An artificial intelligence-based carbon emissions trading market management system

Abstract

The present invention provides an artificial intelligence-based carbon emission trading market management system, which includes an identity authentication module, a member management module, a trading order management module, and a fee management module; the identity authentication module is used to obtain a live authentication video and a face image of a preset time length when a management personnel conducts a login verification, and input the live authentication video and the face image into a preset artificial intelligence recognition algorithm to verify the identity of the management personnel; the member management module is used for the management personnel who have passed the identity authentication to manage the members of the carbon emission trading market; the trading order management module is used for the management personnel who have passed the identity authentication to manage the trading orders of the carbon emission trading market; the fee management module is used for the management personnel who have passed the identity authentication to manage the trading fees generated by completing the trading orders. The present invention completes the process of live detection and identity authentication without disturbing the management personnel undergoing identity authentication, improving the user experience.

Description

Technical Field

[0001] The present invention relates to the field of management, and particularly to a carbon emission trading market management system based on artificial intelligence. Background Art

[0002] Carbon emission trading refers to the public trading activity of carbon emission allowances obtained based on carbon emission rights by carbon emission trading entities at designated trading institutions. Therefore, it is necessary to manage the carbon emission trading market. Existing market management systems generally use methods such as account passwords or face recognition for identity verification of trading systems. Since account passwords are easily leaked, the security is not high enough. And the existing face recognition methods generally simply verify whether it is a live body through actions of the mouth, eyes or head. This kind of live body verification method requires users to make additional actions for login verification, which affects the user experience. Summary of the Invention

[0003] In view of the above problems, the purpose of the present invention is to provide a carbon emission trading market management system based on artificial intelligence, including an identity verification module, a member management module, a trading order management module and a fee management module;

[0004] The identity verification module is used to obtain a live body verification video and a face image of a preset time length when a management staff member conducts login verification, and input the live body verification video and the face image into a preset artificial intelligence recognition algorithm to verify the identity of the management staff member;

[0005] The member management module is used for the management staff members who have passed the identity verification to manage the members of the carbon emission trading market;

[0006] The trading order management module is used for the management staff members who have passed the identity verification to manage the trading orders in the carbon emission trading market;

[0007] The fee management module is used for the management staff members who have passed the identity verification to manage the trading fees generated by completing trading orders.

[0008] Preferably, the identity verification module includes a shooting sub-module and an artificial intelligence sub-module;

[0009] The shooting sub-module is used to obtain a live body verification video and a face image of a preset time length when a management staff member conducts login verification;

[0010] The artificial intelligence sub-module is used to input the live body verification video and the face image into a preset artificial intelligence recognition algorithm to verify the identity of the management staff member.

[0011] Preferably, the shooting sub-module includes a video shooting unit and an image shooting unit;

[0012] The video shooting unit is used to obtain a live verification video of a preset time length when the management personnel perform login verification;

[0013] The image shooting unit is used to obtain the facial image of the management personnel when the management personnel perform login verification.

[0014] Preferably, the artificial intelligence sub-module includes a live body verification unit and an identity verification unit;

[0015] The live body verification unit is used to identify the live verification video by using a preset artificial intelligence recognition algorithm, and judge whether the live verification video passes the live body detection;

[0016] The identity verification unit is used to obtain the feature information included in the facial image and judge whether the management personnel pass the identity verification based on the feature information when the live verification video passes the live body detection.

[0017] Preferably, the membership management module includes a membership information management unit and a membership information statistics unit;

[0018] The membership information management unit is used to manage the attribute information and permission information of the members in the carbon emission trading market;

[0019] The membership information statistics unit is used to statistically analyze the information of the members in the carbon emission trading market.

[0020] Preferably, the trading order management module includes an abnormal order monitoring unit and a prompt unit;

[0021] The abnormal order monitoring unit is used to judge whether the trading order meets the preset monitoring criteria;

[0022] The prompt unit is used to send a warning prompt to the management personnel according to the preset prompt method when the trading order meets the preset monitoring criteria.

