CN114418922A - Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server - Google Patents

Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server Download PDF

Info

Publication number
CN114418922A
CN114418922A CN202011093265.2A CN202011093265A CN114418922A CN 114418922 A CN114418922 A CN 114418922A CN 202011093265 A CN202011093265 A CN 202011093265A CN 114418922 A CN114418922 A CN 114418922A
Authority
CN
China
Prior art keywords
tea cake
image
tea
consumption end
identification code
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011093265.2A
Other languages
Chinese (zh)
Inventor
高辉德
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Pu'er Jingmai Jiuze Investment Development Management Co ltd
Original Assignee
Pu'er Jingmai Jiuze Investment Development Management Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Pu'er Jingmai Jiuze Investment Development Management Co ltd filed Critical Pu'er Jingmai Jiuze Investment Development Management Co ltd
Priority to CN202011093265.2A priority Critical patent/CN114418922A/en
Publication of CN114418922A publication Critical patent/CN114418922A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
    • G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
    • G06K7/10544—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation by scanning of the records by radiation in the optical part of the electromagnetic spectrum
    • G06K7/10821—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation by scanning of the records by radiation in the optical part of the electromagnetic spectrum further details of bar or optical code scanning devices
    • G06K7/10861—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation by scanning of the records by radiation in the optical part of the electromagnetic spectrum further details of bar or optical code scanning devices sensing of data fields affixed to objects or articles, e.g. coded labels
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/018—Certifying business or products
    • G06Q30/0185—Product, service or business identity fraud
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Forestry; Mining
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30128—Food products

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Strategic Management (AREA)
  • Data Mining & Analysis (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Marketing (AREA)
  • Molecular Biology (AREA)
  • Electromagnetism (AREA)
  • Biomedical Technology (AREA)
  • General Business, Economics & Management (AREA)
  • Computational Linguistics (AREA)
  • Economics (AREA)
  • Biophysics (AREA)
  • Agronomy & Crop Science (AREA)
  • Primary Health Care (AREA)
  • Human Resources & Organizations (AREA)
  • Tourism & Hospitality (AREA)
  • Mining & Mineral Resources (AREA)
  • Marine Sciences & Fisheries (AREA)
  • Animal Husbandry (AREA)
  • Toxicology (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)

Abstract

The invention provides a tea cake anti-counterfeiting method, a tea cake anti-counterfeiting system and a cloud server, wherein the tea cake anti-counterfeiting method comprises the following steps: receiving the surface image of the tea cake before warehousing; distributing a unique identification code to the surface image of the tea cake before being put into the warehouse; storing the surface image and the unique identification code of the tea cake; the method comprises the steps that a consumption end uploads an obtained tea cake surface image through scanning a unique identification code; comparing the surface image of the tea cake before warehousing with the surface image of the tea cake uploaded by the consumption end, and judging whether the tea cake obtained by the consumption end is a genuine product; and sending the judgment result to the consumer. By storing the surface image of the tea cake before warehousing and the corresponding unique identification code, after the consumption end uploads the obtained surface image of the tea cake by scanning the unique identification code, the surface image of the tea cake before warehousing is compared with the surface image of the tea cake uploaded by the consumption end, whether the surface image of the tea cake obtained by the consumption end is a genuine product or not is judged, and the judgment result is sent to the consumption end, so that the consumption end can conveniently know whether the obtained tea cake is the genuine product or not.

