The addition operation is performed to obtain an average value, or the interval image 23 is performed to obtain an average value after only subtracting the empty background value. The identification system 100 according to the present invention can effectively eliminate the noise in the dynamic image 21, so that the RV value (ridge valley value) of the dynamic image 21 is greatly improved, and a better fingerprint identification effect is produced.

It is worth mentioning that since both the processing module 14 and the computing module 15 are used to process the interval image 23, in some embodiments, the products that can carry out the algorithm of subtracting the interval images 23 with each other usually also have corresponding functions. The function of subtracting signal multiple amplification, so the processing module 14 and the computing module 15 can be combined into the same role, but the present invention is not limited thereto.

Please refer to FIG. 10 , according to the present invention, based on the identification system 100 of the second embodiment, an identification method for implementing the identification system 100 of the second embodiment is further provided, which includes the following steps:

In the starting step S21 ′, when the object under test 200 touches the sensing area 11 , the image sensor 12 starts to generate a dynamic image 21 , the dynamic image 21 includes a plurality of interval images 23 , and then the subtraction step S22 ′ is executed.

In the subtraction step S22', the processing module 14 executes an algorithm, the algorithm performs a subtraction operation on the interval image 23, and generates a subtraction signal, and then executes the signal amplification step S23'.

In the signal amplification step S23', the subtraction signal is amplified by the arithmetic module 15. The peaks and valleys of the amplified subtraction signal are clear and within the signal processing range, and then the averaging step S24' is performed.

In the averaging step S24', the subtraction signal is averaged through the computing module 15, and then the identification step S25' is executed.

In the identification step S25', the identification module 13 takes an average value according to the amplified subtraction signal as the basis for whether the identification system is unlocked.

Specifically, please refer to FIG. 11A and FIG. 11B , together with those shown in FIG. 8 to FIG. 10 . According to the second embodiment of the present invention, the actual execution algorithm process of the recognition system 100 is described as follows: As shown in FIG. 11A, FIG. 11A is an exemplary illustration of interval images 23-0, 23-4 to 23-8 in the exposure interval 32 of 60 The intensity in the case of microseconds, it should be further explained that the interval image 23 - 0 is the image intensity information generated by no object on the image sensor 12 , that is, the background value in the conventional technology. As shown in FIG. 11B, FIG. 11B is an exemplary illustration of the interval image 23 (23-4 to 23-8) after executing the algorithm according to the second embodiment of the present invention, and the interval after deducting the empty background value The images 23-0 are compared with each other. It can be understood that since the fingerprint identification system of the prior art only deducts the background value of nothing (i.e. 23-8 minus 23-0), the subtracted background value and the interval image 23 The gap between them is too large, so that the noise in the interval image 23 cannot be effectively deducted, and the image intensity information does not decrease significantly. On the other hand, using the algorithm of the second embodiment of the present invention, the exposure interval 32 within the time range 31 Interval image 23 subtracts interval image 23 of previous exposure interval 32 (as shown in Fig. 11B, 23-5 subtracts 23-4, 23-7 subtracts 23-6, 23-8 subtracts 23-7), due to effectively eliminate The noise in the interval image 23 causes the image intensity information to decrease significantly, so that the RV value (ridge valley value) of the dynamic image 21 according to the present invention is greatly improved, resulting in a better fingerprint recognition effect. Thereby, the Accepted Error Rate (FAR) and Rejected Error Rate (FRR) of the identification system 100 according to the present invention are greatly improved, achieving the goals of high accuracy and wide applicability.

Please refer to Figures 12-13B, Figure 12 is a block diagram illustrating the steps of the identification method according to the third embodiment of the present invention; Figure 13A is a schematic diagram illustrating the algorithm after the dynamic image is executed according to the third embodiment of the present invention ; FIG. 13B is another schematic diagram illustrating the execution algorithm of the dynamic image according to the third embodiment of the present invention. Please refer to FIG. 12 , according to the present invention, based on the identification system 100 described above, an identification method for implementing the identification system 100 of the third embodiment is further provided, which includes the following steps:

In the starting step S21 ″, when the object under test 200 touches the sensing area 11 , the image sensor 12 is activated and generates a dynamic image 21 , the dynamic image 21 includes a plurality of interval images 23 , and then the selection step S22 ″ is executed.

Step S22 ″ is selected, and the processing module 14 uses an algorithm to use one of the interval images 23 as a background interval image (not shown), and then executes the subtraction step S23 ″.

In the subtraction step S23 ″, the processing module 14 executes an algorithm, and the algorithm further performs a subtraction operation on the background interval image and the interval image 23 to generate a subtraction signal, and then executes the signal amplification step S24 ″.

In the signal amplification step S24'', the arithmetic module 15 amplifies the subtraction signal by a multiple, and the peaks and valleys of the amplified subtraction signal are clear and within the signal processing range, and then the averaging step S25'' is performed.

In the averaging step S25'', the subtraction signal is averaged through the computing module 15, and then the identification step S26'' is executed.

In the identification step S25 ″, the identification module 13 uses the amplified subtraction signal as a basis for identifying whether the system 100 is unlocked.

Specifically, please refer to FIG. 13A and FIG. 13B , together with FIG. 12 . According to the third embodiment of the present invention, the actual algorithm execution process of the recognition system 100 is described as follows: As shown in Figure 13A, Figure 13A is an exemplary illustration of the algorithm according to the third embodiment of the present invention to select different interval images 23 as the background Interval images, when the exposure interval 32 is 60 microseconds, compare the interval images 23 minus the different background interval images. Image 23 (i.e. 23-8 minus 23-0), when one of the better interval images 23 is selected as the background interval image through the selection step S22'', the noise in the interval image 23 can be effectively eliminated, resulting in an image Intensity information yields a noticeable drop. It is worth mentioning that the algorithm according to the present invention can self-learn through a machine learning algorithm or a deep learning algorithm, and automatically select one of the better interval images 23 as the background dynamic image. The algorithm can be but Not limited to K-Means Clustering (K-Means Clustering), Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO). As shown in FIG. 13B , FIG. 13B is an exemplary illustration of performing the zoom-in operation after deducting different background interval images from the interval image 23 according to the third embodiment of the present invention. Because the better interval image 23 is selected through the selection step S22 ″ One of them is used as the background interval image, so that the peaks and valleys of the image signal of the enlarged interval image 23 are clear and within the signal processing range (as shown in Figure 13B (23-8 minus 23-6)*4 and (23-8 minus 23-7)*4), effectively eliminate the noise in the interval image 23, causing the image intensity information to drop significantly, so that the RV value (ridge valley value) after the dynamic image 21 is enlarged and calculated according to the present invention ) is further improved.

Thus, the present invention has the following implementation effects and technical effects:

Firstly, based on the identification system 100 of the present invention, combined with the identification method provided by the present invention, the image sensor 12 can generate a dynamic image 21 and a perspective image 22, through the change process of the dynamic image 21 and the perspective image 22 to Judging whether the dynamic image 21 is a biological image can effectively prevent others from cracking the identification system with fingerprint images, pictures, or arbitrary models, and greatly increase the security and identification capabilities of the identification system.

Second, based on the recognition system 100 of the present invention, combined with the recognition method provided by the present invention, by executing an algorithm, the algorithm subtracts the interval images 23 of different exposure intervals 32 from each other and obtains an average value , so that the RV value (ridge valley value) of the dynamic image according to the present invention is greatly improved, resulting in a better fingerprint recognition effect. Thereby, the Accepted Error Rate (FAR) and Rejected Error Rate (FRR) of the identification system 100 according to the present invention are greatly improved, achieving the goals of high accuracy and wide applicability.

Third, the identification module 13 according to the present invention is different from the prior art in using blood vessel and vein identification as an identification method. The identification module 13 judges whether the dynamic image 21 is Biological images and dual authentication effectively prevent others from cracking the identification system 100 with fingerprint images, pictures, or arbitrary models. At the same time, the identification results do not need to go through complicated machine learning mechanisms and accumulate a large number of identification features, which greatly improves the feasibility of offline identification. Wide applicability and high security.

Fourth, in the third embodiment of the present invention, one of the better interval images 23 is selected as the background interval image through the selection step S23'' to effectively eliminate the noise in the interval image 23, resulting in image intensity information An obvious drop occurs, which further increases the RV value (ridge valley value) of the dynamic image 21 according to the present invention after the enlargement operation.

The above is to illustrate the implementation of the present invention through specific specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention; all other equivalent changes or modifications that do not deviate from the spirit disclosed in the present invention should be included in the scope of the following patents Inside.

100: Identification system 11: Sensing area 12: Image sensor 121: Light sensor 122: Complementary metal oxide semiconductors 13: Identification module 14: Processing module 15: Operation module 21: Dynamic image 22: Perspective image 23, 23-1, 23-2, 23-3, 23-4, 23-5, 23-6: interval images 31: time frame 32: Exposure interval 200: The object to be tested S11: Startup steps S12: Sensing step S13: identification step S21: start step S22: subtraction step S23: identification step S11': start step S12': sensing step S13': operation steps S14': identification step S21': start step S22': subtraction step S23': Signal amplification step S24': average step S25': identification step S21'': start step S22'': selected step S23'': subtraction step S24'': Signal amplification step S25'': average step S26'': identification step

1 is a schematic diagram of an identification system according to the present invention; Fig. 2 is a block diagram illustrating the steps of implementing the identification method of the present invention; Fig. 3 is a flow chart illustrating the steps of the actual execution process of the identification method according to the present invention; 4 is a schematic diagram of an identification system according to a first embodiment of the present invention; 5 is a schematic diagram of an image sensor according to a first embodiment of the present invention; 6 is a block diagram illustrating the steps of the identification method according to the first embodiment of the present invention; FIG. 7 is a block diagram illustrating another step of implementing the identification method of the first embodiment of the present invention; FIG. 8 is a flow chart illustrating the steps of the actual execution process of the identification method according to the first embodiment of the present invention; 9 is a schematic diagram of an identification system according to a second embodiment of the present invention; 10 is a block diagram illustrating steps of an identification method according to a second embodiment of the present invention; FIG. 11A is a schematic diagram illustrating an interval image according to a second embodiment of the present invention; FIG. 11B is a schematic diagram illustrating an algorithm performed on an interval image according to the second embodiment of the present invention; 12 is a block diagram illustrating steps of an identification method according to a third embodiment of the present invention; FIG. 13A is a schematic diagram illustrating an algorithm performed on an interval image according to a third embodiment of the present invention; FIG. 13B is another schematic diagram illustrating an algorithm after the interval image is executed according to the third embodiment of the present invention.

100: Identification system

11: Sensing area

12: Image sensor

13: Identification module

Claims (9)

An identification method, which is applied to an identification system, the identification system includes a sensing area, an image sensor and an identification module, the identification module is coupled to the sensing area and the image sensor, The identification method includes the following steps: an activation step, when an object to be tested touches the sensing area, the image sensor is activated and generates a dynamic image for the object to be tested; a sensing step, the object to be tested is moved from touch to cover the sensing area, so that the image sensor further generates a perspective image; and an identification step, according to the change process of the dynamic image and whether the perspective image is associated with the characteristics of the biological body, the identification module judges Whether the dynamic image is a biological image. The identification method as described in Claim 1, wherein the dynamic image includes a plurality of interval images generated by the object under test within a time range, and the identification module judges the dynamic image according to whether the clarity value of the interval images exceeds a threshold Whether the image has the change process, when one of the sharpness values exceeds the threshold, the identification module determines that the dynamic image has the change process. The identification method according to claim 2, wherein the sharpness value is calculated through one of image difference method and image gradient value. The identification method as described in Claim 1, wherein the image sensor further comprises: a plurality of light sensors arranged in an array, the light sensors are used to generate complex image intensity information, and by The dynamic image is generated from the image intensity information; a plurality of complementary metal-oxide semiconductors (Complementary Metal-Oxide Semiconductor, CMOS), which is coupled to the light sensors, and the complementary metal-oxide semiconductors are used to control output of the image intensity information. The identification method according to claim 1, wherein the image sensor is disposed below the sensing area, and the image sensor has a shutter mechanism for controlling the exposure interval. The identification method as described in Claim 5, wherein the shutter mechanism is a global shutter (Global Shutter, GS), so that the light sensors are simultaneously exposed to generate the image intensity information. An identification method, which is applied to the identification system described in claim 1, the identification method includes the following steps: an activation step, when the object to be tested touches the sensing area, the image sensor is activated and targets the The object to be tested generates a dynamic image, the dynamic image includes complex interval images; a subtraction step, a processing module executes an algorithm to perform subtraction operations on these interval images, and generates complex subtraction signals; a signal In the amplification step, an operation module amplifies the subtraction signals by multiples, and the peaks and valleys of the amplified subtraction signals are clear and within the signal processing range; and an identification step, the identification module is based on the amplified The equal and subtractive signals are used as the basis for whether the identification system is unlocked. The identification method as described in claim 7, wherein, after performing the signal amplification step, the identification method further includes: an averaging step, the calculation module takes an average value of the subtracted signals; wherein, the identification step Further, through the identification module, the average value is used as a basis for whether the identification system is unlocked. The identification method as described in claim 7, wherein, after performing the starting step, the identification method further includes: a selection step, the processing module takes one of the interval images as a background interval image; Wherein, the subtraction step further performs a subtraction operation on the background interval image and the interval images, and generates the subtraction signals.

Images