Intelligent facial object information recognition system

Abstract

The invention relates to an intelligent facial object information recognition system. The system comprises an information application component and a control component, the identification module is used for intelligently identifying the entity area corresponding to the face of the human body by adopting a BP neural network model according to the imaging focal length, the total number of each component pixel point of the image region, and each part of field depth value, each part of horizontal coordinate value and each part of vertical coordinate value corresponding to each component pixel point of the image region; and the numerical value analysis component is used for enabling the current payment operation when the entity area corresponding to the human face subjected to intelligent identification is within a set area numerical value range. Through the system, the BP neural network model of a customized structure can be established, the BP neural network model is adopted to intelligently identify the entity area occupied by the human face based on various visual data screened in a targeted manner, and the intelligently identified entity area occupied by the human face is adopted to complete financial payment. Therefore, the security of financial payment is improved.

Description

Technical Field

[0001] This invention relates to the field of facial recognition, and more specifically, to an intelligent facial object information recognition system. Background Technology

[0002] The primary material basis for studying artificial intelligence (AI), and the machine that enables AI technology platforms, is the computer. The history of AI development is inextricably linked to the history of computer science and technology. Besides computer science, AI also involves information theory, cybernetics, automation, bionics, biology, psychology, mathematical logic, linguistics, medicine, and philosophy, among other disciplines. The main research areas of AI include: knowledge representation, automated reasoning and search methods, machine learning and knowledge acquisition, knowledge processing systems, natural language understanding, computer vision, intelligent robots, and automatic programming.

[0003] However, artificial intelligence is a relatively new technology. Although it has been clearly demonstrated that it can be applied to various sub-fields, the technical solutions for each sub-field are either not yet under research or are not mature enough. For example, in financial payments based on human facial object recognition, it is still limited to a coarse recognition mode based on facial contours. For instance, a payment method, device, and equipment proposed by Alipay (Hangzhou) Information Technology Co., Ltd. (application publication number: CN115035578A) includes: acquiring an image to be processed, wherein the image to be processed is a video image; obtaining a suspected target face image from the image to be processed; determining the target face image based on the face ID of the suspected target face image and the distance between the suspected target face image and the facial recognition payment device; performing liveness detection and recognition on the target face image; and if both the liveness detection and recognition are successful, proceeding with payment directly. It is evident that existing technologies cannot effectively identify the physical area occupied by a human face based on various visual data of the human face, and cannot use the identified physical area of ​​the human face to replace the facial contours for high-precision recognition processing. Summary of the Invention

[0004] To address the technical problems in related fields, this invention provides an intelligent facial object information recognition system, the system comprising:

[0005] The authorization execution mechanism, located within the user's payment terminal, is used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or denial processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or denial access instruction accordingly.

[0006] A scene capture mechanism, connected to the authorization execution mechanism, is used to perform an image capture operation on the scene in front of the user's payment terminal when an authorization access instruction is received, so as to obtain and output the corresponding scene image in front;

[0007] A directional blurring device, installed inside the user payment terminal and connected to the scene capture mechanism, is used to perform directional blurring processing on the received front scene image to obtain and output the corresponding directional blurring image;

[0008] A smoothing filter device, connected to the directional blurring device, is used to perform edge-preserving smoothing filtering on the received directional blurring image to obtain and output the corresponding smoothed filter image;

[0009] A salt-and-pepper filter device, connected to the smoothing filter device, is used to perform salt-and-pepper noise removal processing based on median filtering on the received smoothed filtered image, so as to obtain and output the corresponding salt-and-pepper filtered image.

[0010] An information application component, installed within a user payment terminal and connected to the salt-and-pepper filter, includes a main control chip, a data acquisition unit, and an intelligent analysis unit. The data acquisition unit is connected to the intelligent analysis unit, and the main control chip is connected to both the data acquisition unit and the intelligent analysis unit, respectively, for executing the timing control of the data acquisition unit and the intelligent analysis unit. The information application component is used to capture the image area occupied by a human face in the received salt-and-pepper filtered image, obtain the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image area, analyze the imaging focal length of the scene capture mechanism, and use a BP neural network model to intelligently identify the corresponding entity area of ​​the human face based on the imaging focal length of the scene capture mechanism, the total number of each component pixel of the image area, and the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image area.