Description

Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server
Technical Field
The invention relates to the technical field of product anti-counterfeiting, in particular to a tea cake anti-counterfeiting method, a tea cake anti-counterfeiting system and a cloud server.
Background
In the internet era, some lawbreakers forge products, modify information, lie and mutually repudiate by all means for pursuing benefits, so that many consumers cannot know the true aspects of the products, and benefit disputes occur between the consumers and merchants or between the merchants and the merchants. Therefore, product traceability becomes a hot spot in this interconnected era, and is more focused. In the field of tea, rich tea tasting experience is required for tasting the quality of tea, and people who just come into tea ceremony can easily buy secondary-grade fake and fake products without self-knowledge.
Disclosure of Invention
The invention provides a tea cake anti-counterfeiting method, a tea cake anti-counterfeiting system and a cloud server, which are used for realizing anti-counterfeiting tracing of tea cakes and facilitating a consumer to identify whether the tea cakes are genuine or not.
In a first aspect, the invention provides an anti-counterfeiting method for a tea cake, which comprises the following steps: receiving the surface image of the tea cake before warehousing; distributing a unique identification code to the surface image of the tea cake before being put into the warehouse; storing the surface image and the unique identification code of the tea cake; the method comprises the steps that a consumption end uploads an obtained tea cake surface image through scanning a unique identification code; comparing the surface image of the tea cake before warehousing with the surface image of the tea cake uploaded by the consumption end, and judging whether the tea cake obtained by the consumption end is a genuine product; and sending the judgment result to the consumer.
In the scheme, the unique identification code corresponding to the tea cake surface image before warehousing is stored, the unique identification code is scanned at the consumption end to upload the obtained tea cake surface image, the tea cake surface image before warehousing is compared with the tea cake surface image uploaded at the consumption end to judge whether the tea cake surface image obtained at the consumption end is a genuine product, and the judgment result is sent to the consumption end, so that the consumption end can conveniently know whether the obtained tea cake is the genuine product.
In one particular embodiment, the tea cake has opposing obverse and reverse sides: the image of the surface of the tea cake before being received and put in storage comprises the image of the front side and the image of the back side of the tea cake before being received and put in storage, so that the anti-counterfeiting effect is improved.
In one specific embodiment, receiving the front image and the back image of the tea cake before warehousing comprises: receiving an image containing front texture and color information of the tea cake before warehousing; segmenting an image containing front texture and color information of the tea cake; picking out the front image of the tea cake from the image containing the front texture and color information of the tea cake; receiving an image containing the back texture and color information of the tea cake before warehousing; segmenting an image containing the back texture and color information of the tea cake; and (4) separating the back side image of the tea cake from the image containing the back side texture and color information of the tea cake. The image containing the texture and the color information is extracted from the received image, so that the comparison accuracy is improved, and the size of the stored image is reduced.
In one specific embodiment, the step of storing the image of the surface of the tea cake and the unique identification code comprises the following steps: extracting feature vector information of a front image and a back image of the tea cake before warehousing by adopting a neural network; binding the feature vector information of the front image and the back image with the unique identification code; and storing the feature vector information and the unique identification code of the front image and the back image. By storing the characteristic vector information, comparison is facilitated, the later comparison speed is improved, the data size is reduced, and the space occupied by storage is reduced.
In a specific embodiment, the extracting, by using a neural network, feature vector information of the front image and the back image of the tea cake before warehousing comprises: extracting feature vector information of a front image of the tea cake before warehousing by adopting a neural network; extracting feature vector information of a reverse image of the tea cake before warehousing by adopting a neural network; comparing the characteristic vector information of the front image with the characteristic vector information of the back image, and judging whether the characteristic vector information of the front image and the characteristic vector information of the back image are from the same surface image of the tea cake; if so, prompting to upload the front image and the back image of the tea cake again. The tea cakes in the two images are ensured to come from the front and the back of the tea cakes respectively, and the abnormality that the two uploaded images come from the same surface of the tea cakes is eliminated.
In a specific embodiment, the receiving and consuming end uploads the obtained tea cake surface image by scanning the unique identification code, and the receiving and consuming end comprises the following steps: receiving information of a unique identification code scanned by a consumer; prompting the consumption end to upload the obtained front image and back image of the tea cake; and the receiving consumption end uploads the obtained front and back images of the tea cake.
In a specific embodiment, after the receiving and consuming terminal uploads the front image and the back image of the obtained tea cake, the receiving and consuming terminal uploads the surface image of the obtained tea cake by scanning the unique identification code further comprises: judging whether the front image and the back image of the tea cake obtained by uploading the tea cake from the consumption end can be identified or not; and if the image is not identifiable, prompting the consumer to upload the obtained front image and back image again. So as to ensure that the tea cake images uploaded by the consumption end have identifiability.
In a specific embodiment, comparing the image of the surface of the tea cake before warehousing with the image of the surface of the tea cake uploaded by the consumption end, and judging whether the tea cake obtained by the consumption end is a genuine product comprises:
extracting feature vector information of the front image and the back image of the tea cake uploaded by a consumption end by adopting a neural network;
comparing the characteristic vector information of the front image of the tea cake before warehousing with the characteristic vector information of the front image of the tea cake uploaded by the consumption end, and acquiring the similarity between the front image of the tea cake before warehousing and the front image of the tea cake uploaded by the consumption end;
comparing the characteristic vector information of the reverse image of the tea cake before warehousing with the characteristic vector information of the reverse image of the tea cake uploaded by the consumption end, and acquiring the similarity between the reverse image of the tea cake before warehousing and the reverse image of the tea cake uploaded by the consumption end;
judging whether the similarity between the front image of the tea cake before warehousing and the front image of the tea cake uploaded by the consumption end is greater than a set threshold value or not; if not, judging that the tea cake obtained by the consumption end is an authentic product;
if so, judging whether the similarity between the back image of the tea cake before warehousing and the back image of the tea cake uploaded by the consumption end is greater than a set threshold value; if not, judging that the tea cake obtained by the consumption end is an authentic product; if so, judging that the tea cake obtained by the consumption end is a genuine product.
In a specific embodiment, the extracting, by using a neural network, the feature vector information of the front image and the back image of the tea cake before warehousing and the feature vector information of the front image and the back image of the tea cake uploaded by the consuming end specifically include:
building a neural network model, wherein the neural network model comprises 52 convolution layers, 51 batch normalization layers and 1 full connection layer; convolution kernels of the convolutional layers have convolution kernels of sizes 7 × 7, 3 × 3 and 1 × 1, and have different convolution kernel size settings at different stages of the neural network; the convolution step is 1 or 2, and different step settings are provided at different stages of the neural network; a batch normalization layer is arranged behind each convolution layer, an activation function is arranged behind each batch normalization layer, and the activation functions are PReLU activation functions; the fully connected layer is positioned at the last layer of the whole neural network and is used for mapping the specific dimension output of the convolutional layer to a feature vector of a specified dimension;
