Financial payment verification system

Abstract

This invention relates to a financial payment verification system, comprising a depth analysis unit, a data acquisition unit, an information capture unit, a learning and conversion unit, a content identification component, and an enabling control component. This invention enhances the security of the payment collection process and prevents entry into invalid payment collection processes.

Description

Technical Field

[0001] This invention relates to the field of financial payments, and more particularly to a financial payment verification system. Background Technology

[0002] Financial payment is a crucial part of financial transactions, encompassing the process of transferring funds from one party to another. In modern economic activities, financial payments are typically conducted electronically, such as online bank transfers, mobile payments, and credit card payments. These payment methods greatly simplify traditional cash transaction processes and improve the efficiency and security of fund transactions. Financial payment methods are diverse and constantly evolving with technological advancements. Currently common payment methods include online payments, mobile payments, and e-wallet payments. Online payments involve fund transfers through channels such as online banking; mobile payments are made through mobile applications; and e-wallet payments are completed using specific e-wallet tools. These payment methods all play important roles in daily life and business activities.

[0003] When upgrading financial payments to be intelligent, it is necessary to detect whether there is a payment terminal on site that matches the standard shape and color of the financial payment terminal at the same time as the financial payment is initiated. The code scanning function of the code scanning terminal is only enabled when such a terminal is found; otherwise, the code scanning function of the code scanning terminal is disabled. This improves the security of the payment process and avoids entering an invalid payment process. Obviously, there is no corresponding solution in the existing technology. Summary of the Invention

[0004] To address technical issues in related fields, this invention provides a financial payment verification system. By introducing a depth-of-field analysis mechanism, when a target with a depth-of-field value less than or equal to a preset depth-of-field threshold is present in a real-time captured image, the imaging area of ​​this target in the real-time captured image is used as a reference imaging area for subsequent dual identification based on shape and color. The real-time captured image originates from a miniature imaging mechanism embedded in the front panel of a code scanning terminal used to scan payment codes. When the code scanning terminal scans the payment code displayed on the financial terminal, it initiates a payment process to the corresponding account of the financial terminal. Furthermore, a learning conversion mechanism is introduced to perform multiple learning operations on a backpropagation (BP) neural network to obtain a BP neural network model output after multiple learning operations. The number of learning operations performed by the BP neural network is positively correlated with the resolution of the real-time captured image, thereby completing the structural design of the BP neural network model and using the BP neural network model based on the reference imaging. The system intelligently identifies whether a target with a depth value less than or equal to a preset depth threshold in the real-time captured image matches the standard shape and color of the financial payment terminal in both aspects. This is achieved by considering the number of occupied pixels in the region, the number of edge pixels, the difference between the maximum and minimum depth values ​​for each pixel, the hue component values ​​for each pixel in the reference imaging region, the luminance component values ​​for each pixel in the reference imaging region, and the saturation component values ​​for each pixel in the reference imaging region. Furthermore, an enable control component is introduced to enable the code scanning function of the code scanning terminal when the BP neural network model intelligently identifies that a target with a depth value less than or equal to the preset depth threshold in the real-time captured image matches the standard shape and color of the financial payment terminal in both aspects. Otherwise, the code scanning function of the code scanning terminal is disabled. This improves the security of the payment process and avoids entering an invalid payment process.

[0005] According to the present invention, a financial payment verification system is provided, the system comprising:

[0006] A depth-of-field analysis mechanism is used to output the imaging area of ​​the target with a depth value less than or equal to a preset depth-of-field threshold in the real-time captured image as a reference imaging area when there is a target with a depth value less than or equal to the preset depth-of-field threshold in the real-time captured image. The real-time captured image comes from a miniature imaging mechanism embedded in the front panel of a code scanning terminal used to scan payment codes.

[0007] The data acquisition mechanism is connected to the depth analysis mechanism and is used to acquire the number of occupied pixels, the number of edge pixels, and the difference between the maximum and minimum depth values ​​of each pixel in the reference imaging area.

[0008] An information capture mechanism, connected to the depth analysis mechanism, is used to capture the hue component values, luminance component values, and saturation component values ​​corresponding to each pixel in the reference imaging area.

[0009] The learning conversion mechanism is used to perform multiple learning operations on the BP neural network to obtain a BP neural network after multiple learning operations and output it as a BP neural network model. The number of learning operations completed by the BP neural network is positively correlated with the resolution of the real-time captured image.

[0010] The content identification component, connected to the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to intelligently identify, using a BP neural network model, whether a target with a depth of field value less than or equal to a preset depth of field threshold in the real-time captured image matches the standard shape and color of the financial payment terminal in terms of both shape and color. This identification is based on the number of occupied pixels in the reference imaging area, the number of edge pixels, the difference between the maximum and minimum depth of field values ​​of each pixel, the hue component values ​​of each pixel in the reference imaging area, the luminance component values ​​of each pixel in the reference imaging area, and the saturation component values ​​of each pixel in the reference imaging area.