[0011] The numerical analysis component is connected to the information application component and the authorized execution mechanism respectively. It is used to enable the current payment operation when the corresponding physical area of ​​the human face identified by the intelligent identification is within the set area value range of the face of the verified user corresponding to the current payment operation. It is also used to close the current payment operation when the corresponding physical area of ​​the human face identified by the intelligent identification is outside the set area value range of the face of the verified user corresponding to the current payment operation.

[0012] The BP neural network model is a BP neural network that completes multiple learning iterations, and the number of learning iterations is inversely correlated with the pixel value redundancy of the salt-and-pepper filtered image.

[0013] The intelligent facial object information recognition system of this invention is logically reliable and operates stably. Because it can establish a customized BP neural network model and use this model to intelligently identify the physical area occupied by the human face based on targeted selection of visual data, and then use the intelligently identified physical area occupied by the human face to complete financial payments, the security of financial payments is improved. Attached Figure Description

[0014] Those skilled in the art can better understand the many advantages of the present invention by referring to the accompanying drawings.

[0015] Figure 1 This is a schematic diagram of the structure of an intelligent facial object information recognition system according to a primary embodiment of the present invention.

[0016] Figure 2 This is a schematic diagram of the structure of an intelligent facial object information recognition system according to a secondary embodiment of the present invention.

[0017] Figure 3 This is a schematic diagram of the structure of an intelligent facial object information recognition system according to a further embodiment of the present invention. Detailed Implementation

[0018] Figure 1 This is a schematic diagram of an intelligent facial object information recognition system according to a primary embodiment of the present invention, the system comprising:

[0019] The authorization execution mechanism, located within the user's payment terminal, is used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or denial processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or denial access instruction accordingly.

[0020] For example, an authorization execution mechanism, installed within a user payment terminal, is used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or rejection processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or rejection access instruction accordingly. This includes: implementing the authorization execution mechanism using a programmable logic device, installed within the user payment terminal, and used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or rejection processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or rejection access instruction accordingly.

[0021] A scene capture mechanism, connected to the authorization execution mechanism, is used to perform an image capture operation on the scene in front of the user's payment terminal when an authorization access instruction is received, so as to obtain and output the corresponding scene image in front;

[0022] A directional blurring device, installed inside the user payment terminal and connected to the scene capture mechanism, is used to perform directional blurring processing on the received front scene image to obtain and output the corresponding directional blurring image;

[0023] A smoothing filter device, connected to the directional blurring device, is used to perform edge-preserving smoothing filtering on the received directional blurring image to obtain and output the corresponding smoothed filter image;

[0024] A salt-and-pepper filter device, connected to the smoothing filter device, is used to perform salt-and-pepper noise removal processing based on median filtering on the received smoothed filtered image, so as to obtain and output the corresponding salt-and-pepper filtered image.

[0025] An information application component, installed within a user payment terminal and connected to the salt-and-pepper filter, includes a main control chip, a data acquisition unit, and an intelligent analysis unit. The data acquisition unit is connected to the intelligent analysis unit, and the main control chip is connected to both the data acquisition unit and the intelligent analysis unit, respectively, for executing the timing control of the data acquisition unit and the intelligent analysis unit. The information application component is used to capture the image area occupied by a human face in the received salt-and-pepper filtered image, obtain the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image area, analyze the imaging focal length of the scene capture mechanism, and use a BP neural network model to intelligently identify the corresponding entity area of ​​the human face based on the imaging focal length of the scene capture mechanism, the total number of each component pixel of the image area, and the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image area.

[0026] The numerical analysis component is connected to the information application component and the authorized execution mechanism respectively. It is used to enable the current payment operation when the corresponding physical area of ​​the human face identified by the intelligent identification is within the set area value range of the face of the verified user corresponding to the current payment operation. It is also used to close the current payment operation when the corresponding physical area of ​​the human face identified by the intelligent identification is outside the set area value range of the face of the verified user corresponding to the current payment operation.

[0027] The BP neural network model is a BP neural network that completes multiple learning iterations, and the number of learning iterations is inversely correlated with the pixel value redundancy of the salt and pepper filtered image.

[0028] The authorization execution mechanism, located within the user's payment terminal, is used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or denial processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or denial access instruction accordingly. This includes issuing an authorization access instruction when performing authorization processing of the facial access authorization request based on the user's manual authorization action.

[0029] Figure 2 This is a schematic diagram of the structure of an intelligent facial object information recognition system according to a secondary embodiment of the present invention.

[0030] and Figure 1 different, Figure 2 The intelligent facial object information recognition system may also include the following components:

[0031] A vibration measurement mechanism is connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively, and is used to measure the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component.