on a data set of a set number of training samples, selecting a cross entropy loss function to train the built neural network model; setting the total training iteration times of the neural network as set times; selecting an Adam optimizer, wherein the initial learning rate is 0.01, and the weight attenuation coefficient is 0.0005; when training iteration times are every 30k, the learning rate is attenuated to 0.1 time;
compressing and quantizing the trained neural network model to obtain a target neural network model;
inputting the front image and the back image of the tea cake before warehousing into a neural network model of a target, and extracting the characteristic vector information of the front image and the back image of the tea cake before warehousing;
and inputting the front image and the back image of the tea cake uploaded by the consumption end into a neural network model of a target, and extracting the feature vector information of the front image and the back image of the tea cake uploaded by the consumption end.
In a second aspect, the invention also provides an anti-counterfeiting system for tea cakes, which comprises: the receiving module is used for receiving the surface image of the tea cake before warehousing; the identification code distribution module is used for distributing a unique identification code to the surface image of the tea cake before the tea cake is stored in the warehouse; the storage module is used for storing the surface image and the unique identification code of the tea cake; the receiving module is also used for receiving the tea cake surface image obtained by the consumer end through scanning the unique identification code and uploading. The tea cake anti-counterfeiting system further comprises: the comparison module is used for comparing the surface image of the tea cake before warehousing with the surface image of the tea cake uploaded by the consumption end and judging whether the tea cake obtained by the consumption end is a genuine product; and the sending module is used for sending the judgment result to the consumption end.
In the scheme, the storage module is used for storing the tea cake surface image before warehousing and the unique identification code corresponding to the tea cake surface image, the comparison module is used for comparing the tea cake surface image before warehousing with the tea cake surface image uploaded by the consumption end after the consumption end scans the unique identification code to judge whether the tea cake surface image obtained by the consumption end is a genuine product, and then the sending module is used for sending the judgment result to the consumption end, so that the consumption end can conveniently know whether the obtained tea cake is the genuine product.
In a third aspect, the present invention further provides a cloud server, where the cloud server includes a processor and a memory; wherein the memory has stored therein a computer program; the processor is used for executing the tea cake anti-counterfeiting method of any one of the above items by calling the computer program stored in the memory. So that the consumption end can conveniently identify whether the obtained tea cake is a genuine product or not.
Drawings
FIG. 1 is a flow chart of an anti-counterfeiting method for tea cakes according to an embodiment of the invention;
FIG. 2 is a flow chart of an anti-counterfeiting method for tea cakes according to an embodiment of the invention;
fig. 3 is a block diagram of a flow of uploading a surface image of a tea cake to a warehousing end and storing the image by a cloud server according to an embodiment of the present invention;
fig. 4 is a block diagram of another flow of uploading a surface image of a tea cake at a warehousing end and storing the image by a cloud server according to the embodiment of the present invention;
fig. 5 is a block diagram of a process of uploading a surface image of a tea cake obtained by a consumer and comparing the image with the surface image, according to an embodiment of the present invention;
fig. 6 is a block diagram of another process for uploading and comparing the obtained surface image of the tea cake by the consumer according to the embodiment of the present invention;
fig. 7 is a schematic diagram of comparing a surface image of a tea cake by using a neural network according to an embodiment of the present invention.
Reference numerals:
10-warehousing end 20-cloud server 30-consuming end
41-tea cake 42-unique identification code
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to facilitate understanding of the tea cake anti-counterfeiting method provided by the embodiment of the invention, an application scenario of the tea cake anti-counterfeiting method provided by the embodiment of the invention is described below, and the tea cake anti-counterfeiting method is applied to the production, packaging and selling processes of tea cakes. The anti-counterfeiting method of the tea cake is described in detail with reference to the accompanying drawings.
Referring to fig. 1, the tea cake anti-counterfeiting method provided by the embodiment of the invention comprises the following steps:
s10: receiving the surface image of the tea cake before warehousing;
s20: distributing a unique identification code to the surface image of the tea cake before being put into the warehouse;
s30: storing the surface image and the unique identification code of the tea cake;
s40: the method comprises the steps that a consumption end uploads an obtained tea cake surface image through scanning a unique identification code;
s50: comparing the surface image of the tea cake before warehousing with the surface image of the tea cake uploaded by the consumption end, and judging whether the tea cake obtained by the consumption end is a genuine product;
s60: and sending the judgment result to the consumer.
In the scheme, the unique identification code corresponding to the tea cake surface image before warehousing is stored, the unique identification code is scanned at the consumption end to upload the obtained tea cake surface image, the tea cake surface image before warehousing is compared with the tea cake surface image uploaded at the consumption end to judge whether the tea cake surface image obtained at the consumption end is a genuine product, and the judgment result is sent to the consumption end, so that the consumption end can conveniently know whether the obtained tea cake is the genuine product. And because different tea cakes have different pressures when forming the tea cakes, different color changes of different positions of tea leaves and different textures on the surfaces of the tea cakes, the surface image of each tea cake is bound with the unique identification code to form a one-to-one corresponding relation, and the later-stage consumption end compares the obtained surface image of the tea cake with the surface image of the tea cake before warehousing, which corresponds to the unique identification code, by scanning the unique identification code, so that the tracing and anti-counterfeiting are carried out. The texture information and the color information of the surfaces of different tea cakes are utilized to perform tracing and anti-counterfeiting, so that additional anti-counterfeiting marks do not need to be arranged on the package or the tea cakes in a sticking mode. Each of the above steps will be described in detail with reference to the accompanying drawings.
First, referring to fig. 1 and 2, an image of the surface of the tea cake 41 before being put in storage is received. The tea cake 41 is formed by pressing tea leaves into a cake-like structure of a circle, a rectangle, etc., and the tea cake 41 has opposite obverse and reverse sides. The texture information and the color information on the surface of different tea cakes 41 are different because the pressure of different tea cakes 41 is different when the tea cakes 41 are formed and the color of different positions of tea leaves is different. Referring to fig. 3, at the warehousing end 10, the producer of the tea cake 41 collects images of the surface of the tea cake 41. Specifically, a manual unpacking mode can be adopted to collect the surface image of the packaged tea cake 41. In the specific collection, the image of the front side of the tea cake 41 can be collected, and the image of the back side of the tea cake 41 can also be collected. When the front image and the back image of the tea cake 41 are uploaded to the cloud server 20 by the manufacturer, the cloud server 20 receives the front image and the back image of the tea cake 41 before warehousing, so that the front image and the back image of the tea cake 41 can be compared at a later stage, or the back image and the front image and the back image can be compared, and the anti-counterfeiting effect is improved. In implementation, a receiving module may be arranged in the cloud server 20, and the receiving module receives the image of the surface of the tea cake 41 before entering the warehouse, which is uploaded by the warehousing end 10. The collected front image and back image of the tea cake 41 can be photographed by a camera, and the photographed images are color images which can display not only the texture information of the surface of the tea cake 41 but also the color information of the surface of the tea cake 41.
