Search

CN-121981790-A - Intelligent intelligent purchasing method and system for smart city commodity big data based on block chain

CN121981790ACN 121981790 ACN121981790 ACN 121981790ACN-121981790-A

Abstract

The invention discloses a block chain-based intelligent purchasing method and system for big data of smart city commodities, and relates to the technical field of voice analysis. The method comprises the steps of collecting voice signals of a user, denoising the voice signals to obtain target voice signals, identifying the target voice signals to generate identification results, confirming user identities, deducting money in a personal account system corresponding to the user according to the confirmation results, recording checkout time, dispatching products, acquiring and recording a receipt image according to commodity receipt confirmation information of the user, optimizing the receipt image to obtain a receipt optimized image, and uploading the user identities in the confirmation results, commodity information in the identification results, checkout time and the receipt optimized image to a block chain. The invention combines a plurality of methods to realize accurate voice recognition, and utilizes the block chain to realize the uplink storage of core information, thereby ensuring the safety and reliability of data.

Inventors

  • GAO XIBO

Assignees

  • 北京吉道尔科技有限公司

Dates

Publication Date
20260505
Application Date
20240301
Priority Date
20230303

Claims (7)

  1. 1. A block chain-based intelligent purchasing method for big data of smart city commodities is characterized by comprising the following steps: When a user needs to purchase goods, collecting voice signals of the user; denoising the voice signal by using a self-adaptive voice denoising method based on frequency domain weighted segmentation signal-to-noise ratio detection so as to obtain a target voice signal; Identifying the target voice signal by using a voice identification method based on a multi-voice enhancement model combined with low consumption gain, and generating an identification result; confirming the user identity according to the user identity information in the identification result and the pre-recorded user basic voice signal by using a small sample voiceprint identification method based on voice coding, and generating a confirmation result; Deducting money in a personal account system of a corresponding user according to the confirmation result, recording the checkout time, generating and sending a delivery order according to commodity information in the identification result and a reserved receiving address of the user, and carrying out commodity delivery; Acquiring and inputting an signing receipt image according to commodity signing confirmation information of a user, and optimizing the signing receipt image by utilizing an image depth enhancement method based on image super-resolution reconstruction so as to obtain a signing optimization image; And uploading the user identity in the confirmation result, commodity information in the identification result, the checkout time and the signing optimization image to the blockchain.
  2. 2. The intelligent purchasing method of smart city commodity big data based on block chain according to claim 1, wherein the method for denoising the voice signal by using the adaptive voice denoising method based on frequency domain weighted segmentation signal-to-noise ratio detection comprises the following steps: performing frequency domain weighted segmentation signal-to-noise ratio detection on the voice signal to obtain a corresponding frequency domain weighted segmentation signal-to-noise ratio value; If the frequency domain weighted segmentation signal-to-noise ratio is larger than a preset signal-to-noise ratio threshold, denoising the voice signal by using the simple voice denoising model, otherwise, denoising the voice signal by using the voice denoising model based on deep learning.
  3. 3. The intelligent purchasing method of smart city commodity big data based on block chain according to claim 1, wherein the method for identifying the target voice signal and generating the identification result by utilizing the voice identification method based on the multi-voice enhancement model combined with low consumption gain comprises the following steps: Sequentially selecting a plurality of voice enhancement models, carrying out iterative enhancement processing on a target voice signal, and calculating a voice quality perception evaluation value of the voice signal after each enhancement processing; Calculating the difference value of the voice quality perception evaluation values of the voice signals after the adjacent two times of enhancement processing, stopping iterative enhancement processing when the difference value is smaller than a preset difference value threshold value, and taking the voice signal after the last time of enhancement processing as a high-quality enhancement voice signal; And recognizing the high-quality enhanced voice signal by utilizing the voice recognition model to generate a recognition result.
  4. 4. The intelligent purchasing method of intelligent city commodity large data based on block chain according to claim 1, wherein the method for identifying the user identity by utilizing the small sample voiceprint recognition method based on voice coding according to the user identity information in the recognition result and the pre-recorded user basic voice signal, and generating the identification result comprises the following steps: extracting a plurality of sections of user basic voice signals of the corresponding user which are pre-recorded according to the user identity information in the identification result; Respectively carrying out voice coding on each section of user basic voice signal, calculating every two, and determining a preferred positive sample according to the similarity between each section of user basic voice signal; selecting and carrying out voice coding on multiple sections of user basic voice signals of other users, calculating every two sections of user basic voice signals of other users, and determining a preferable negative sample according to the similarity between the sections of user basic voice signals of other users; Training the optimal positive sample and the optimal negative sample by using the SVM model to obtain a target voiceprint recognition model of the user corresponding to the recognition result; And carrying out voiceprint recognition on the voice signal by adopting a target voiceprint recognition model, and confirming the identity of the user to generate a confirmation result.
  5. 5. A block chain-based intelligent purchasing system for big data of smart city commodities is characterized by comprising a voice acquisition module, a denoising module, an identification module, an identity confirmation module, a transaction sending module, a signing module and an information uplink module, wherein: The voice acquisition module is used for acquiring voice signals of a user when the user needs to purchase goods; The denoising module is used for denoising the voice signal by using a self-adaptive voice denoising method based on frequency domain weighted segmentation signal-to-noise ratio detection so as to obtain a target voice signal; the recognition module is used for recognizing the target voice signal by utilizing a voice recognition method based on the combination of the multi-voice enhancement model and the low consumption gain, and generating a recognition result; The identity confirmation module is used for confirming the user identity according to the user identity information in the identification result and the pre-recorded user basic voice signal by using a small sample voiceprint identification method based on voice coding, and generating a confirmation result; The transaction dispatching module is used for deducting money in the personal account system of the corresponding user according to the confirmation result, recording the checkout time, generating and sending a transaction dispatching order according to commodity information in the identification result and the reserved receiving address of the user, and dispatching the commodity; The signing module is used for acquiring and inputting signing bill images according to commodity signing confirmation information of a user, and optimizing the signing bill images by utilizing an image depth enhancement method based on image super-resolution reconstruction so as to obtain signing optimized images; and the information uplink module is used for uploading the user identity in the confirmation result, the commodity information in the identification result, the checkout time and the signing optimization image to the blockchain.
  6. 6. An electronic device, comprising: A memory for storing one or more programs; A processor; The method of any of claims 1-4 is implemented when the one or more programs are executed by the processor.
  7. 7. A computer readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the method according to any of claims 1-4.