[0032] For example, the vibration measurement mechanism includes multiple vibration measurement units, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, for measuring the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component;

[0033] The vibration measurement mechanism is connected to the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component, respectively, and is used to measure the current vibration amplitude of each of the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component. The vibration measurement mechanism includes multiple vibration measurement units, each connected to the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component, to perform separate measurements of the current vibration amplitude of each of the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component.

[0034] The vibration measurement mechanism includes multiple vibration measurement units, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component to perform separate measurements of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component. The multiple vibration measurement units are multiple vibration sensing circuits, respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component to perform separate measurements of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component.

[0035] The plurality of vibration measurement units are plurality of vibration sensing circuits, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component to complete the separate measurement of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component, including: the plurality of vibration sensing circuits have the same structure;

[0036] Furthermore, the plurality of vibration measurement units are plurality of vibration sensing circuits, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component to complete the separate measurement of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component. The plurality of vibration sensing circuits have the same upper limit value and lower limit value for vibration measurement.

[0037] Figure 3 This is a schematic diagram of the structure of an intelligent facial object information recognition system according to a further embodiment of the present invention.

[0038] and Figure 1 different, Figure 3 The intelligent facial object information recognition system may also include the following components:

[0039] A data caching device is disposed near the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, and is connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively, for providing data caching services to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively.

[0040] The data caching device is disposed near the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, and is connected to each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively. It provides data caching services to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively, including: the data caching device is connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component through different data caching interfaces.

[0041] The data caching device, which is disposed near the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component and is connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component respectively, and is used to provide data caching services for the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component respectively, further includes: the data caching devices using different data caching interfaces have the same data transmission bandwidth.

[0042] In addition, in the intelligent facial object information recognition system, the authorization execution mechanism is set in the user payment terminal. When a human face access authorization request triggered by the current payment operation is received, the mechanism performs authorization processing or rejection processing of the human face access authorization request based on the user's manual authorization action, and issues an authorization access instruction or rejection access instruction accordingly. The mechanism also includes issuing a rejection access instruction when the rejection processing of the human face access authorization request is performed based on the user's manual authorization action.

[0043] Furthermore, the scene capture mechanism is also used to suspend the image capture operation of the scene in front of the user's payment terminal when it receives an access denial instruction.

[0044] Based on the above specific embodiments, the present invention possesses at least the following important innovative aspects:

[0045] First, the image region occupied by the human face in the salt-and-pepper filtered image obtained after targeted optimization processing of the payment scene image is captured. The depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image region are obtained. The imaging focal length of the scene capture mechanism is analyzed. A BP neural network model is used to intelligently identify the corresponding entity area of ​​the human face based on the imaging focal length of the scene capture mechanism, the total number of each component pixel of the image region, and the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image region. This enables intelligent identification of the entity area value of the face of the user verifying the current payment request.

[0046] Secondly: The BP neural network model used for intelligent identification is a BP neural network that has completed multiple learning cycles, and the number of learning cycles is inversely correlated with the pixel value redundancy of the salt-and-pepper filtered image after targeted optimization, thereby completing the customized design of the neural network model.

[0047] Furthermore, a numerical analysis component is introduced to enable the current payment operation when the physical area corresponding to the intelligently identified human face is within the set area value range corresponding to the face of the verified user in the current payment operation. It is also used to close the current payment operation when the physical area corresponding to the intelligently identified human face is outside the set area value range corresponding to the face of the verified user in the current payment operation. This completes the automatic payment control based on the intelligent identification result of the physical area value of the verified user's face, improving the security and reliability of the payment operation.

[0048] While the invention has been fully described by way of example and with reference to the accompanying drawings, it should be understood that various changes and modifications will be apparent to those skilled in the art. Therefore, unless otherwise indicated, such changes and modifications should be considered to be included within the scope of the invention.