Referring to fig. 4, after the cloud server 20 receives the image of the surface of the tea cake 41 before being put in storage, the cloud server 20 may directly store the received image of the surface of the tea cake 41 and the unique identification code 42, or may process the received image of the surface of the tea cake 41 to remove background information in the image, so as to store only an image portion including texture information and color information of the surface of the tea cake 41. For example, after receiving the image containing the front texture and color information of the tea cake 41 before warehousing, the image containing the front texture and color information of the tea cake 41 may be segmented; then, the front image of the tea cake 41 is extracted from the image containing the front texture and color information of the tea cake 41. Similarly, after receiving the image containing the back texture and color information of the tea cake 41 before warehousing, segmenting the image containing the back texture and color information of the tea cake 41; the reverse image of the tea cake 41 is divided from the image containing the reverse texture and color information of the tea cake 41. Since the size of the image taken by the camera is larger than the image of the surface of the tea cake 41, the image of the tea cake 41 taken by the camera includes not only the texture information and color information of the tea cake 41 but also other background information. The image that will include texture and color information is scratched out from the image of receiving to the adoption mode of cutting apart and scratching to get rid of background information to the later stage is compared, improves the accuracy of comparing, and reduces the size that needs the save data, saves storage space.
Specifically, when the front image and the back image of the tea cake 41 are extracted from the image uploaded by the warehousing end 10, referring to fig. 4, a segmentation module may be arranged on the cloud server 20 for segmenting the image uploaded by the warehousing end 10; a cutout module is also arranged for cutout of the segmented image, so that the front image and the back image of the tea cake 41 are cutout from the image uploaded by the warehousing end 10.
Next, referring to fig. 1 and 2, a unique identification code 42 is assigned to the surface image of the tea cake 41 before being put into the warehouse. Referring to fig. 3, the cloud server 20 assigns a unique identification code 42 to the image of the surface of the tea cake 41 uploaded by the warehousing end 10. The specific implementation manner may be that an identification code distribution module is arranged in the cloud server 20, after the surface image of one tea cake 41 is uploaded by the warehousing end 10, a unique identification code 42 is issued to the warehousing end 10, and the uploaded surface image of the tea cake 41 is bound with the unique identification code 42 one by one. When the later consumption end 30 scans the unique identification code 42, the cloud server 20 can receive the information of the unique identification code 42 scanned by the consumption end 30, and according to the unique identification code 42 scanned by the consumption end 30, the image of the surface of the tea cake 41 before warehousing corresponding to the unique identification code 42 is called from the cloud server 20 for comparison. The unique identification code 42 may be a one-dimensional identification code, a two-dimensional identification code, or the like having an identification function. After the warehousing end 10 collects and finishes uploading the surface image of the tea cake 41, the warehousing end 10 may repack the tea cake 41 and paste the unique identification code 42 corresponding to the tea cake 41 on the package of the tea cake 41, so that the later-stage consumption end 30 uploads the obtained surface image of the tea cake 41 by scanning the unique identification code 42 on the package.
Next, the image of the surface of the tea cake 41 and the unique identification code 42 are saved. The stored information may include image information of the surface of the tea cake 41 and information of the unique identification code 42 corresponding to the tea cake 41. In a specific storage manner, a storage module may be arranged in the cloud server 20, and is used to store the image of the surface of the tea cake 41 before being put in storage and the unique identification code 42 uploaded by the put-in terminal 10.
When the surface image of the tea cake 41 and the unique identification code 42 are stored, the surface image of the tea cake 41 uploaded by the warehousing terminal 10 can be directly stored. The front image of the tea cake 41 and the back image of the tea cake 41, which are extracted and only contain the texture information and the color information of the front and the back of the tea cake 41, but not contain background information, can also be saved. The method can also be realized by storing the characteristic vector information of the image on the surface of the tea cake 41 and the unique identification code 42. For example, referring to fig. 4, after the front image of the tea cake 41 and the back image of the tea cake 41 are extracted, a neural network can be used to extract feature vector information of the front image and the back image of the tea cake 41 before warehousing; then, binding the feature vector information of the front image and the back image with the unique identification code 42; then, the feature vector information and the unique identification code 42 of the front image and the back image are stored. By storing the characteristic vector information, comparison is facilitated, the later comparison speed is improved, the data size is reduced, and the space occupied by storage is reduced. In a specific implementation manner, a retrieval module may be arranged in the cloud server 20, and is configured to extract feature vector information of the front image and the back image of the tea cake 41 from the front image and the back image of the tea cake 41 after matting.
When the characteristic vector information of the front image and the back image of the tea cake 41 before warehousing is extracted by the neural network, the following method can be adopted:
a neural network model is firstly established, and the neural network model comprises 52 convolutional layers, 51 batch normalization layers and 1 full connection layer. Convolution kernels of the convolutional layers have convolution kernels of sizes 7 × 7, 3 × 3, 1 × 1, with different convolution kernel size settings at different stages of the neural network. The convolution step is 1 or 2, and different step settings are provided at different stages of the neural network; each convolutional layer is followed by a batch normalization layer to stabilize the training convergence process. There is one activation function after each batch normalization layer, and the activation functions are PReLU activation functions. The fully-connected layer is located at the last layer of the whole neural network and is used for mapping the specific dimension output of the convolutional layer to the feature vector of the specified dimension.
And then, training the built neural network model. Specifically, on a data set of a set number of training samples, a cross entropy loss function is selected to train the built neural network model, wherein the set number of training samples are related to the training samples in the training set. And setting the total training iteration number of the neural network as a set number, wherein the set number can be 100k and the like. Selecting an Adam optimizer, wherein the initial learning rate is 0.01, and the weight attenuation coefficient is 0.0005; the learning rate decays 0.1 times earlier every 30k training iterations.
And then, compressing and quantizing the trained neural network model to obtain the target neural network model. The trained neural network model is compressed and quantized, so that the size of the neural network model, forward reasoning calculation amount, time consumption and the like are reduced, and the real-time processing capacity of the tea cake verification process is ensured.
Then, the front image and the back image of the tea cake 41 uploaded by the warehousing end 10 are input into the neural network model of the target, and the feature vector information of the front image and the back image of the tea cake 41 before warehousing is extracted.
By extracting the feature vector information of the front image and the back image of the tea cake 41 in the above manner, the matching degree of the extracted feature vector information and the texture information and the color information in the image of the tea cake 41 is improved.
With continued reference to fig. 4, when the characteristic vector information of the front image and the back image of the tea cake 41 before warehousing is extracted by using the neural network, the characteristic vector information of the front image of the tea cake 41 and the characteristic vector information of the back image of the tea cake 41 may be extracted in a time-sharing manner or at the same time, and then the extracted characteristic vector information of the front image of the tea cake 41 and the extracted characteristic vector information of the back image of the tea cake 41 are compared to eliminate an error that the front image of the tea cake 41 and the back image of the tea cake 41 uploaded by the warehousing end 10 are from the same image of the tea cake 41.