[0011] The enable control component, connected to the content identification component, is used to enable the code scanning function of the code scanning terminal when the BP neural network model intelligently identifies that the target in the real-time captured image has a depth value less than or equal to a preset depth threshold and matches the standard shape and color of the financial payment terminal in both aspects. Otherwise, the code scanning function of the code scanning terminal is disabled.

[0012] Therefore, it can be seen that the present invention has at least the following four key inventive points:

[0013] Firstly, a depth-of-field analysis mechanism is introduced to detect targets with depth values ​​less than or equal to a preset depth-of-field threshold in the real-time captured image. The imaging area of ​​such targets in the real-time captured image is used as a reference imaging area for the suspected presence of a financial payment terminal for subsequent dual identification of shape and color. The real-time captured image comes from a miniature imaging mechanism embedded in the front panel of a code scanning terminal used to scan payment codes. The code scanning terminal initiates the payment process to the corresponding account of the financial payment terminal when it scans the payment code displayed on the financial payment terminal.

[0014] Secondly, a learning conversion mechanism is introduced to perform multiple learning operations on the BP neural network to obtain a BP neural network after multiple learning operations, which is then used as the output of the BP neural network model. The number of learning operations completed by the BP neural network is positively correlated with the resolution of the real-time captured image, thereby completing the structural design of the BP neural network model.

[0015] Furthermore, a BP neural network model is used to intelligently identify whether targets with depth values ​​less than or equal to a preset depth threshold in the real-time captured image match the standard shape and color of the financial payment terminal in terms of both shape and color. This is based on the number of occupied pixels in the reference imaging area, the number of edge pixels, the difference between the maximum and minimum depth values ​​of each pixel, the tonal component values ​​of each pixel in the reference imaging area, the luminance component values ​​of each pixel in the reference imaging area, and the saturation component values ​​of each pixel in the reference imaging area.

[0016] Finally, an enable control component is introduced to enable the code scanning function of the code scanning terminal when the BP neural network model intelligently identifies a target in the real-time captured image whose depth value is less than or equal to a preset depth threshold and whose shape and color match the standard shape and color of the financial payment terminal. Otherwise, the code scanning function of the code scanning terminal is disabled, thereby improving the security of the payment process and preventing entry into an invalid payment process. Detailed Implementation

[0017] The implementation scheme of the financial payment verification system of the present invention will be described in detail below.

[0018] The financial payment verification system shown in embodiment A of the present invention includes:

[0019] A depth-of-field analysis mechanism is used to output the imaging area of ​​the target with a depth value less than or equal to a preset depth-of-field threshold in the real-time captured image as a reference imaging area when there is a target with a depth value less than or equal to the preset depth-of-field threshold in the real-time captured image. The real-time captured image comes from a miniature imaging mechanism embedded in the front panel of a code scanning terminal used to scan payment codes.

[0020] Specifically, the depth-of-field analysis mechanism is used to output the imaging area of ​​the target with a depth value less than or equal to a preset depth-of-field threshold in the real-time captured image as a reference imaging area when there is a target with a depth value less than or equal to the preset depth-of-field threshold in the real-time captured image. The real-time captured image comes from the micro-capture mechanism embedded in the front panel of the code scanning terminal used to scan the payment code. The micro-capture mechanism includes: the depth-of-field analysis mechanism content data caching component, used to cache the preset depth-of-field threshold.

[0021] The data acquisition mechanism is connected to the depth analysis mechanism and is used to acquire the number of occupied pixels, the number of edge pixels, and the difference between the maximum and minimum depth values ​​of each pixel in the reference imaging area.

[0022] An information capture mechanism, connected to the depth analysis mechanism, is used to capture the hue component values, luminance component values, and saturation component values ​​corresponding to each pixel in the reference imaging area.

[0023] The learning conversion mechanism is used to perform multiple learning operations on the BP neural network to obtain a BP neural network after multiple learning operations and output it as a BP neural network model. The number of learning operations completed by the BP neural network is positively correlated with the resolution of the real-time captured image.

[0024] The content identification component, connected to the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to intelligently identify, using a BP neural network model, whether a target with a depth of field value less than or equal to a preset depth of field threshold in the real-time captured image matches the standard shape and color of the financial payment terminal in terms of both shape and color. This identification is based on the number of occupied pixels in the reference imaging area, the number of edge pixels, the difference between the maximum and minimum depth of field values ​​of each pixel, the hue component values ​​of each pixel in the reference imaging area, the luminance component values ​​of each pixel in the reference imaging area, and the saturation component values ​​of each pixel in the reference imaging area.

[0025] The enable control component, connected to the content identification component, is used to enable the code scanning function of the code scanning terminal when the target with a depth value less than or equal to a preset depth threshold in the real-time captured image is matched with the standard shape and color of the financial payment terminal in both shape and color. Otherwise, the code scanning function of the code scanning terminal is disabled.

[0026] The method employs a BP neural network model to intelligently identify whether targets with depth values ​​less than or equal to a preset depth threshold in the real-time captured image match the standard shape and color of the financial payment terminal in terms of both shape and color. This identification is based on the number of occupied pixels in the reference imaging area, the number of edge pixels, the difference between the maximum and minimum depth values ​​of each pixel, the hue component values ​​of each pixel in the reference imaging area, the luminance component values ​​of each pixel in the reference imaging area, and the saturation component values ​​of each pixel in the reference imaging area.

[0027] The process of performing multiple learning operations on the BP neural network to obtain a BP neural network after multiple learning operations is used as the output of the BP neural network model. The positive correlation between the number of learning operations performed by the BP neural network and the resolution of the real-time captured image includes: using a signal conversion function to represent the signal conversion relationship between the number of learning operations performed by the BP neural network and the resolution of the real-time captured image.

[0028] Compared to embodiment A of the present invention, embodiment B of the present invention illustrates a financial payment verification system that may further include:

[0029] A data parsing component is connected to the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism, respectively, and is used to measure the noise decibels near each of the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism.

[0030] The data parsing component is connected to the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism, respectively, and is used to measure the nearby noise decibels of each of the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism. The data parsing component includes multiple decibel measurement units, which are connected to the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism respectively, to perform separate measurements of the nearby noise decibels of each of the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism.

[0031] The data parsing component includes multiple decibel measurement units, which are respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism to perform separate measurements of the nearby noise decibels of each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism. The multiple decibel measurement units are multiple decibel sensing circuits, respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism to perform separate measurements of the nearby noise decibels of each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism.

[0032] The plurality of decibel measurement units are plurality of decibel sensing circuits, which are respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism to complete the separate measurement of the nearby noise decibels of the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism, including: the plurality of decibel sensing circuits have the same structure;

[0033] The plurality of decibel measurement units are plurality of decibel sensing circuits, which are respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism to complete the separate measurement of the nearby noise decibels of the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism. The plurality of decibel sensing circuits have the same upper limit value and lower limit value for decibel measurement.

[0034] Compared to embodiment A of the present invention, the financial payment verification system shown in embodiment C of the present invention may further include:

[0035] An on-site service component is located near the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism, and is used to provide the voice control services required by each of the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism respectively.

[0036] The on-site service component, located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to provide the voice control services required by each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism. The on-site service component includes a sound acquisition unit, a numerical conversion unit, and a signal transmission unit.

[0037] The on-site service component, located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, and used to provide the voice control services required by the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism respectively, further includes: the numerical conversion unit is connected to the sound acquisition unit and the signal transmission unit respectively;

[0038] The on-site service component, located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to provide the required voice control services to the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism respectively. It further includes: the sound acquisition unit is used to collect sound signals from the vicinity of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, and to determine whether each collected sound signal belongs to a sound control signal, so that only the sound control signal is sent to the value conversion unit for control signal parsing.

[0039] In addition, in the financial payment verification system, the BP neural network is subjected to multiple learning operations to obtain a BP neural network after multiple learning operations, which is then used as the output of the BP neural network model. The positive correlation between the number of learning operations completed by the BP neural network and the resolution of the real-time captured image further includes: in the signal conversion function, the resolution of the real-time captured image is the input data, and the number of learning operations completed by the BP neural network is the output data.

[0040] The financial payment verification system of this invention addresses the technical problem that the security and intelligence level of existing financial payment models still have room for improvement. It can intelligently identify whether a target with a depth of field value less than or equal to a preset depth of field threshold in a real-time captured image matches the financial payment terminal in terms of both shape and color. The system only enables the code scanning function of the code scanning terminal when a match is found, thereby improving the security of the payment process and preventing invalid payment processes from being entered, thus solving the aforementioned technical problems.

[0041] Although the invention has been described and illustrated in detail with reference to exemplary embodiments, those skilled in the art will understand that various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the claims.

Claims

1. A financial payment verification system, characterized in that, The system includes: A depth-of-field analysis mechanism is used to output the imaging area of ​​the target with a depth value less than or equal to a preset depth-of-field threshold in the real-time captured image as a reference imaging area when there is a target with a depth value less than or equal to the preset depth-of-field threshold in the real-time captured image. The real-time captured image comes from a miniature imaging mechanism embedded in the front panel of a code scanning terminal used to scan payment codes. The data acquisition mechanism is connected to the depth analysis mechanism and is used to acquire the number of occupied pixels, the number of edge pixels, and the difference between the maximum and minimum depth values ​​of each pixel in the reference imaging area. An information capture mechanism, connected to the depth analysis mechanism, is used to capture the hue component values, luminance component values, and saturation component values ​​corresponding to each pixel in the reference imaging area. The learning conversion mechanism is used to perform multiple learning operations on the BP neural network to obtain a BP neural network after multiple learning operations and output it as a BP neural network model. The number of learning operations completed by the BP neural network is positively correlated with the resolution of the real-time captured image. The content identification component, connected to the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to intelligently identify, using a BP neural network model, whether a target with a depth of field value less than or equal to a preset depth of field threshold in the real-time captured image matches the standard shape and color of the financial payment terminal in terms of both shape and color. This identification is based on the number of occupied pixels in the reference imaging area, the number of edge pixels, the difference between the maximum and minimum depth of field values ​​of each pixel, the hue component values ​​of each pixel in the reference imaging area, the luminance component values ​​of each pixel in the reference imaging area, and the saturation component values ​​of each pixel in the reference imaging area. The enable control component, connected to the content identification component, is used to enable the code scanning function of the code scanning terminal when the BP neural network model intelligently identifies that the target in the real-time captured image has a depth value less than or equal to a preset depth threshold and matches the standard shape and color of the financial payment terminal in both aspects. Otherwise, the code scanning function of the code scanning terminal is disabled.

2. The financial payment verification system as described in claim 1, characterized in that: The BP neural network model intelligently identifies whether targets with depth values ​​less than or equal to a preset depth threshold in the real-time captured image match the standard shape and color of the financial payment terminal in terms of both shape and color. This identification is based on the number of occupied pixels in the reference imaging area, the number of edge pixels, the difference between the maximum and minimum depth values ​​of each pixel, the hue component values ​​of each pixel in the reference imaging area, the luminance component values ​​of each pixel in the reference imaging area, and the saturation component values ​​of each pixel in the reference imaging area. The process of performing multiple learning operations on the BP neural network to obtain a BP neural network after multiple learning operations is used as the output of the BP neural network model. The positive correlation between the number of learning operations performed by the BP neural network and the resolution of the real-time captured image includes: using a signal conversion function to represent the signal conversion relationship between the number of learning operations performed by the BP neural network and the resolution of the real-time captured image.

3. The financial payment verification system as described in claim 2, characterized in that, The system also includes: A data parsing component is connected to the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism, respectively, and is used to measure the noise decibels near each of the content identification component, the learning and conversion mechanism, the data acquisition mechanism, and the information capture mechanism. The data parsing component, connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to measure the nearby noise decibels of each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism. Specifically, the data parsing component includes multiple decibel measurement units, each connected to one of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, to measure the nearby noise decibels of each of these components.

4. The financial payment verification system as described in claim 3, characterized in that: The data parsing component includes multiple decibel measurement units, which are respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism to perform separate measurements of the nearby noise decibels of each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism. The multiple decibel measurement units are multiple decibel sensing circuits, respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism to perform separate measurements of the nearby noise decibels of each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism.

5. The financial payment verification system as described in claim 4, characterized in that: The plurality of decibel measurement units are plurality of decibel sensing circuits, which are respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism to complete the separate measurement of the nearby noise decibels of the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism, including: the plurality of decibel sensing circuits have the same structure.

6. The financial payment verification system as described in claim 5, characterized in that: The plurality of decibel measurement units are plurality of decibel sensing circuits, which are respectively connected to the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism to complete the separate measurement of the nearby noise decibels of the content identification component, the learning conversion mechanism, the data acquisition mechanism and the information capture mechanism. The plurality of decibel sensing circuits have the same upper limit value and lower limit value for decibel measurement.

7. The financial payment verification system as described in claim 3, characterized in that, The system also includes: The on-site service component is located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, and is used to provide the voice control services required by each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism.

8. The financial payment verification system as described in claim 7, characterized in that: The on-site service component, located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to provide the voice control services required by each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism. The on-site service component includes a sound acquisition unit, a value conversion unit, and a signal transmission unit.

9. The financial payment verification system as described in claim 8, characterized in that: The on-site service component, located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, and used to provide the voice control services required by the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism respectively, further includes: the numerical conversion unit is connected to the sound acquisition unit and the signal transmission unit respectively; The on-site service component, located near the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, is used to provide the necessary voice control services to each of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism. It further includes: the sound acquisition unit collecting sound signals from the vicinity of the content identification component, the learning conversion mechanism, the data acquisition mechanism, and the information capture mechanism, and determining whether each collected sound signal is a sound control signal, so that only sound control signals are sent to the value conversion unit for control signal parsing.