Claims

1. An intelligent facial object information recognition system, characterized in that, The system includes: The authorization execution mechanism, located within the user's payment terminal, is used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or denial processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or denial access instruction accordingly. A scene capture mechanism, connected to the authorization execution mechanism, is used to perform an image capture operation on the scene in front of the user's payment terminal when an authorization access instruction is received, so as to obtain and output the corresponding scene image in front; A directional blurring device, installed inside the user payment terminal and connected to the scene capture mechanism, is used to perform directional blurring processing on the received front scene image to obtain and output the corresponding directional blurring image; A smoothing filter device, connected to the directional blurring device, is used to perform edge-preserving smoothing filtering on the received directional blurring image to obtain and output the corresponding smoothed filter image; A salt-and-pepper filter device, connected to the smoothing filter device, is used to perform salt-and-pepper noise removal processing based on median filtering on the received smoothed filtered image, so as to obtain and output the corresponding salt-and-pepper filtered image. An information application component, installed within a user payment terminal and connected to the salt-and-pepper filter, includes a main control chip, a data acquisition unit, and an intelligent analysis unit. The data acquisition unit is connected to the intelligent analysis unit, and the main control chip is connected to both the data acquisition unit and the intelligent analysis unit, respectively, for executing the timing control of the data acquisition unit and the intelligent analysis unit. The information application component is used to capture the image area occupied by a human face in the received salt-and-pepper filtered image, obtain the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image area, analyze the imaging focal length of the scene capture mechanism, and use a BP neural network model to intelligently identify the corresponding entity area of ​​the human face based on the imaging focal length of the scene capture mechanism, the total number of each component pixel of the image area, and the depth values, horizontal coordinate values, and vertical coordinate values ​​corresponding to each component pixel of the image area. The numerical analysis component is connected to the information application component and the authorized execution mechanism respectively. It is used to enable the current payment operation when the corresponding physical area of ​​the human face identified by the intelligent identification is within the set area value range of the face of the verified user corresponding to the current payment operation. It is also used to close the current payment operation when the corresponding physical area of ​​the human face identified by the intelligent identification is outside the set area value range of the face of the verified user corresponding to the current payment operation. The BP neural network model is a BP neural network that completes multiple learning iterations, and the number of learning iterations is inversely correlated with the pixel value redundancy of the salt-and-pepper filtered image.

2. The intelligent facial object information recognition system as described in claim 1, characterized in that: An authorization execution mechanism, located within the user's payment terminal, is used to, upon receiving a facial access authorization request triggered by the current payment operation, perform authorization processing or denial processing of the facial access authorization request based on the user's manual authorization action, and issue an authorization access instruction or denial access instruction accordingly. This includes issuing an authorization access instruction when performing authorization processing of the facial access authorization request based on the user's manual authorization action.

3. The intelligent facial object information recognition system as described in claim 2, characterized in that, The system also includes: A vibration measurement mechanism is connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively, and is used to measure the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component. The vibration measurement mechanism is connected to the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component, respectively, and is used to measure the current vibration amplitude of each of the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component. The vibration measurement mechanism includes multiple vibration measurement units, each connected to the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component, to perform separate measurements of the current vibration amplitude of each of the directional blurring device, the smoothing filter, the salt-and-pepper filter, and the information application component.

4. The intelligent facial object information recognition system as described in claim 3, characterized in that: The vibration measurement mechanism includes multiple vibration measurement units, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component to perform separate measurements of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component. The multiple vibration measurement units are multiple vibration sensing circuits, respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component to perform separate measurements of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component.

5. The intelligent facial object information recognition system as described in claim 4, characterized in that: The plurality of vibration measurement units are plurality of vibration sensing circuits, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component to complete the separate measurement of the current vibration amplitude of each of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component, including: the plurality of vibration sensing circuits have the same structure.

6. The intelligent facial object information recognition system as described in claim 5, characterized in that: The plurality of vibration measurement units are plurality of vibration sensing circuits, which are respectively connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component to complete the separate measurement of the current vibration amplitude of the directional blurring device, the smoothing filter device, the salt-and-pepper filter device and the information application component. The plurality of vibration sensing circuits have the same upper limit value and lower limit value for vibration measurement.

7. The intelligent facial object information recognition system as described in any one of claims 3-6, characterized in that, The system also includes: A data caching device is disposed near the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, and is connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively, for providing data caching services to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively.

8. The intelligent facial object information recognition system as described in claim 7, characterized in that: A data caching device is disposed near and connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively, for providing data caching services to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, respectively. This includes the data caching device connecting to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component through different data caching interfaces.

9. The intelligent facial object information recognition system as described in claim 8, characterized in that: A data caching device, disposed near the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component, and connected to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component respectively, for providing data caching services to the directional blurring device, the smoothing filter device, the salt-and-pepper filter device, and the information application component respectively, further includes: the different data caching interfaces used by the data caching device have the same data transmission bandwidth.

Images