For example, a neural network may be used to extract feature vector information of the front image of the tea cake 41 before being put in storage; then, extracting the characteristic vector information of the reverse image of the tea cake 41 before warehousing by adopting a neural network; then, comparing the feature vector information of the front image with the feature vector information of the back image, and judging whether the feature vector information of the front image and the feature vector information of the back image are from the same surface image of the tea cake 41; if yes, the prompt is to upload the front image and the back image of the tea cake 41 again. By comparing whether the feature vector information of the collected front image is consistent with the feature vector information of the back image or not, if not, the front image and the back image of the tea cake 41 are prompted to be uploaded again, so that the tea cakes 41 in the two images are ensured to come from the front and the back of the tea cake 41 respectively, and the abnormality that the two uploaded images come from the same surface of the tea cake 41 is eliminated. In a specific implementation manner, a positive and negative verification module may be arranged on the cloud server 20, and is configured to compare feature vector information of the positive image and the negative image of the tea cake 41 extracted by the retrieval module, and feed back a verification result to the warehousing end 10. The specific implementation manner can be implemented by a sending module of the cloud server 20, that is, the sending module feeds back the verification result to the warehousing end 10. When the image of the front side of the tea cake 41 and the image of the back side of the tea cake 41 are from the same surface, the verification result is that the images do not pass, the sending module sends an indication that the images need to be uploaded again to the warehousing end 10, and prompts that the image of the front side of the tea cake 41 and the image of the back side of the tea cake 41 which are just uploaded come from the same surface of the tea cake 41. If the front image of the tea cake 41 and the back image of the tea cake 41 do not come from the same surface of the tea cake 41, the verification result is passed, the subsequent storage process is entered, and the feature vector information and the unique identification code 42 of the front image and the back image of the tea cake 41 are stored in the storage module.
It should be understood that the way of saving the image of the surface of the tea cake 41 is not limited to the above-described implementation of keeping the feature vector information, and other ways may also be adopted. For example, the front image and the back image of the tea cake 41 may be directly stored, and then, in the comparison, feature vector information may be extracted from the front image and the back image stored before the storage, and the feature vector information of the surface image of the tea cake 41 obtained by uploading the extracted feature vector information from the consumer 30 may be compared.
Next, referring to fig. 1, 2 and 5, the receiving and consuming terminal 30 uploads the image of the surface of the tea cake 41 obtained by scanning the unique identification code 42. After the consumer 30 obtains the tea cake 41 by purchasing, giving away, etc., the unique identification code 42 on the package can be scanned to upload the obtained image of the surface of the tea cake 41, so as to verify whether the obtained tea cake 41 is genuine. Referring to fig. 5, the way of implementing the receiving and consuming terminal 30 uploading the obtained image of the surface of the tea cake 41 by scanning the unique identification code 42 may be: for example, when the consumption end 30 scans the unique identification code 42 through a terminal such as a mobile phone or a notebook computer, the terminal of the consumption end 30 such as the mobile phone or the notebook computer sends the information of the unique identification code 42 to the cloud server 20. Then, prompting the consumption end 30 to upload the obtained front image and back image of the tea cake 41, and the cloud server 20 sends a prompt message to the terminals of the consumption end 30, such as a mobile phone, a notebook computer and the like, to prompt the consumption end 30 to upload the obtained front image and back image of the tea cake 41; then, the consumption end 30 uploads the obtained front and back images of the tea cake 41, and after the consumption end 30 uploads the obtained front and back images of the tea cake 41 through a camera module on a terminal such as a mobile phone or a notebook computer, the cloud server 20 receives the front and back images of the tea cake 41 uploaded by the consumption end 30 for later comparison. Specifically, the way that the cloud server 20 receives the surface image of the tea cake 41 uploaded by the consumption end 30 may be that a receiving module in the cloud server 20 receives the surface image of the tea cake 41 uploaded by the consumption end 30.
In addition, referring to fig. 6, after the receiving and consuming terminal 30 uploads the obtained front image and back image of the tea cake 41, it can also be determined whether the front image and back image of the tea cake 41 uploaded by the consuming terminal 30 can be identified; if not, the consumer 30 is prompted to upload the obtained front and back images again. If the determination result is recognizable, the front and back images of the tea cake 41 uploaded by the consumption end 30 are clear and can be used as images for comparison with the surface images of the tea cake 41 before warehousing, so that the identifiability of the images of the tea cake 41 uploaded by the consumption end 30 is ensured, and the accuracy of later comparison is improved. In a specific implementation manner, the segmentation module in the cloud server 20 may be used to segment the surface image of the tea cake 41 uploaded by the consumption end 30, and then the quality evaluation module in the cloud server 20 is used to evaluate the segmented surface image of the tea cake 41, so as to determine whether the surface image of the tea cake 41 uploaded by the consumption end 30 is recognizable. When the quality evaluation module judges that the surface image of the tea cake 41 divided by the division module is not recognizable, the cloud server 20 sends an indication prompting that the consumption end 30 needs to shoot again and upload the image, and the steps are sequentially circulated until the surface image of the tea cake 41 uploaded by the consumption end 30 is recognizable.
Next, referring to fig. 1, fig. 2, fig. 5, and fig. 6, the surface image of the tea cake 41 before warehousing is compared with the surface image of the tea cake 41 uploaded by the consuming terminal 30, and whether the tea cake 41 obtained by the consuming terminal 30 is a genuine product is determined. The specific implementation manner can realize comparison between the surface image of the tea cake 41 before entering the warehouse and the surface image of the tea cake 41 uploaded by the consumption terminal 30 through a comparison module arranged in the cloud server 20. When the consumption end 30 scans the unique identification code 42 through a device such as a mobile phone end, and the cloud server 20 receives the unique identification code 42 scanned by the device such as the mobile phone end of the consumption end 30, the cloud server calls the feature vector information of the surface image of the tea cake 41 before being put in storage or the surface image of the tea cake 41 before being put in storage, which is stored in the storage module and bound with the unique identification code 42, according to the scanned unique identification code 42. And then acquiring the feature vector information of the front image and the back image of the tea cake uploaded by the consumption end, and extracting the feature vector information of the front image and the back image of the tea cake uploaded by the consumption end by adopting a neural network. Specifically, the front image and the back image of the tea cake 41 uploaded by the consuming terminal 30 may be input into the neural network model of the target, and the feature vector information of the front image and the back image of the tea cake 41 uploaded by the consuming terminal 30 is extracted, so as to improve the matching degree between the extracted feature vector information and the texture information and the color information in the image of the tea cake 41. Then, the cloud server 20 compares the image of the surface of the tea cake 41 or the feature vector information of the image of the surface of the tea cake 41 stored in the storage module with the image of the surface of the tea cake 41 uploaded by the consumer 30.
Referring to fig. 7, specifically, comparing the image of the surface of the tea cake 41 before warehousing with the image of the surface of the tea cake 41 uploaded by the consuming terminal 30, determining whether the tea cake 41 obtained by the consuming terminal 30 is a genuine product may adopt a method of comparing the feature vector information of the image of the front side of the tea cake 41 before warehousing with the feature vector information of the image of the front side of the tea cake 41 uploaded by the consuming terminal 30, and comparing the feature vector information of the image of the back side of the tea cake 41 before warehousing with the feature vector information of the image of the back side of the tea cake 41 uploaded by the consuming terminal 30, that is, adopting a respective comparison method. When the feature vector information of the front image and the back image of the tea cake 41 before warehousing is stored in the storage module of the cloud server 20, referring to fig. 6, after the receiving module of the cloud server 20 receives the surface image of the tea cake 41 uploaded by the consuming terminal 30, the feature vector information of the front image and the back image of the tea cake 41 uploaded by the consuming terminal 30 needs to be extracted through the segmentation module, the tea cake 41 matting module and the retrieval module of the cloud server 20, and then the feature vector information of the surface image of the tea cake 41 before warehousing is compared with the feature vector information of the surface image of the tea cake 41 uploaded by the consuming terminal 30. Specifically, the following method may be adopted:
comparing the feature vector information of the front image of the tea cake 41 before warehousing with the feature vector information of the front image of the tea cake 41 uploaded by the consumption end 30, and acquiring the similarity between the front image of the tea cake 41 before warehousing and the front image of the tea cake 41 uploaded by the consumption end 30;
comparing the feature vector information of the reverse image of the tea cake 41 before warehousing with the feature vector information of the reverse image of the tea cake 41 uploaded by consumption, and acquiring the similarity between the reverse image of the tea cake 41 before warehousing and the reverse image of the tea cake 41 uploaded by the consumption end 30;
judging whether the similarity between the front image of the tea cake 41 before warehousing and the front image of the tea cake 41 uploaded by the consumption end 30 is greater than a set threshold value or not; if not, the judgment result is that the tea cake 41 obtained by the consumption end 30 is an authentic product;
if so, judging whether the similarity between the back image of the tea cake 41 before warehousing and the back image of the tea cake 41 uploaded by the consumption end 30 is greater than a set threshold value; if not, the judgment result is that the tea cake 41 obtained by the consumption end 30 is an authentic product; if yes, the result is that the tea cake 41 obtained by the consuming terminal 30 is genuine.
Namely, whether the similarity between the front image of the tea cake 41 before warehousing and the front image of the tea cake 41 uploaded by the consumption end 30 is greater than a set threshold is judged, and if the similarity is not greater than the set threshold, the fact that the front image of the tea cake 41 obtained by the consumption end 30 is greatly different from the front image of the tea cake 41 before warehousing is explained, so that the conclusion that the tea cake 41 obtained by the consumption end 30 is a non-genuine product can be explained, a subsequent judgment process is not needed, and the judgment time is saved. Only when the similarity between the front image of the tea cake 41 before warehousing and the front image of the tea cake 41 uploaded by the consumption end 30 is greater than a set threshold, further judging whether the similarity between the back image of the tea cake 41 before warehousing and the back image of the tea cake 41 uploaded by the consumption end 30 is greater than the set threshold, so as to judge whether the tea cake 41 obtained by the consumption end 30 is a genuine product.
It should be understood that, when the front image of the tea cake 41 and the back image of the tea cake 41 before being put in storage are stored in the cloud server 20, the cloud server 20 needs to process the called front image of the tea cake 41 and the called back image of the tea cake 41, and the front image of the tea cake 41 and the back image of the tea cake 41 uploaded by the consuming terminal 30 through the segmentation module, the tea cake 41 matting module, and the retrieval module to obtain feature vector information of corresponding images, and then perform comparison.
Next, the judgment result is transmitted to the consumer 30. The specific implementation manner may be implemented by a sending module disposed in the cloud server 20.
By storing the surface image of the tea cake 41 before warehousing and the unique identification code 42 corresponding to the surface image of the tea cake 41, after the consumption terminal 30 uploads the obtained surface image of the tea cake 41 by scanning the unique identification code 42, the surface image of the tea cake 41 before warehousing is compared with the surface image of the tea cake 41 uploaded by the consumption terminal 30 to judge whether the surface image of the tea cake 41 obtained by the consumption terminal 30 is a genuine product, and the judgment result is sent to the consumption terminal 30, so that the consumption terminal 30 can conveniently know whether the obtained tea cake 41 is a genuine product. And because different tea cakes have different pressures when forming the tea cakes, different color changes of different positions of tea leaves and different textures on the surfaces of the tea cakes, the surface image of each tea cake is bound with the unique identification code to form a one-to-one corresponding relation, and the later-stage consumption end compares the obtained surface image of the tea cake with the surface image of the tea cake before warehousing, which corresponds to the unique identification code, by scanning the unique identification code, so that the tracing and anti-counterfeiting are carried out. The texture information and the color information of the surfaces of different tea cakes are utilized to perform tracing and anti-counterfeiting, so that additional anti-counterfeiting marks do not need to be arranged on the package or the tea cakes in a sticking mode.
In addition, an embodiment of the present invention further provides an anti-counterfeiting system for a tea cake, and referring to fig. 2, the anti-counterfeiting system for a tea cake includes: the receiving module is used for receiving the surface image of the tea cake 41 before warehousing; the identification code distribution module is used for distributing a unique identification code 42 to the surface image of the tea cake 41 before warehousing; the storage module is used for storing the surface image of the tea cake 41 and the unique identification code 42; the receiving module is also used for receiving the surface image of the tea cake 41 obtained by uploading the unique identification code 42 by the consuming terminal 30 through scanning. The tea cake 41 anti-counterfeiting system further comprises: the comparison module is used for comparing the surface image of the tea cake 41 before warehousing with the surface image of the tea cake 41 uploaded by the consumption end 30 and judging whether the tea cake 41 obtained by the consumption end 30 is a genuine product; and a sending module, configured to send the determination result to the consuming end 30.
In the above scheme, the storage module stores the surface image of the tea cake 41 before warehousing and the unique identification code 42 corresponding to the surface image of the tea cake 41, after the consumption terminal 30 uploads the obtained surface image of the tea cake 41 by scanning the unique identification code 42, the comparison module compares the surface image of the tea cake 41 before warehousing with the surface image of the tea cake 41 uploaded by the consumption terminal 30 to determine whether the surface image of the tea cake 41 obtained by the consumption terminal 30 is a genuine product, and then the transmission module transmits the determination result to the consumption terminal 30, so that the consumption terminal 30 can conveniently know whether the obtained tea cake 41 is a genuine product. And because the pressure of different tea cakes 41 is different when the tea cakes 41 are formed, the color change of different tea positions is different, and the texture of the surface of the tea cake 41 is different, the surface image of each tea cake 41 is bound with the unique identification code 42 to form a one-to-one corresponding relationship, and the later consumption end 30 scans the unique identification code 42 to compare the obtained surface image of the tea cake 41 with the surface image of the tea cake 41 before warehousing corresponding to the unique identification code 42, thereby tracing the source and preventing the counterfeiting. The texture information and the color information of the surfaces of different tea cakes 41 are utilized to perform tracing and anti-counterfeiting, so that additional anti-counterfeiting marks do not need to be arranged on the package or the tea cakes 41 in an attaching mode.
Each of the above modules is a functional module formed by a computer program in the cloud server 20, the program of the functional module may be stored in the memory of the cloud server 20, and the processor of the cloud server 20 executes the function realized by each of the above functional modules by calling the computer program stored in the memory. The functions realized by each module are described above, and are not described herein again.
In addition, the embodiment of the invention also provides a cloud server, which comprises a processor and a memory; wherein the memory has stored therein a computer program; the processor is used for executing the tea cake anti-counterfeiting method of any one of the above items by calling the computer program stored in the memory. So that the consumption end can conveniently identify whether the obtained tea cake is a genuine product or not.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (11)