[0023] Preferably, the management of the transaction fees generated by the completed trading orders includes:

[0024] Statistically analyze the transaction fees according to the query period selected by the management personnel who have passed the identity verification, and display them in a chart.

[0025] The carbon emission trading market management system of the present invention uses an artificial intelligence algorithm to process the live verification video and facial image of the management personnel undergoing identity verification, thereby completing the process of live detection and identity verification without disturbing the management personnel undergoing identity verification, which is beneficial to improving the user experience. In addition, the present invention also sets up various types of management modules to manage the members, trading orders, trading fees, etc. of the carbon emission trading market respectively, which is beneficial to improving the management efficiency of the carbon emission trading market. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The present invention will be further described with reference to the accompanying drawings. However, the embodiments in the drawings do not constitute any limitation to the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to the following drawings without creative efforts.

[0027] Figure 1 , which is an exemplary embodiment diagram of a carbon emission trading market management system based on artificial intelligence of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation to the present invention.

[0029] As Figure 1 shown in an embodiment, the present invention provides a carbon emission trading market management system based on artificial intelligence, including an identity verification module, a member management module, a trading order management module, and a fee management module;

[0030] The identity verification module is used to obtain a live verification video and a facial image of a preset time length when the management personnel perform a login verification, and input the live verification video and the facial image into a preset artificial intelligence recognition algorithm to verify the identity of the management personnel;

[0031] The member management module is used for the management personnel who have passed the identity verification to manage the members of the carbon emission trading market;

[0032] The trading order management module is used for the management personnel who have passed the identity verification to manage the trading orders of the carbon emission trading market;

[0033] The fee management module is used for the management personnel who have passed the identity verification to manage the trading fees generated by completing the trading orders.

[0034] The carbon emission trading market management system of the present invention uses an artificial intelligence algorithm to process the live verification video and facial image of the management personnel undergoing identity verification, thereby completing the process of live detection and identity verification without disturbing the management personnel undergoing identity verification, which is beneficial to improving the user experience. In addition, the present invention also sets up various types of management modules to manage the members, trading orders, trading fees, etc. of the carbon emission trading market, which is beneficial to improving the management efficiency of the carbon emission trading market.

[0035] Preferably, the identity verification module includes a shooting sub-module and an artificial intelligence sub-module;

[0036] The shooting sub-module is used to obtain a live verification video and a facial image of a preset time length when the management personnel perform login verification;

[0037] The artificial intelligence sub-module is used to input the live verification video and the facial image into a preset artificial intelligence recognition algorithm to verify the identity of the management personnel.

[0038] Specifically, the preset time length can be 3S, 5S, or other time lengths, and can be adaptively adjusted according to the network speed of the management personnel undergoing identity verification.

[0039] Preferably, the shooting sub-module includes a video shooting unit and an image shooting unit;

[0040] The video shooting unit is used to obtain a live verification video of a preset time length when the management personnel perform login verification;

[0041] The image shooting unit is used to obtain the facial image of the management personnel when the management personnel perform login verification.

[0042] Specifically, when obtaining the live verification video, it can be prompted on the screen that the video recording is in progress and the remaining recording time.

[0043] Preferably, the artificial intelligence sub-module includes a live verification unit and an identity verification unit;

[0044] The live verification unit is used to identify the live verification video by using a preset artificial intelligence recognition algorithm to determine whether the live verification video passes the live detection;

[0045] The identity verification unit is used to obtain the feature information contained in the facial image when the live verification video passes the live detection, and determine whether the management personnel pass the identity verification based on the feature information.

[0046] Preferably, the live verification video is identified by a preset artificial intelligence recognition algorithm to determine whether the live verification video passes the live detection, including:

[0047] Obtain the image frames of the live verification video;

[0048] Obtain the image features included in the image frames;

[0049] Input the image features into a pre-trained support vector machine classifier for judgment to obtain a live judgment result.

[0050] Preferably, the obtaining of the image frames of the live verification video includes:

[0051] Perform frame splitting on the live verification video to obtain multiple image frames.