Description

Intelligent intelligent purchasing method and system for smart city commodity big data based on block chain Technical Field The invention relates to the technical field of voice analysis, in particular to a block chain-based intelligent purchasing method and system for big data of smart city commodities. Background With the progress of society, smart cities offer great convenience to citizens through various intelligent means, and are increasingly accepted by vast citizens. The intelligent shopping is taken as an important component in the intelligent city, and can provide direct support for citizens to efficiently select target commodities. For the intelligent shopping system of a part of smart cities, although an intelligent means has been used, high-quality interactive shopping cannot be realized. Meanwhile, the blockchain is used as an emerging technology in recent years, so that the safety of the whole intelligent shopping system can be improved. Therefore, how to combine the blockchain technology to realize intelligent purchasing of big data of smart city commodities becomes a new technical problem. Disclosure of Invention In order to overcome the problems or at least partially solve the problems, the embodiment of the invention provides a smart city commodity big data intelligent purchasing method and system based on a blockchain, which realize high-quality and accurate voice recognition by utilizing a self-adaptive voice denoising method based on frequency domain weighted segmentation signal-to-noise ratio detection, a voice recognition method based on multi-voice enhancement combined low-consumption gain and a small sample voiceprint recognition method based on voice coding, and realize core information uplink storage by utilizing a blockchain technology, so that data safety and reliability are ensured. Embodiments of the present invention are implemented as follows: In a first aspect, an embodiment of the present invention provides a blockchain-based intelligent purchasing method for smart city commodity big data, including the following steps: When a user needs to purchase goods, collecting voice signals of the user; denoising the voice signal by using a self-adaptive voice denoising method based on frequency domain weighted segmentation signal-to-noise ratio detection so as to obtain a target voice signal; Identifying the target voice signal by using a voice identification method based on a multi-voice enhancement model combined with low consumption gain, and generating an identification result; confirming the user identity according to the user identity information in the identification result and the pre-recorded user basic voice signal by using a small sample voiceprint identification method based on voice coding, and generating a confirmation result; Deducting money in a personal account system of a corresponding user according to the confirmation result, recording the checkout time, generating and sending a delivery order according to commodity information in the identification result and a reserved receiving address of the user, and carrying out commodity delivery; Acquiring and inputting an signing receipt image according to commodity signing confirmation information of a user, and optimizing the signing receipt image by utilizing an image depth enhancement method based on image super-resolution reconstruction so as to obtain a signing optimization image; And uploading the user identity in the confirmation result, commodity information in the identification result, the checkout time and the signing optimization image to the blockchain. In order to solve the problems in the prior art, the self-adaptive voice denoising method based on frequency domain weighted segmentation signal-to-noise ratio detection is utilized to match a proper voice denoising model according to the actual characteristics of voice signals, so that the citizen voice signals are denoised more pertinently. And the voice recognition method based on multi-voice enhancement combined low-consumption gain is utilized, so that the high-quality enhancement is performed on the voice signal to the maximum extent on the basis of reducing the consumption of computing resources, and direct support is provided for the accurate recognition of the voice signal. On the basis, a small sample voiceprint recognition method based on voice coding is utilized, and a voiceprint recognition model with the highest quality is trained by using as few training samples as possible, so that direct support is provided for accurate identification of low-consumption citizens. The invention also utilizes the blockchain technology to store the core information of the commodity purchased by citizens in a uplink way, thereby ensuring the safety and reliability of the data. Based on the first aspect, in some embodiments of the present invention, the method for denoising a speech signal by using the adaptive speech denoising method based on frequency domain w