1. An anti-counterfeiting method for tea cakes is characterized by comprising the following steps:
receiving the surface image of the tea cake before warehousing;
distributing a unique identification code to the surface image of the tea cake before warehousing;
storing the surface image of the tea cake and the unique identification code;
receiving a tea cake surface image obtained by the consumer end through scanning the unique identification code and uploading;
comparing the image of the surface of the tea cake before warehousing with the image of the surface of the tea cake uploaded by the consumption end, and judging whether the tea cake obtained by the consumption end is a genuine product;
and sending the judgment result to the consumption end.
2. The tea cake anti-counterfeiting method according to claim 1, wherein the tea cake has opposite obverse and reverse sides;
the receiving of the surface image of the tea cake before warehousing comprises the following steps: and receiving the front image and the back image of the tea cake before warehousing.
3. The tea cake anti-counterfeiting method according to claim 2, wherein the receiving of the front image and the back image of the tea cake before warehousing comprises:
receiving an image containing front texture and color information of the tea cake before warehousing;
segmenting the image containing the front texture and color information of the tea cake;
picking out the front image of the tea cake from the image containing the front texture and color information of the tea cake;
receiving an image containing the back texture and color information of the tea cake before warehousing;
segmenting the image containing the back texture and color information of the tea cake;
and (4) segmenting the reverse side image of the tea cake from the image containing the reverse side texture and color information of the tea cake.
4. The anti-counterfeiting method for the tea cake as claimed in claim 3, wherein the step of storing the image on the surface of the tea cake and the unique identification code comprises the following steps:
extracting feature vector information of the front image and the back image of the tea cake before warehousing by adopting a neural network;
binding the feature vector information of the front image and the back image with the unique identification code;
and storing the feature vector information of the front image and the back image and the unique identification code.
5. The tea cake anti-counterfeiting method according to claim 4, wherein the extracting of the feature vector information of the front image and the back image of the tea cake before warehousing by using the neural network comprises:
extracting feature vector information of the front image of the tea cake before warehousing by adopting a neural network;
extracting feature vector information of the reverse image of the tea cake before warehousing by adopting a neural network;
comparing the feature vector information of the front image with the feature vector information of the back image, and judging whether the feature vector information of the front image and the feature vector information of the back image are from the same surface image of the tea cake;
and if so, prompting to upload the front image and the back image of the tea cake again.
6. The anti-counterfeiting method for the tea cake as claimed in claim 4, wherein the receiving and consuming end uploads the obtained tea cake surface image by scanning the unique identification code comprises the following steps:
receiving the information of the unique identification code scanned by the consumer;
prompting the consumption end to upload the obtained front image and back image of the tea cake;
and receiving the front image and the back image of the tea cake obtained by uploading the image by the consumption end.
7. The anti-counterfeiting method for the tea cake as claimed in claim 6, wherein after the receiving and consuming end uploads the obtained front image and back image of the tea cake, the receiving and consuming end uploads the obtained surface image of the tea cake by scanning the unique identification code further comprises:
judging whether the front image and the back image of the tea cake obtained by uploading the tea cake from the consumption end can be identified or not;
and if the image cannot be identified, prompting the consumer to upload the obtained front image and back image again.
8. The tea cake anti-counterfeiting method according to claim 6, wherein the comparing the tea cake surface image before warehousing with the tea cake surface image uploaded by the consuming end to judge whether the tea cake obtained by the consuming end is a genuine product comprises:
extracting feature vector information of the front image and the back image of the tea cake uploaded by the consumption end by adopting a neural network;
comparing the characteristic vector information of the image of the front side of the tea cake before warehousing with the characteristic vector information of the image of the front side of the tea cake uploaded by the consumption end, and acquiring the similarity between the image of the front side of the tea cake before warehousing and the image of the front side of the tea cake uploaded by the consumption end;
comparing the characteristic vector information of the back image of the tea cake before warehousing with the characteristic vector information of the back image of the tea cake uploaded by the consumption end to obtain the similarity between the back image of the tea cake before warehousing and the back image of the tea cake uploaded by the consumption end;
judging whether the similarity between the image of the front side of the tea cake before warehousing and the image of the front side of the tea cake uploaded by the consumption end is greater than a set threshold value or not; if not, judging that the tea cake obtained by the consumption end is an authentic product;
if so, judging whether the similarity between the back image of the tea cake before warehousing and the back image of the tea cake uploaded by the consumption end is greater than a set threshold value; if not, judging that the tea cake obtained by the consumption end is an authentic product; if so, judging that the tea cake obtained by the consumption end is a genuine product.
9. The tea cake anti-counterfeiting method according to claim 8, wherein the extracting of the feature vector information of the front image and the back image of the tea cake before warehousing and the feature vector information of the front image and the back image of the tea cake uploaded by the consuming terminal by using a neural network specifically comprises:
building a neural network model, wherein the neural network model comprises 52 convolutional layers, 51 batch normalization layers and 1 full connection layer; the convolution kernels of the convolutional layers have convolution kernels with sizes of 7 x 7, 3 x 3 and 1 x 1, and have different convolution kernel size settings at different stages of the neural network; the convolution step is 1 or 2, and different step settings are provided at different stages of the neural network; each convolution layer is followed by a batch normalization layer, each batch normalization layer is followed by an activation function, and the activation functions are PReLU activation functions; the fully-connected layer is positioned at the last layer of the whole neural network and is used for mapping the specific dimension output of the convolutional layer to a feature vector of a specified dimension;
on a data set of a set number of training samples, selecting a cross entropy loss function to train the built neural network model; setting the total training iteration times of the neural network as set times; selecting an Adam optimizer, wherein the initial learning rate is 0.01, and the weight attenuation coefficient is 0.0005; when training iteration times are every 30k, the learning rate is attenuated to 0.1 time;
compressing and quantizing the trained neural network model to obtain a target neural network model;
inputting the front image and the back image of the tea cake before warehousing into the neural network model of the target, and extracting the characteristic vector information of the front image and the back image of the tea cake before warehousing;
and inputting the front image and the back image of the tea cake uploaded by the consumption end into the neural network model of the target, and extracting the feature vector information of the front image and the back image of the tea cake uploaded by the consumption end.
10. A tea cake anti-counterfeiting system is characterized by comprising:
the receiving module is used for receiving the surface image of the tea cake before warehousing;
the identification code distribution module is used for distributing a unique identification code to the image on the surface of the tea cake before warehousing;
the storage module is used for storing the surface image of the tea cake and the unique identification code;
the receiving module is also used for receiving a tea cake surface image obtained by scanning the unique identification code and uploading by the consumption end;
further comprising:
the comparison module is used for comparing the tea cake surface image before warehousing with the tea cake surface image uploaded by the consumption end and judging whether the tea cake obtained by the consumption end is a genuine product;
and the sending module is used for sending the judgment result to the consumption end.
11. A cloud server is characterized by comprising a processor and a memory; wherein the memory has stored therein a computer program; the processor is used for executing the tea cake anti-counterfeiting method according to any one of claims 1-9 by calling the computer program stored in the memory.
CN202011093265.2A 2020-10-13 2020-10-13 Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server Pending CN114418922A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011093265.2A CN114418922A (en) 2020-10-13 2020-10-13 Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011093265.2A CN114418922A (en) 2020-10-13 2020-10-13 Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server