[0052] Preferably, the obtaining of the image features included in the image frames includes:

[0053] Perform grayscale processing on the image frames to obtain grayscale images;

[0054] Perform noise reduction processing on the grayscale images to obtain noise-reduced images;

[0055] Perform restoration processing on the noise-reduced images to obtain restored images;

[0056] Use a preset image feature extraction algorithm to obtain the image features included in the restored images.

[0057] Preferably, the preset image feature extraction algorithm includes the HOG algorithm, the SIFT algorithm, etc.

[0058] Preferably, the performing of grayscale processing on the image frames to obtain grayscale images includes:

[0059] Perform grayscale processing on the image frames using the following formula:

[0060] G(x,y) = w 1 ×R(x,y) + w 2 ×G(x,y) + w 3 ×B(x,y)

[0061] where G(x,y) represents the grayscale value of the pixel point with coordinates (x,y), w 1 、w 2 、w 3Denote as the preset weight parameter. R(x, y), G(x, y), and B(x, y) respectively denote the pixel values of the pixel points with coordinates (x, y) in the red component image, green component image, and blue component image. The red component image, green component image, and blue component image are the images corresponding to the red component, green component, and blue component of the image frame in the RGB color space.

[0062] In the above embodiments of the present invention, by using a weighted method to obtain a grayscale image, compared with the method of directly using a single component image as the grayscale image, it can retain the difference information between pixel points as much as possible.

[0063] Preferably, the noise reduction process for the grayscale image to obtain a denoised image includes:

[0064] Perform differential enhancement processing on the grayscale image in the following manner to obtain a differentially enhanced image:

[0065] aG(x, y) = G(x, y) × Φ

[0066] Wherein, G(x, y) and aG(x, y) respectively denote the pixel values of the pixel points with coordinates (x, y) before and after differential enhancement processing, and Φ denotes the preset differential enhancement coefficient;

[0067] Perform wavelet decomposition processing on the differentially enhanced image to obtain a high-frequency wavelet coefficient image and a low-frequency wavelet coefficient image;

[0068] Perform the following processing on the high-frequency wavelet coefficient image to obtain a processed high-frequency wavelet coefficient image:

[0069] If gpximg(x, y) ≤ gpxthr 1 , then calculate the processing result of the high-frequency wavelet coefficient image through the following formula:

[0070]

[0071] If gpxthr 1 <gpximg(x, y)<gpxthr 2 , then calculate the processing result of the high-frequency wavelet coefficient image through the following formula:

[0072] agpximg(x, y) = |gpxthr 1 -gpxthr 2 |×[gpximg(x, y)+tfm[gpximg(x, y)]]

[0073] If gpxthr 2If ≤ gpximg(x,y), then the processing result of the high-frequency wavelet coefficient image is calculated by the following formula:

[0074] agpximg(x,y) = gpximg(x,y) × Γ

[0075] Where, gpximg(x,y) and agpximg(x,y) respectively represent the pixel values of the pixel point with coordinates (x,y) before and after processing; tfm[gpximg(x,y)] represents the adjustment function. If gpximg(x,y) is greater than the preset comparison parameter, the value of tfm[gpximg(x,y)] is 1. If gpximg(x,y) is equal to the preset comparison parameter, the value of tfm[gpximg(x,y)] is 0.6, otherwise the value of tfm[gpximg(x,y)] is 0.3, gpxthr 1 and gpxthr 2 respectively represent the preset first judgment parameter and second judgment parameter, tq represents the preset control parameter, tq belongs to (0, 0.8), and Γ represents the preset proportional parameter;

[0076] Perform wavelet reconstruction processing on the low-frequency wavelet coefficient image and the processed high-frequency wavelet coefficient image to obtain a denoised image.

[0077] In the above embodiment of the present invention, before denoising, the gray values of the pixel points in the gray image are first enhanced by the difference enhancement function, thereby amplifying the difference in pixel values between the noise pixel points and the normal pixel points. Then, after decomposing the difference-enhanced image into a high-frequency wavelet coefficient image and a low-frequency wavelet coefficient image, denoising processing is performed. This processing method is beneficial to increasing the pixel values of the noise pixel points in the high-frequency wavelet coefficient image, making the noise pixel points more prominent. It is beneficial to improving the accuracy of the denoising effect.

[0078] In addition, when denoising the high-frequency wavelet coefficient image, mainly the first judgment parameter and the second judgment parameter are used to automatically select an appropriate processing function for the high-frequency wavelet coefficient image under different conditions for processing, improving the pertinence of the processing function, and thus being beneficial to improving the accuracy of the denoising processing result.

[0079] In addition, compared with the traditional spatial domain denoising processing method, the present invention performs denoising processing in the transform domain, which is beneficial to effectively removing the noise in the image while retaining more image edge information.

[0080] Preferably, the repairing the denoised image to obtain a repaired image includes:

[0081] Obtain the pixel points to be repaired in the denoised image:

[0082] Calculate the change degree coefficient of each pixel point in the denoised image respectively:

[0083] chgidx(x,y) = lowG(x,y) - G(x,y)

[0084] where lowG(x,y) and G(x,y) respectively represent the gray values of the pixel point with coordinates (x,y) in the denoised image and the grayscale image, and chgidx(x,y) represents the change degree coefficient of the pixel point with coordinates (x,y);

[0085] If chgidx(x,y) is less than or equal to the preset change degree threshold, it means that the pixel point with coordinates (x,y) is a pixel point to be repaired;

[0086] Adopt the following method to perform repair processing on each pixel point to be repaired respectively to obtain a repaired image:

[0087]

[0088] where δ ∈ {L,a,b}; reptre(x,y) and areptre(x,y) respectively represent the gray values of the pixel point with coordinates (x,y) before and after the repair processing, reptre[(x,y),δ] = |δ(x,y) - δ(midpix)|, δ(x,y) represents the pixel value of the pixel point corresponding to the pixel point with coordinates (x,y) in the δ component image in the Lab color space corresponding to the image frame, and δ(midpix) represents the pixel value of the pixel point corresponding to the pixel point with the largest pixel value in the denoised image in the δ component image in the Lab color space corresponding to the image frame, represents the weight value corresponding to the δ component.

[0089] In the process of converting the image into a grayscale image, since the difference information contained in the three components is expressed only through the gray value, even if weighted summation is used for grayscale conversion, the difference degree information between pixel points is still lost to a certain extent. For example, the differences between two pixel points in the R, G, and B components were originally relatively large values. However, due to the weighted processing in the grayscale conversion process, in the grayscale image, the difference between the two pixel points is reduced, which also loses the information of the original image edge to a certain extent. Therefore, after the noise reduction processing of the present invention, the pixel points to be repaired are first judged, and then the pixel points to be repaired are repaired to obtain a repaired image. Thus, the difference information between pixel points in the original image frame is transmitted to the repaired image, greatly retaining the edge detail information of the image, which is beneficial to obtaining accurate image features, and further improving the security of the identity verification process of the present invention.

[0090] In the calculation process of the repair process, the present invention puts the pixel points to be repaired and the reference pixel points, that is, the pixel points with the largest pixel value, into the Lab color space to obtain the difference information. This processing method can effectively reduce the influence of the illumination condition on the difference information and is beneficial to improving the accuracy of the obtained difference information.

[0091] Preferably, the support vector machine classifier is obtained by the following method:

[0092] Obtain a set of training images that have been normalized;

[0093] Input each training image into a neural network for image feature extraction to obtain the image features of the training image;

[0094] Use the image features of the training images to train the support vector machine classifier to obtain a pre-trained support vector machine classifier.

[0095] Preferably, the membership management module includes a membership information management unit and a membership information statistics unit;

[0096] The membership information management unit is used to manage the attribute information and permission information of the members in the carbon emission trading market;

[0097] The membership information statistics unit is used to statistically analyze the information of the members in the carbon emission trading market.

[0098] Specifically, the attribute information of the members includes membership number, enterprise name, corporate account, enterprise address, enterprise contact, annual output value method, etc.

[0099] The permission information includes the trading permissions of the members, and different trading permissions correspond to different trading segments. For example, the members are classified according to the annual output value, and different membership levels correspond to trading segments with different transaction scales. For example, only members with an annual output value exceeding 1 billion can participate in trading segments with a daily average trading volume of more than 100 million.

[0100] Statistically analyzing the membership information includes classifying and counting the members according to their attribute information, for example, counting the members according to the provinces where they are located.

[0101] Preferably, the trading order management module includes an abnormal order monitoring unit and a prompt unit;

[0102] The abnormal order monitoring unit is used to judge whether a trading order meets the preset monitoring criteria;

[0103] The prompt unit is used to send a warning prompt to the management personnel according to the preset prompt method when the trading order meets the preset monitoring criteria.

[0104] For example, the preset monitoring standard can be whether the transaction amount is greater than the preset amount threshold within a preset time period. Conducting abnormal orders can further prevent the carbon emission trading market from being maliciously manipulated.

[0105] The ways of warning prompts include pop-up warnings, sound warnings, etc.

[0106] Preferably, the management of the transaction fees generated by the completed transaction orders includes:

[0107] Statistical analysis of the transaction fees according to the query period selected by the authenticated management personnel and display them in the form of charts.

[0108] The query period includes one year, one month, one week, one day, etc., and can also be a custom time interval. The transaction fees can include stamp duty, handling fees, etc.

[0109] Although the embodiments of the present invention have been shown and described, those skilled in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the claims and their equivalents.

Claims

1. An artificial intelligence-based carbon emission trading market management system, characterized in that, it includes an identity verification module, a member management module, a trading order management module, and a fee management module; The identity verification module is used to obtain a live verification video and a facial image of a preset time length when a management staff conducts a login verification, and input the live verification video and the facial image into a preset artificial intelligence recognition algorithm to verify the identity of the management staff; The member management module is used for the management staff who has passed the identity verification to manage the members of the carbon emission trading market; The trading order management module is used for the management staff who has passed the identity verification to manage the trading orders of the carbon emission trading market; The fee management module is used for the management staff who has passed the identity verification to manage the transaction fees generated by completing the trading orders; The identity verification module includes a shooting sub-module and an artificial intelligence sub-module; The shooting sub-module is used to obtain a live verification video and a facial image of a preset time length when a management staff conducts a login verification; The artificial intelligence sub-module is used to input the live verification video and the facial image into a preset artificial intelligence recognition algorithm to verify the identity of the management staff; The artificial intelligence sub-module includes a live verification unit and an identity verification unit; The live verification unit is used to use a preset artificial intelligence recognition algorithm to recognize the live verification video and judge whether the live verification video passes the live detection; The identity verification unit is used to obtain the feature information contained in the facial image when the live verification video passes the live detection, and judge whether the management staff passes the identity verification based on the feature information; The using a preset artificial intelligence recognition algorithm to recognize the live verification video and judge whether the live verification video passes the live detection includes: Obtaining the image frames of the live verification video; Obtaining the image features contained in the image frames; Inputting the image features into a pre-trained support vector machine classifier for judgment to obtain a live judgment result; The obtaining the image features contained in the image frames includes: Performing grayscale processing on the image frames to obtain grayscale images; Performing noise reduction processing on the grayscale images to obtain noise-reduced images; Performing repair processing on the noise-reduced images to obtain repaired images; Using a preset image feature extraction algorithm to obtain the image features contained in the repaired images; The performing noise reduction processing on the grayscale images to obtain noise-reduced images includes: Performing differential enhancement processing on the grayscale images in the following manner to obtain a differentially enhanced image: aG(x,y) = G(x,y) × Φ where G(x,y) and aG(x,y) respectively represent the pixel values of the pixel point with coordinates (x,y) before and after differential enhancement processing, and Φ represents a preset differential enhancement coefficient; Performing wavelet decomposition processing on the differentially enhanced image to obtain a high-frequency wavelet coefficient image and a low-frequency wavelet coefficient image; Performing the following processing on the high-frequency wavelet coefficient image to obtain a processed high-frequency wavelet coefficient image: If gpximg(x,y) ≤ gpxthr 1 , then the processing result of the high-frequency wavelet coefficient image is calculated by the following formula: If gpxthr 1 <gpximg(x,y)<gpxthr 2 , then the processing result of the high-frequency wavelet coefficient image is calculated by the following formula: agpximg(x,y) = |gpxthr 1 -gpxthr 2 | × [gpximg(x,y) + tfm[gpximg(x,y)]] If gpxthr 2 ≤ gpximg(x, y), then the processing result of the high-frequency wavelet coefficient image is calculated by the following formula: agpximg(x,y) = gpximg(x,y) × Γ Among them, gpximg(x, y) and agpximg(x, y) respectively represent the pixel values of the pixel point with coordinates (x, y) before and after processing; tfm[gpximg(x, y)] represents the adjustment function. If gpximg(x, y) is greater than the preset comparison parameter, the value of tfm[gpximg(x, y)] is 1. If gpximg(x, y) is equal to the preset comparison parameter, the value of tfm[gpximg(x, y)] is 0.

6. Otherwise, the value of tfm[gpximg(x, y)] is 0.3, gpxthr 1 and gpxthr 2 respectively represent the preset first judgment parameter and the second judgment parameter, tq represents the preset control parameter, tq belongs to (0, 0.8), and Γ represents the preset proportional parameter; Perform wavelet reconstruction on the low-frequency wavelet coefficient image and the processed high-frequency wavelet coefficient image to obtain a denoised image; Obtain the pixels to be repaired in the denoised image: Calculate the change degree coefficients of each pixel in the denoised image respectively: chgidx(x,y) = lowG(x,y) - G(x,y) where lowG(x,y) and G(x,y) respectively represent the gray values of the pixel at coordinates (x,y) in the denoised image and the gray-scale image, and chgidx(x,y) represents the change degree coefficient of the pixel at coordinates (x,y); If chgidx(x,y) is less than or equal to a preset change degree threshold, it means that the pixel at coordinates (x,y) is a pixel to be repaired; Obtain and perform repair processing on each pixel to be repaired respectively in the following manner to obtain a repaired image: Among them, δ ∈ {L, a, b}; reptre(x, y) and areptre(x, y) respectively represent the gray values of the pixel at coordinates (x, y) before and after the repair process, and reptre[(x, y), δ] = δ(x, y) - δ(midpix) | , where δ(x, y) represents the pixel value of the pixel corresponding to the pixel at coordinates (x, y) in the δ component image in the Lab color space corresponding to the image frame, and δ(midpix) represents the pixel value of the pixel corresponding to the pixel with the largest pixel value in the denoised image in the δ component image in the Lab color space corresponding to the image frame represents the weight value corresponding to the δ component 2. A management system for carbon emission trading market based on artificial intelligence according to claim 1, characterized in that, the shooting sub-module includes a video shooting unit and an image shooting unit; the video shooting unit is used to obtain a live verification video of a preset time length when the management personnel perform login verification; the image shooting unit is used to obtain the face image of the management personnel when the management personnel perform login verification.

3. A management system for carbon emission trading market based on artificial intelligence according to claim 1, characterized in that, the member management module includes a member information management unit and a member information statistics unit; the member information management unit is used to manage the attribute information and permission information of the members in the carbon emission trading market; the member information statistics unit is used to perform statistical analysis on the information of the members in the carbon emission trading market.

4. A management system for carbon emission trading market based on artificial intelligence according to claim 1, characterized in that, the trading order management module includes an abnormal order monitoring unit and a prompt unit; the abnormal order monitoring unit is used to judge whether a trading order meets the preset monitoring criteria; the prompt unit is used to send a warning prompt to the management personnel according to the preset prompt method when the trading order meets the preset monitoring criteria.

5. A management system for carbon emission trading market based on artificial intelligence according to claim 1, characterized in that, the management of the transaction fees generated by the completed trading orders includes: Statistically analyze the transaction fees according to the query period selected by the management personnel who have passed the identity verification and display them in a chart.

Images