TWI774577B - Identity identification system - Google Patents

Abstract

An identity identification system includes a camera module, a capture module, and an identification module. The camera module shoots a person to generate a face image. The capture module acquires human body features from the person. The identification module receives the face image and human body features. The identification module carries out analysis to acquire the gray scale value of a plurality of pixels and each pixel from the facial image, and compares the gray scale value of each pixel with the gray scale values of neighboring nine-square division with the pixel as the center. The comparison result is calculated and converted into an eight-digit binary sequence value and the corresponding decimal system, acquiring the comparison value corresponding to each pixel. The identification module classifies and analyzes the comparison value of each pixel and stores a facial feature diagram. When a new facial image is received, a new facial feature diagram is accordingly generated after analysis. When the new facial feature diagram matches the stored facial feature diagram, and the human body features are also acquired at the same time, an authentication signal is generated.

Description

identification system

The present invention relates to the related field of identification analysis, in particular to an identity identification system.

With the development of artificial intelligence, the use of CNN image recognition in deep learning has made great progress. Therefore, the use of product recognition in artificial intelligence for face recognition has become a very challenging and difficult task. The work of meaning.

At present, the application of face recognition can be roughly divided into the following 6 ways: 1. Access control system: identify the identity of the entrant through the face recognition system; 2. Image surveillance system: monitor the crowd in public places to achieve the purpose of identity recognition 3. Network application: use face recognition technology to prevent non-credit card owners from using it; 4. Student attendance system: the school uses smart cards to cooperate with face recognition to carry out roll call and record students’ attendance records; 5. Camera: a new type of digital camera The built-in face recognition function can assist in focusing when shooting people; 6. Smart phone: used to unlock the phone and identify the user.

However, when face recognition is generally used for identity verification, since the face recognition system uses a single image for recognition without other auxiliary recognition targets, it is not limited to use dynamic images as the recognition target, so photos are used as the recognition target. It is also feasible, but using a photo as an identification object may be regarded by the face recognition system as authenticating its identity, so photos may be used to deceive the face recognition system, resulting in the identity being used fraudulently; therefore, the current face recognition system When the system is used for identity verification, there may be a risk of potential use of hidden identities.

In order to solve the above problem, the present invention provides an identity recognition system, which utilizes facial images to match human body features to increase the complexity of identification and reduce the chance of being fraudulently used.

An embodiment of the present invention provides an identity recognition system, which includes: a camera module for photographing a person and generating a face image, the camera module can transmit the face image to the outside; a capture module A group, which can acquire and generate a human body feature from the person, and the capture module can transmit the human body feature to the outside; an identification module, which is connected with the signal of the camera module and the capture module, and the identification module has mutual coupling A communication unit, a processing unit, a verification unit and a memory unit are connected, the communication unit can receive and transmit the facial image to the processing unit, the communication unit can receive the human body feature and transmit it to the verification unit; the processing unit analyzes the facial image The complex number of pixels and the gray value of each pixel are obtained. The processing unit arranges the gray value of each pixel in a vertical and horizontal grid, and calculates the gray value of each pixel in the neighborhood of the nine grid grid with its center as the center. Size comparison, convert the comparison result operation into an octet binary sequence value, and then perform the corresponding decimal system operation to obtain a comparison value corresponding to each pixel, and the processing unit converts the comparison value generated by each pixel. Perform classification and analysis to generate a facial feature map, and the processing unit stores the facial feature map to the memory unit; wherein, when the processing unit receives a new facial image transmitted by the camera module again, and analyzes it to generate a new facial image face feature map, and the new face feature map can be matched with the face feature map already stored in the memory unit, and the verification unit obtains the human body features at the same time, the verification unit generates an authentication signal.

Through the above, the present invention utilizes facial images to generate high-accuracy face feature maps through image analysis and processing, and uses human features as verification judgments, thereby increasing the complexity of identification and reducing the chance of being fraudulently used. Therefore, it can effectively improve the conventional risk of identifying only a single image as a hidden identity risk.

Furthermore, the present invention processes the facial image through image analysis, and performs gray value comparison, octet binary sequence value and decimal system operation, so that it can be seen that the changed image is compared with the original image. , which can more clearly reflect the texture of each typical area, and at the same time dilute the characteristics of smooth areas that are of little research value, and reduce the dimension of the features, so as to obtain a high-accuracy face feature map and achieve a high recognition effect. .

In order to facilitate the description of the central idea of the present invention expressed in the column of the above-mentioned summary of the invention, specific embodiments are hereby expressed. Various objects in the embodiments are drawn in proportions, dimensions, deformations or displacements suitable for illustration, rather than scales of actual elements.

The directional terms mentioned in the present invention, such as "up", "down", "front", "rear", "left", "right", "inside", "outside", "side", etc., are only drawings. Therefore, the directional terms used are used to describe and understand the present invention, rather than to limit the present invention, and will be described together first.

Please refer to FIG. 1 to FIG. 4 , the present invention provides an identification system 100, which includes:

A camera module 10 has an image unit 11 and a first transmission unit 12 coupled to each other. The image unit 11 is used to capture a person and generate a facial image F, and the first transmission unit 12 transmits the facial image F. outgoing transmission.

A capture module 20 has a feature unit 21 and a second transmission unit 22 coupled to each other. The feature unit 21 can capture and generate a human body feature from a person, and the second transmission unit 22 can transmit the human body feature to the outside. transmission.

The feature unit 21 of the capture module 20 can be a photographing device or a sound-receiving device. When the feature unit 21 is a photographing device, the feature unit 21 can capture body movements, gestures or facial muscle changes of the person to generate static images or dynamic images When the feature unit 21 is a radio device, the feature unit 21 can acquire the voice of the person to generate the human body feature of the voiceprint signal.

An identification module 30 is connected to the camera module 10 and the capture module 20 for signals. The identification module 30 has a communication unit 31 , a processing unit 32 , a verification unit 33 and a memory unit 34 coupled to each other. The communication unit 31 can be signal-connected with the first transmission unit 12 and the second transmission unit 22 , the communication unit 31 can receive and transmit the facial image F to the processing unit 32 , and the communication unit 31 can receive the human body features and transmit them to the verification unit 33 . It is particularly noted that the communication unit 31 can be connected to the first transmission unit 12 and the second transmission unit 22 simultaneously or non-simultaneously, that is, the camera module 10 and the capture module 20 can function simultaneously or non-simultaneously.

Please refer to FIG. 1 , FIG. 3 and FIG. 4 , the processing unit 32 analyzes the complex pixel P and the grayscale value G of each pixel P from the facial image F, and the processing unit 32 analyzes the grayscale value G of each pixel P Arrange in vertical and horizontal grids, and compare the size of the gray value G in the neighborhood of the nine grid grid with the gray value G of each pixel P as the center, wherein, when the gray value G of each pixel P is greater than or equal to When the gray value G of the pixel P in the neighborhood is coded as 1 in the corresponding grid of the pixel P in the neighborhood; when the gray value G of each pixel P is smaller than the gray value of the pixel P in the neighborhood When G, the code is 0 in the corresponding grid of the pixel P in the neighborhood, and the processing unit 32 converts the comparison result operation into an octet binary sequence value. In the embodiment of the present invention, each pixel P and the neighborhood The comparison result of the pixel P is to extract the remaining grids of the non-center grid of the nine-square grid, arrange them clockwise as an octet binary sequence value, and then perform the corresponding decimal system operation to obtain a ratio corresponding to each pixel P. For the values, the processing unit 32 classifies and analyzes the comparison values generated by each pixel P to generate a facial feature map M, and the processing unit 32 stores the facial feature map M in the memory unit 34 .

It should be noted that when there is no corresponding face feature map M stored in the memory unit 34, the facial image F can be obtained through the camera module 10, and the image analysis processing can be performed through the processing unit 32 of the recognition module 30 to Obtain the facial feature map M, and store the facial feature map M in the memory unit 34; when the processing unit 32 receives the new facial image F transmitted by the camera module 10 again, and generates a new facial feature map through analysis M, and the new face feature map M can match the face feature map M that has been stored in the memory unit 34, and the verification unit 33 obtains the human body features at the same time, then the verification unit 33 generates an authentication signal, wherein the authentication signal is As for the trigger signal, the authentication signal can be used as the relevant basis for unlocking, identification, etc. The present invention does not limit the trigger type of the authentication signal.

Furthermore, in another embodiment of the present invention, please refer to FIG. 2 , the identification module 30 can be coupled to a terminal device 40 , and the identification module 30 can transmit and display the authentication signal on the terminal device 40 , wherein, The identification module 30 can be set up on a server and mounted on the terminal device 40 in the form of software or an application (APP), and the terminal device 40 can be various personal computers, notebook computers, smart mobile devices or tablet computers. The device 40 is various personal computers or notebook computers, and the use means of the identification module 30 can be software; when the terminal device 40 is a smart mobile device or a tablet computer, the use means of the identification module 30 can be an application program (APP).

When the means of use of the identification module 30 is software, the camera module 10 and the capture module 20 are separately installed in the use environment; when the means of use of the identification module 30 is an application, the camera module 10 and the capture module 20 are separately installed in the use environment; The capture module 20 obtains the function authorization of the terminal device 40 and utilizes the camera and microphone functions of the terminal device 40 to achieve the effect of capturing the facial image F and acquiring human body features. The first transmission unit 12 and the second transmission unit 22 It utilizes the information transmission function of the terminal device 40 to achieve the effect of transmitting the facial image F and human body features to the recognition module 30 .

In addition, when the verification unit 33 generates an authentication signal, it means that the face feature map M and the human body feature have matched each other, so the verification unit 33 can store the human body feature with the face feature map M in the memory unit 34 to facilitate the next time Identify work.

The identification module 30 further has a establishment unit 35, the establishment unit 35 provides for setting a personal information and sets a basic data, a face feature map M and a human body feature corresponding to the personal information, and the establishment unit 35 can store the set personal information in the memory unit. 34.

In addition, the identity recognition system 100 of the present invention can be applied to a merchant, and the camera module 10 and the capture module 20 can be installed in the merchant, the recognition module 30 can be mounted on the terminal device 40, and the recognition module 30 can be set with the sales time point set by the merchant. The information system signal is connected, and the identification module 30 receives a checkout consumption information sent by the point-of-sale information system; the identification module 30 further has a member unit 36, and the member unit 36 can record the checkout consumption information corresponding to the personal information; Through the membership unit 36 and the integration of big data analysis, merchants can grasp the personal information and consumption behavior of consumers, and further understand the shopping orientation of consumers.

Therefore, the identity recognition system 100 of the present invention can use the facial image F to generate a high-accuracy face feature map M through image analysis and processing, and use the human body features as a verification judgment, thereby increasing the complexity of identification and reducing fraudulent use. opportunity for identity.

Furthermore, the present invention processes the facial image F through image analysis, and performs the size comparison of the gray value G, the octet binary sequence value and the decimal system operation, so that the changed image and the original image can be seen. In contrast, it can more clearly reflect the texture of each typical area, while at the same time diluting the features of smooth areas that are of little research value, and at the same time reducing the dimension of the features, and then obtaining a high-accuracy face feature map M, reaching Highly recognizable effect.

The above-mentioned embodiments are only used to illustrate the present invention, but not to limit the scope of the present invention. All the modifications or changes that do not violate the spirit of the present invention belong to the intended protection category of the present invention.

100: Identification System

10: Camera module

11: Image unit

12: The first transmission unit

20: Capture module

21: Feature unit

22: The second transmission unit

30: Identification module

31: Communication unit

32: Processing unit

33: Verification Unit

34: Memory Unit

35: Build Unit

36: Membership Unit

40: Terminal device

F: face image

P: pixel

G: gray value

M: face feature map

FIG. 1 is a block diagram of a system architecture according to an embodiment of the present invention. FIG. 2 is a block diagram of a system architecture according to another embodiment of the present invention. FIG. 3 is a schematic diagram of analyzing a complex number of pixels from a facial image according to the present invention. FIG. 4 is a schematic diagram of the calculation of the gray value of each pixel according to the present invention.

100: Identification System

10: Camera module

11: Image unit

12: The first transmission unit

20: Capture module

21: Feature unit

22: The second transmission unit

30: Identification module

31: Communication unit

32: Processing unit

33: Verification Unit

34: Memory Unit

35: Build Unit

36: Membership Unit

Claims (8)

An identity recognition system, comprising: a camera module, which has an image unit coupled to each other and a first transmission unit, the image unit is used to photograph a person and generate a face image, the first transmission unit can The facial image is transmitted to the outside; a capture module has a feature unit coupled with each other and a second transmission unit, the feature unit can be obtained from the person and generate a human feature, the second transmission unit The unit can transmit the human body features to the outside; an identification module is connected to the camera module and the capture module for signals, and the identification module has a communication unit, a processing unit, and a verification unit coupled to each other and a memory unit, the communication unit is coupled with the first transmission unit and the second transmission unit, the communication unit can receive and transmit the facial image to the processing unit, the communication unit can receive the human body feature and transmit it to the verification unit; the processing unit analyzes the multiple pixels and the gray value of each pixel from the facial image, the processing unit arranges the gray value of each pixel in a vertical and horizontal grid, and analyzes the gray value of each pixel The size of the gray value is compared in the neighborhood of the nine-square grid with the value as the center, and the operation of the comparison result is converted into an octet binary sequence value, and then the corresponding decimal system operation is performed to obtain a comparison corresponding to each pixel. value, the processing unit classifies and analyzes the comparison value generated by each pixel to generate a facial feature map, and the processing unit stores the facial feature map to the memory unit; wherein, when the processing unit The new facial image transmitted by the camera module is received again, and a new facial feature map is generated through analysis, and the new facial feature map and the facial features that have been stored in the memory unit are If the images can be matched, and the verification unit obtains the human body feature at the same time, the verification unit generates an authentication signal. The identification system according to claim 1, wherein when the gray value of each pixel is greater than or equal to the gray value of the pixel in the neighborhood, a code of 1 is generated in the corresponding grid of the pixel in the neighborhood; When the gray value of each pixel is smaller than the gray value of the pixel in the neighborhood, a code of 0 is generated in the corresponding grid of the pixel in the neighborhood. The identity recognition system according to claim 2, wherein the comparison result of each pixel with the pixels in the neighborhood is to extract the rest of the non-central grids in the nine-square grid, and arrange them clockwise as an octet binary sequence value. The identity recognition system according to claim 1 or 3, wherein the capture module is a photographing device, and the capture module can capture body movements, gestures or facial muscle changes of the person to generate static images or dynamic images The human body feature of the image; the verification unit can store the human body feature in the memory unit. The identity recognition system according to claim 4, wherein the capture module is a radio device, the capture module can capture the voice of the person to generate the human body feature of the voiceprint signal; the verification unit can Human features are stored in the memory unit. The identity recognition system according to claim 5, wherein the recognition module has a creation unit, the creation unit provides for setting a personal information and setting a basic data, the face feature map and the human body feature corresponding to the personal information, the The establishment unit can store the set personal information into the memory unit. The identification system according to claim 6, wherein the identification module is connected with a signal of a point-of-sale information system set by the merchant, and the identification module receives a checkout consumption information sent by the point-of-sale information system; the identification module is further There is a member unit, which can record the checkout consumption information corresponding to the personal information. The identification system according to claim 1 or 3, wherein the identification module can be coupled with a terminal device, and the identification module can transmit and display the authentication signal on the terminal device.

Images