Publications (1)

Publication Number Publication Date
CN114418922A true CN114418922A (en) 2022-04-29

Family

ID=81260275

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011093265.2A Pending CN114418922A (en) 2020-10-13 2020-10-13 Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server

Country Status (1)

Country Link
CN (1) CN114418922A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102465850B1 (en) 2022-06-15 2022-11-09 안성희 Pure tea identification system based on artificial intelligent learing
KR102465848B1 (en) 2022-06-15 2022-11-09 안성희 Puer tea identification system capable of certification of seller
KR102465849B1 (en) 2022-06-15 2022-11-09 안성희 Pure tea identification system based on blockchain technique
CN115690758A (en) * 2022-12-12 2023-02-03 吉林农业大学 Tea face truth verification method and truth verification system based on depth measurement learning

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102465850B1 (en) 2022-06-15 2022-11-09 안성희 Pure tea identification system based on artificial intelligent learing
KR102465848B1 (en) 2022-06-15 2022-11-09 안성희 Puer tea identification system capable of certification of seller
KR102465849B1 (en) 2022-06-15 2022-11-09 안성희 Pure tea identification system based on blockchain technique
CN115690758A (en) * 2022-12-12 2023-02-03 吉林农业大学 Tea face truth verification method and truth verification system based on depth measurement learning
CN115690758B (en) * 2022-12-12 2023-08-08 吉林农业大学 Tea face verification method and system based on deep measurement learning

Similar Documents

Publication Publication Date Title
CN114418922A (en) Tea cake anti-counterfeiting method, tea cake anti-counterfeiting system and cloud server
CN108416902B (en) Real-time object identification method and device based on difference identification
CN111191568A (en) Method, device, equipment and medium for identifying copied image
CN110532746B (en) Face checking method, device, server and readable storage medium
CN109815823B (en) Data processing method and related product
CN112906806A (en) Data optimization method and device based on neural network
CN111680181A (en) Abnormal object identification method and terminal equipment
CN112561662A (en) E-commerce platform commodity return order remote verification method based on artificial intelligence and cloud computing verification platform
CN109949050B (en) Block chain-based product identification analysis method, terminal device and storage medium
CN112668562B (en) Client identification method and device
CN113676533A (en) Transmission channel determination method and system based on data compression of Internet of things
CN109858306A (en) Internet method for anti-counterfeit and system based on natural grain
CN112836807A (en) Data processing method and device based on neural network
CN107103632A (en) Method for compressing image and device
CN111950562A (en) Picture processing method and device, terminal equipment and storage medium
CN111353526A (en) Image matching method and device and related equipment
CN113609954B (en) Social network image source identification method and system based on deep learning
CN111611993A (en) Method and device for identifying volume of food in refrigerator and computer storage medium
CN110874602A (en) Image identification method and device
CN114997884A (en) Dual anti-counterfeiting authentication system based on tea cake image and information code
CN114708247A (en) Cigarette case packaging defect identification method and device based on deep learning
CN111445267A (en) Food anti-counterfeiting traceability system and method based on RFID
CN116912606B (en) Tea cake damp analysis method and device based on convolutional neural network and storage medium
Rouhi et al. A cluster-based approach of smartphone camera fingerprint for user profiles resolution within social network
CN111797922A (en) Text image classification method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination