CN115392938B - Electronic commerce platform counterfeit commodity big data identification method based on blockchain - Google Patents
Electronic commerce platform counterfeit commodity big data identification method based on blockchain Download PDFInfo
- Publication number
- CN115392938B CN115392938B CN202211314659.5A CN202211314659A CN115392938B CN 115392938 B CN115392938 B CN 115392938B CN 202211314659 A CN202211314659 A CN 202211314659A CN 115392938 B CN115392938 B CN 115392938B
- Authority
- CN
- China
- Prior art keywords
- image
- detected
- similarity
- commodity
- reference commodity
- 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.)
- Active
Links
Images
Classifications
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/19173—Classification techniques
Abstract
The invention provides a block chain-based large data identification method for counterfeit goods of an electronic commerce platform, and relates to the field of counterfeit goods identification. A large data identification method for counterfeit goods of an electronic commerce platform based on a block chain comprises the following steps: collecting a reference commodity image of first release of commodities; calculating the similarity of the image to be detected and the reference commodity image, respectively carrying out significance detection on the image to be detected and the obtained reference commodity image when the similarity is higher, and detecting the similarity of two significant areas; the method comprises the steps that brand areas of an image to be detected and a reference commodity image are detected respectively when the similarity of the salient areas is high; respectively detecting characters in trademark areas of the image to be detected and the reference commodity image, and if the detected characters are the same and the entropy value is approximate, identifying the image to be detected and the reference commodity image as the same brand, and listing the images as suspected infringement commodity release; and the commodity infringement identification accuracy and screening efficiency are improved.
Description
Technical Field
The invention relates to the field of counterfeit commodity identification, in particular to a large data identification method for a counterfeit commodity of an electronic commerce platform based on a blockchain.
Background
With the rise of network e-commerce platforms, more and more commodities are sold on the e-commerce platforms, and great convenience is provided for enterprises and consumers. Meanwhile, the counterfeit goods inevitably flow into the electronic commerce platform, and losses are brought to regular enterprises, electronic commerce platforms and consumers. The efficiency of manually screening infringed commodities is extremely low, and the traditional image detection, identification and other related methods can detect partial counterfeit commodities, but the accuracy is not high.
The accuracy of image recognition is generally improved by adopting a traditional image denoising method, but more calculation resources are consumed, and an excellent denoising result is difficult to achieve. There is therefore a need for a method that improves the accuracy of identifying infringement of merchandise and the efficiency of screening.
Disclosure of Invention
The invention aims to provide a large data identification method for counterfeit goods of an electronic commerce platform based on a blockchain, which can improve the accuracy of identifying infringement of goods and the screening efficiency.
Embodiments of the present invention are implemented as follows:
the embodiment of the application provides a large data identification method for counterfeit goods of an electronic commerce platform based on a blockchain, which comprises the following steps:
(1) A plurality of viewers collect reference commodity images of first release of commodities through an electronic commerce platform, and link the first release platform and the reference commodity images;
(2) Acquiring an image to be detected of the commodity, calculating the similarity between the image to be detected and the reference commodity image based on an image pyramid similarity detection method, and entering a next step (3) if the similarity is higher than a preset threshold value;
(3) Performing saliency detection on the image to be detected and the obtained reference commodity image respectively, performing similarity detection on two salient regions based on similarity detection of multiple convolution kernels, entering a next step (4) if the similarity of the salient regions is higher than a preset threshold value, and performing uplink on the two salient regions with high similarity;
(4) Respectively detecting trademark areas of the image to be detected and the reference commodity image by using a target detection method, and linking the two trademark areas;
(5) Detecting characters in the trademark areas of the image to be detected and the reference commodity image respectively by utilizing an OCR technology, and judging whether the detected two characters are identical; respectively detecting entropy values of the trademark areas by utilizing an image entropy technology, and judging whether the two detected entropy values are similar; if the detected characters are the same and the entropy values are similar, the image to be detected and the reference commodity image are considered to be the same brand, the image to be detected is listed as a suspected infringement commodity to be issued, and the suspected infringement commodity is sent to an inspector for inspection; if the detected characters are the same and any one of the entropy values is similar, the images to be detected are classified as suspected counterfeit goods, and the suspected counterfeit goods are sent to an inspector for inspection; if the detected characters are different and the entropy values are not similar, the image to be detected is listed as the suspected counterfeit commodity to be released and sent to an inspector for inspection; and (5) carrying out uplink storage on the image to be detected, which is confirmed to be the counterfeit commodity after examination.
In some embodiments of the present invention, the above image pyramid-based similarity detection method specifically includes the following steps:
(a) Performing sparse coding on the reference commodity image and the image to be detected, calculating the similarity between the reference commodity image and the image to be detected by using the Euclidean distance, if the similarity is low, determining that the two images are dissimilar, and if the similarity is high, continuing to perform the next step;
(b) 4 equally dividing the reference commodity image and the image to be detected, performing sparse coding on each equally divided region, respectively calculating the similarity of 4 corresponding regions, if the similarity in the 4 regions is higher and lower than 3, determining that the two images are dissimilar, and if the similarity in the 4 regions is higher and not lower than 3, continuing to perform the next step;
(c) Dividing the reference commodity image and the image to be detected by 16 equally, performing sparse coding on each equally divided region respectively, and calculating the similarity of 16 corresponding regions respectively, wherein if the similarity in the 16 regions is higher and lower than 12, the two images are considered to be dissimilar; and if the similarity in the 16 areas is higher and is not lower than 12, the similarity between the reference commodity image and the image to be detected is higher.
In some embodiments of the present invention, the similarity detection based on multiple convolution kernels specifically includes the steps of: processing the salient region of the reference commodity image and the salient region of the image to be detected by using a smooth convolution kernel respectively to obtain a filtered result; similarity is calculated using Euclidean distance after sparse coding.
In some embodiments of the present invention, the similarity detection based on multiple convolution kernels specifically includes the steps of: the sharpened convolution is utilized to check the salient region of the reference commodity image and the salient region of the image to be detected to obtain a filtered result; similarity is calculated using Euclidean distance after sparse coding.
In some embodiments of the present invention, the similarity detection based on multiple convolution kernels specifically includes the steps of: the significant area of the reference commodity image and the significant area of the image to be detected are checked by using a denoising convolution to be processed, and a filtered result is obtained; similarity is calculated using Euclidean distance after sparse coding.
In some embodiments of the present invention, in the step (5), the detecting the entropy of the trademark region by using the image entropy technique specifically includes the following steps:
(a) Calculating a gray value i (0 < =i < =255) of each pixel point in the image;
(b) Calculating probability of occurrence of certain gray scale in the image
；
(c) Using the formula
And calculating the image entropy.
In some embodiments of the present invention, the similarity detection based on the multiple convolution kernels specifically further includes step (6): and (3) detecting the images to be detected of the plurality of commodities by utilizing the steps (1) to (5).
Compared with the prior art, the embodiment of the invention has at least the following advantages or beneficial effects:
according to the method and the device, whether the image to be detected is infringed or not is judged according to the reference commodity image by collecting the reference commodity image which is released for the first time by the commodity; the similarity between the image to be detected and the reference commodity image is utilized, so that whether the commodity belongs to the counterfeit commodity of the reference commodity image is judged; the salient areas of the reference commodity image and the image to be detected are further identified and compared, so that whether the commodity is imitated is judged according to the similarity, and the accuracy of the identification of the imitated commodity is improved; judging whether the brand belongs to a counterfeit brand or not by combining the entropy values of the trademarks of the commodity image and the reference commodity image; and the commodities with approximate trademark characters and trademark entropy values are listed as suspected counterfeit commodities, and manual examination is carried out, so that the screening efficiency of brand counterfeiting is further improved, and the rights and interests of brand parties and consumers are maintained. According to the commodity image identification method and device, commodity images are identified through multi-area identification and core area identification, so that accuracy and screening efficiency of commodity brand imitation identification are improved. The similarity detection method based on the image pyramid and the similarity detection method based on the multiple convolution kernels calculate the multi-region similarity and the core region similarity of the reference commodity image and the image to be detected, so that the recognition accuracy of the image can be improved, and the accuracy of the image counterfeit recognition can be ensured; on the basis, whether trademarks are the same or not is judged by utilizing an OCR technology and an image entropy calculation model; further utilizing the local characteristics to confirm the counterfeit infringement behavior of the commodity; the method has the advantages that the reference commodity images are collected by utilizing a plurality of browses looking up the first commodity release platform, so that the evidence integrity and reliability of the infringed commodity are ensured, and the data can be conveniently and backtracked and inquired; the concept depth of the blockchain can be applied to the checking of commodity counterfeit infringement. The method and the device can improve accuracy and efficiency of identifying the counterfeit goods, and are convenient to identify the counterfeit goods applied to mass images. In the whole identification process of the counterfeit commodity, the acquired first release platform and the reference commodity image are linked, so that the evidence protection is enhanced, and the auditing and later commodity maintenance are facilitated; the to-be-detected image and the reference commodity image with high similarity, the two obvious areas with high similarity and the two identified trademark areas are stored in a uplink mode, the risk of data damage is prevented, and the reliability of commodity counterfeiting evidence is improved; the manual checking of each item of information in turn is facilitated for screening; and the checked counterfeit commodities are stored in a uplink manner, so that the infringed commodities can be managed uniformly and the commodity maintenance processing can be performed.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a block chain-based electronic commerce platform counterfeit commodity big data identification method in an embodiment of the invention;
fig. 2 is a schematic flow chart of step (5) in the embodiment of the invention.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, as provided in the accompanying drawings, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
In the description of the present application, it should also be noted that, unless explicitly specified and limited otherwise, the terms "disposed," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the terms in this application will be understood by those of ordinary skill in the art in a specific context.
Some embodiments of the present application are described in detail below with reference to the accompanying drawings. The various embodiments and features of the embodiments described below may be combined with one another without conflict.
Examples
Referring to fig. 1 to fig. 2, fig. 1 to fig. 2 are schematic flow diagrams of a method for identifying big data of a block chain-based electronic commerce platform counterfeit commodity provided in an embodiment of the present application. The block chain-based electronic commerce platform counterfeit commodity big data identification method comprises the following steps:
(1) A plurality of viewers collect reference commodity images of first release of commodities through an electronic commerce platform, and link the first release platform and the reference commodity images;
(2) Acquiring an image to be detected of the commodity, calculating the similarity between the image to be detected and the reference commodity image based on an image pyramid similarity detection method, and entering a next step (3) if the similarity is higher than a preset threshold value;
(3) Performing saliency detection on the image to be detected and the obtained reference commodity image respectively, performing similarity detection on two salient regions based on similarity detection of multiple convolution kernels, entering a next step (4) if the similarity of the salient regions is higher than a preset threshold value, and performing uplink on the two salient regions with high similarity;
(4) Respectively detecting trademark areas of the image to be detected and the reference commodity image by using a target detection method, and linking the two trademark areas;
(5) Detecting characters in the trademark areas of the image to be detected and the reference commodity image respectively by utilizing an OCR technology, and judging whether the detected two characters are identical; respectively detecting entropy values of the trademark areas by utilizing an image entropy technology, and judging whether the two detected entropy values are similar; if the detected characters are the same and the entropy values are similar, the image to be detected and the reference commodity image are considered to be the same brand, the image to be detected is listed as a suspected infringement commodity to be issued, and the suspected infringement commodity is sent to an inspector for inspection; if the detected characters are the same and any one of the entropy values is similar, the images to be detected are classified as suspected counterfeit goods, and the suspected counterfeit goods are sent to an inspector for inspection; if the detected characters are different and the entropy values are not similar, the image to be detected is listed as the suspected counterfeit commodity to be released and sent to an inspector for inspection; and (5) carrying out uplink storage on the image to be detected, which is confirmed to be the counterfeit commodity after examination.
The method comprises the steps of acquiring various matched reference commodity images according to a commodity authentication platform or a commodity first release platform. The reference commodity image of each brand can comprise a plurality of commodities, and each commodity can comprise images with a plurality of directions, so that the images are utilized for precise identification, and more accurate counterfeit commodities are obtained. In detail, the first release platform for collecting the reference commodity image is recorded by the user who browses the commodity, and the first release platform for the commodity image is confirmed according to the number of the recorded people. The two preset thresholds represent different values and can be set automatically or manually according to actual conditions. The method can jointly maintain the first release record of the commodity by a plurality of browses to record the commodity image, and the idea depth of the blockchain is applied to the verification of infringement and counterfeit behaviors. For example, when 300 people watch the commodity image on the e-commerce platform at the same time, 300 people record that the commodity is released for the first time by a certain mall platform of the e-commerce platform at the same time. When the number of the mall platforms recorded by all people is multiple, the reference commodity images with multiple recorded people can be selected for collection.
The website of the first release platform and the reference commodity image are linked, so that evidence is reserved, and manual auditing and commodity maintenance are facilitated in the later period; the image to be detected with high similarity and the reference commodity image are linked, so that comparison is convenient during manual auditing; the two significant areas with high similarity and the two trademark areas which are correspondingly identified are linked, so that accurate identification and manual rechecking are convenient; and (3) the images to be detected of the suspected infringing commodity and the suspected counterfeit commodity which are confirmed to be the counterfeit commodity after being inspected are stored in a uplink mode, and unified management and right-maintaining treatment can be carried out. Optionally, the commodity information in the first release platform website is stored in a uplink mode at the same time, and the record browsed by the browser can be stored in the uplink mode at the same time, so that the data reliability of the first release commodity is further improved. Optionally, the suspected infringed commodity and the suspected counterfeit commodity before examination are also stored in a uplink mode, so that operation evidence for issuing information can be stored after manual review and right-maintaining treatment are carried out by people. The release of the suspected counterfeit commodity and the suspected infringing commodity can be disclosed to related commodity people or public, so that the suspected counterfeit commodity and the suspected infringing commodity are convenient to be subjected to right maintenance and complaint, and can be only indicated as release to auditing personnel for auditing.
Alternatively, the collected image to be detected may be an image to be detected of a commodity source that is not brand-certified, or may be a commodity image uploaded by a user (consumer) by himself. Optionally, the collected multiple reference commodity images of multiple brands can be classified and stored, so that the reference commodity images of one or multiple brands can be manually selected to accurately identify the image to be detected. In detail, the to-be-detected image and the obtained reference commodity image are respectively subjected to significance detection, so that similarity judgment is carried out on the significance areas detected by the to-be-detected image and the obtained reference commodity image. When the similarity of the salient areas is high, the trademark is detected in the next step. In detail, a trademark region is recognized by using a target detection method, and characters therein are recognized by using an OCR character recognition technology. And calculating an image entropy value by using the trademark region, thereby judging the approximation degree of the entropy value. And judging the commodity to be suspected counterfeit when the recognized characters are the same and the entropy value similarity is low, and judging the commodity to be suspected counterfeit when the recognized characters are different and the entropy value similarity is high. When the recognized characters are the same and the entropy value similarity is high, the characters are judged to be the same brand, so that supervision personnel, brand parties or customers can conveniently conduct manual screening, the manual screening range is reduced, and counterfeit goods are found out according to images to be detected of the same brand. The method can authenticate the genuine products or the counterfeit products according to the goods sources of the same brand, and is convenient for supervision and screening. The method for detecting the saliency can be realized by any visual perception technology, such as a saliency detection method based on visual perception positive feedback disclosed in reference to patent application number CN 201611202475.4. OCR text recognition techniques, image entropy techniques, and the like are all prior art and need not be described in detail herein.
In detail, the similarity of multiple local positions is analyzed through multiple convolution kernels, so that the accuracy of counterfeit feature identification is further improved. The salient regions of the image to be detected and the reference commodity image can be one or more. Optionally, when the number of the salient regions is plural, the relative positions of the salient regions are used for one-to-one comparison, so as to judge the similarity. And calculating the similarity between the image to be detected and the reference commodity image by using the similarity detection method of the image pyramid, so that the accuracy of image identification is improved. After the image to be detected and the reference commodity image are the same brand, the non-authenticated commodity can be sent to a censor for verification according to whether the issuing platform is authenticated or not, and the non-authenticated commodity is sent to a consumer after verification is successful. The method can send commodities of the same brand to the censor of the brand party for examination, and send suspected counterfeit commodities to the censor of the supervision department for examination, so that the examination pressure is reduced, and the efficiency of maintaining the brand and consumer rights is greatly improved. Wherein brand identity is indicated when the brand characters are identical, and is considered a suspected counterfeit product when the two characters are not identical. When the trademark characters are the same and the entropy values of the trademark areas are approximate, the commodity is a suspected infringement commodity, and whether the commodity is counterfeited or not is confirmed after manual auditing. In the step (6), the images to be detected which are confirmed to be the counterfeit goods after examination are stored in a uplink mode and are stored and acquired by examination personnel at any time, data are prevented from being lost and tampered, unified management of the counterfeit goods is facilitated, and therefore supervision of the counterfeit goods is enhanced.
In some embodiments of the present invention, the above image pyramid-based similarity detection method specifically includes the following steps:
(a) Performing sparse coding on the reference commodity image and the image to be detected, calculating the similarity between the reference commodity image and the image to be detected by using the Euclidean distance, if the similarity is low, determining that the two images are dissimilar, and if the similarity is high, continuing to perform the next step;
(b) 4 equally dividing the reference commodity image and the image to be detected, performing sparse coding on each equally divided region, respectively calculating the similarity of 4 corresponding regions, if the similarity in the 4 regions is higher and lower than 3, determining that the two images are dissimilar, and if the similarity in the 4 regions is higher and not lower than 3, continuing to perform the next step;
(c) Dividing the reference commodity image and the image to be detected by 16 equally, performing sparse coding on each equally divided region respectively, and calculating the similarity of 16 corresponding regions respectively, wherein if the similarity in the 16 regions is higher and lower than 12, the two images are considered to be dissimilar; and if the similarity in the 16 areas is higher and is not lower than 12, the similarity between the reference commodity image and the image to be detected is higher.
In detail, the dividing manner of the equal division area is any one, for example, taking a rectangular image as an example, and dividing the image into four parts by the center lines of the long side and the short side. Likewise, the image to be detected may be divided into 16 equal divisions in a rectangular distribution manner. Wherein after division into the respective parts, the corresponding region, such as the quarter, corresponds to the upper left part and the lower right part of the rectangle. And multiple judgment of partial and whole images is carried out, and the similarity of the images is judged, so that the investigation range can be reduced, and the counterfeit commodity with the brand height similar to that of the brand is obtained.
In some embodiments of the present invention, the similarity detection based on multiple convolution kernels specifically includes the steps of: processing the salient region of the reference commodity image and the salient region of the image to be detected by using a smooth convolution kernel respectively to obtain a filtered result; similarity is calculated using Euclidean distance after sparse coding.
In some embodiments of the present invention, the similarity detection based on multiple convolution kernels specifically includes the steps of: the sharpened convolution is utilized to check the salient region of the reference commodity image and the salient region of the image to be detected to obtain a filtered result; similarity is calculated using Euclidean distance after sparse coding.
In some embodiments of the present invention, the similarity detection based on multiple convolution kernels specifically includes the steps of: the significant area of the reference commodity image and the significant area of the image to be detected are checked by using a denoising convolution to be processed, and a filtered result is obtained; similarity is calculated using Euclidean distance after sparse coding.
In detail, the size of the convolution kernel is typically 3*3, and selecting different types of multi-convolution kernels can be adapted for different types of image use. By using any one of the three steps to obtain a result with higher similarity, the similarity between the salient region of the reference commodity image and the salient region of the image to be detected can be determined to be higher. Optionally, the calculated similarity can be combined with various detection methods, and a certain weight is added according to the image type, so that the similarity with higher accuracy is obtained.
In some embodiments of the present invention, in the step (5), the detecting the entropy of the trademark region by using the image entropy technique specifically includes the following steps:
(a) Calculating a gray value i (0 < =i < =255) of each pixel point in the image;
(b) Calculating probability of occurrence of certain gray scale in the image
；
(c) Using the formula
And calculating the image entropy.
And calculating a gray value and a probability of occurrence of the gray value, calculating an entropy value, judging the similarity of the corresponding trademark areas, and checking whether the obtained trademark is a counterfeit product or not by utilizing the trademark similarity with high overall similarity degree of the obtained product.
Optionally, the detection result of the commodity image or the inspected result is collected to obtain multiple groups of commodity detection data, each group of commodity detection data comprises an image to be detected, a corresponding reference commodity image and a detection result for marking whether the commodity belongs to a counterfeit commodity, and a counterfeit commodity training model for outputting the detection result is obtained through machine learning training. The data input by the counterfeit commodity training model is a detected commodity image, so that whether counterfeit and the corresponding brand of the counterfeit commodity are judged. Alternatively, the input data may be the detected commodity image and the reference commodity image, so that the corresponding detection result is output.
Compared with the prior art, the embodiment of the invention has at least the following advantages or beneficial effects:
according to the method and the device, whether the image to be detected is infringed or not is judged according to the reference commodity image by collecting the reference commodity image which is released for the first time by the commodity; the similarity between the image to be detected and the reference commodity image is utilized, so that whether the commodity belongs to the counterfeit commodity of the reference commodity image is judged; the salient areas of the reference commodity image and the image to be detected are further identified and compared, so that whether the commodity is imitated is judged according to the similarity, and the accuracy of the identification of the imitated commodity is improved; judging whether the brand belongs to a counterfeit brand or not by combining the entropy values of the trademarks of the commodity image and the reference commodity image; and the commodities with approximate trademark characters and trademark entropy values are listed as suspected counterfeit commodities, and manual examination is carried out, so that the screening efficiency of brand counterfeiting is further improved, and the rights and interests of brand parties and consumers are maintained. According to the commodity image identification method and device, commodity images are identified through multi-area identification and core area identification, so that accuracy and screening efficiency of commodity brand imitation identification are improved. The similarity detection method based on the image pyramid and the similarity detection method based on the multiple convolution kernels calculate the multi-region similarity and the core region similarity of the reference commodity image and the image to be detected, so that the recognition accuracy of the image can be improved, and the accuracy of the image counterfeit recognition can be ensured; on the basis, whether trademarks are the same or not is judged by utilizing an OCR technology and an image entropy calculation model; further utilizing the local characteristics to confirm the counterfeit infringement behavior of the commodity; the method has the advantages that the reference commodity images are collected by utilizing a plurality of browses looking up the first commodity release platform, so that the evidence integrity and reliability of the infringed commodity are ensured, and the data can be conveniently and backtracked and inquired; the concept depth of the blockchain can be applied to the checking of commodity counterfeit infringement. The method and the device can improve accuracy and efficiency of identifying the counterfeit goods, and are convenient to identify the counterfeit goods applied to mass images. In the whole identification process of the counterfeit commodity, the acquired first release platform and the reference commodity image are linked, so that the evidence protection is enhanced, and the auditing and later commodity maintenance are facilitated; the to-be-detected image and the reference commodity image with high similarity, the two obvious areas with high similarity and the two identified trademark areas are stored in a uplink mode, the risk of data damage is prevented, and the reliability of commodity counterfeiting evidence is improved; the manual checking of each item of information in turn is facilitated for screening; and the checked counterfeit commodities are stored in a uplink manner, so that the infringed commodities can be managed uniformly and the commodity maintenance processing can be performed.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners as well. The apparatus embodiments described above are merely illustrative, for example, flow diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the flowchart illustrations, and combinations of blocks in the flowchart illustrations, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, the functional modules in the embodiments of the present application may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing description is only of the preferred embodiments of the present application and is not intended to limit the same, but rather, various modifications and variations may be made by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (7)
1. The large data identification method for the counterfeit commodity of the electronic commerce platform based on the block chain is characterized by comprising the following steps:
(1) A plurality of viewers collect reference commodity images of first release of commodities through an electronic commerce platform, and uplink the first release platform and the reference commodity images;
(2) Acquiring an image to be detected of the commodity, calculating the similarity of the image to be detected and the reference commodity image based on an image pyramid similarity detection method, stopping calculation and entering the next step (3) if the similarity is higher than a preset threshold value, and linking the image to be detected and the reference commodity image with high similarity;
(3) Performing saliency detection on the image to be detected and the obtained reference commodity image respectively, performing similarity detection on two salient regions based on similarity detection of multiple convolution kernels, entering a next step (4) if the similarity of the salient regions is higher than a preset threshold, and performing uplink on the two salient regions with high similarity;
(4) Respectively detecting trademark areas of the image to be detected and the reference commodity image by using a target detection method, and linking the two trademark areas;
(5) Detecting characters in the trademark areas of the image to be detected and the reference commodity image respectively by utilizing an OCR technology, and judging whether the detected two characters are identical; respectively detecting entropy values of the trademark areas by utilizing an image entropy technology, and judging whether the two detected entropy values are similar; if the detected characters are the same and the entropy values are similar, the image to be detected and the reference commodity image are considered to be the same brand, the image to be detected is listed as a suspected infringement commodity to be issued, and the suspected infringement commodity is sent to an inspector for inspection; if the detected characters are the same and any one of the entropy values is similar, the images to be detected are classified as suspected counterfeit goods, and the suspected counterfeit goods are sent to an inspector for inspection; if the detected characters are different and the entropy values are not similar, the images to be detected are listed as the suspected counterfeit goods to be released and sent to an inspector for inspection; and (5) carrying out uplink storage on the image to be detected, which is confirmed to be the counterfeit commodity after examination.
2. The method for identifying big data of counterfeit goods by using an electronic commerce platform based on a blockchain as claimed in claim 1, wherein the method for detecting similarity based on an image pyramid specifically comprises the following steps:
(a) Performing sparse coding on the reference commodity image and the image to be detected, calculating the similarity between the reference commodity image and the image to be detected by using the Euclidean distance, if the similarity is low, determining that the two images are dissimilar, and if the similarity is high, continuing to perform the next step;
(b) Dividing the reference commodity image and the image to be detected by 4 equally, performing sparse coding on each equally divided region respectively, calculating the similarity of 4 corresponding regions respectively, if the similarity in the 4 regions is higher and lower than 3, determining that the two images are dissimilar, and if the similarity in the 4 regions is higher and not lower than 3, continuing to perform the next step;
(c) Dividing the reference commodity image and the image to be detected by 16 equally, performing sparse coding on each equally divided region respectively, and calculating the similarity of 16 corresponding regions respectively, wherein if the similarity in the 16 regions is higher and lower than 12, the two images are considered to be dissimilar; and if the similarity in the 16 areas is higher and is not lower than 12, the similarity between the reference commodity image and the image to be detected is higher.
3. The method for identifying large data of counterfeit goods on the electronic commerce platform based on the blockchain as claimed in claim 1, wherein the similarity detection based on the multi-convolution kernel comprises the following steps: processing the salient region of the reference commodity image and the salient region of the image to be detected respectively by using a smooth convolution kernel to obtain a filtered result; similarity is calculated using Euclidean distance after sparse coding.
4. The method for identifying large data of counterfeit goods on the electronic commerce platform based on the blockchain as claimed in claim 3, wherein the similarity detection based on the multi-convolution kernel specifically comprises the following steps: the significant area of the reference commodity image and the significant area of the image to be detected are checked by utilizing sharpening convolution to be processed, and a filtered result is obtained; similarity is calculated using Euclidean distance after sparse coding.
5. The method for identifying large data of counterfeit goods on the electronic commerce platform based on the blockchain as claimed in claim 3, wherein the similarity detection based on the multi-convolution kernel specifically comprises the following steps: checking the salient region of the reference commodity image and the salient region of the image to be detected by using a denoising convolution to obtain a filtered result; similarity is calculated using Euclidean distance after sparse coding.
6. The method for identifying large data of counterfeit goods by using a blockchain-based electronic commerce platform as defined in claim 1, wherein in the step (5), the detecting entropy values of the trademark areas by using an image entropy technology comprises the following steps:
(a) Calculating a gray value i (0 < =i < =255) of each pixel point in the image;
(b) Calculating probability of occurrence of certain gray scale in the image
；
(c) Using the formula
And calculating the image entropy.
7. The blockchain-based electronic commerce platform counterfeit commodity big data identification method as in claim 1, further comprising the step of (6): and (3) detecting the images to be detected of the plurality of commodities by utilizing the steps (1) to (5).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211314659.5A CN115392938B (en) | 2022-10-26 | 2022-10-26 | Electronic commerce platform counterfeit commodity big data identification method based on blockchain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211314659.5A CN115392938B (en) | 2022-10-26 | 2022-10-26 | Electronic commerce platform counterfeit commodity big data identification method based on blockchain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115392938A CN115392938A (en) | 2022-11-25 |
| CN115392938B true CN115392938B (en) | 2023-06-02 |
Family
ID=84128474
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211314659.5A Active CN115392938B (en) | 2022-10-26 | 2022-10-26 | Electronic commerce platform counterfeit commodity big data identification method based on blockchain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115392938B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408334A (en) * | 2014-11-17 | 2015-03-11 | 广东工业大学 | Design patent early-warning method |
| CN112507160A (en) * | 2020-12-03 | 2021-03-16 | 平安科技（深圳）有限公司 | Automatic judgment method and device for trademark infringement, electronic equipment and storage medium |
| CN114282027A (en) * | 2021-12-24 | 2022-04-05 | 北京畅元国讯科技有限公司 | Data processing method, device, apparatus, storage medium, and program |
| CN115205866A (en) * | 2022-09-16 | 2022-10-18 | 北京吉道尔科技有限公司 | Block chain-based scientific and technological paper big data plagiarism detection method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11443332B2 (en) * | 2014-12-22 | 2022-09-13 | Superior Integrated Solutions Llc | System, method, and software for predicting the likelihood of selling automotive commodities |
-
2022
- 2022-10-26 CN CN202211314659.5A patent/CN115392938B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408334A (en) * | 2014-11-17 | 2015-03-11 | 广东工业大学 | Design patent early-warning method |
| CN112507160A (en) * | 2020-12-03 | 2021-03-16 | 平安科技（深圳）有限公司 | Automatic judgment method and device for trademark infringement, electronic equipment and storage medium |
| CN114282027A (en) * | 2021-12-24 | 2022-04-05 | 北京畅元国讯科技有限公司 | Data processing method, device, apparatus, storage medium, and program |
| CN115205866A (en) * | 2022-09-16 | 2022-10-18 | 北京吉道尔科技有限公司 | Block chain-based scientific and technological paper big data plagiarism detection method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115392938A (en) | 2022-11-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10552848B2 (en) | Method and system for determining whether a barcode is genuine using a deviation from an idealized grid | |
| US11455786B2 (en) | Systems and methods of image processing for remote validation | |
| Dong et al. | Casia image tampering detection evaluation database | |
| US10482303B2 (en) | Methods and a system for verifying the authenticity of a mark | |
| CN102419777B (en) | System and method for filtering internet image advertisements | |
| Artaud et al. | Find it! fraud detection contest report | |
| US10387703B2 (en) | Methods and system for verifying an authenticity of a printed item | |
| EP3111376B1 (en) | Methods and a system for verifying the identity of a printed item | |
| WO2015022538A1 (en) | Image identification marker and method | |
| EP3742340B1 (en) | Methods and a system for verifying the authenticity of a mark | |
| CN107169777B (en) | Overlay code anti-counterfeiting method and system applying same | |
| CN104244016B (en) | A kind of H264 video contents altering detecting method | |
| CN110532855A (en) | Natural scene certificate image character recognition method based on deep learning | |
| CN106056183B (en) | The printed medium of printing press readable image and the system and method for scanning the image | |
| CN110428028A (en) | Fake method, device, equipment and medium are tested in identification based on Quasi dynamic laser label | |
| CN103871044A (en) | Image signature generating method and image verifying method and device | |
| CN114626024A (en) | Internet infringement video low-consumption detection method and system based on block chain | |
| CN115600249A (en) | Meta-universe e-commerce shopping big data security protection method and system based on block chain | |
| CN112487982A (en) | Merchant information auditing method, system and storage medium | |
| CN110533704A (en) | Fake method, device, equipment and medium are tested in the identification of ink label | |
| CN115392938B (en) | Electronic commerce platform counterfeit commodity big data identification method based on blockchain | |
| CN116168275B (en) | Lightweight dual-attention mechanism identification method based on feature grouping and channel replacement | |
| Alkhowaiter et al. | Evaluating perceptual hashing algorithms in detecting image manipulation over social media platforms | |
| CN115171144B (en) | Intelligent education test paper big data evaluation method and system based on block chain | |
| CN112560989B (en) | Artificial intelligence anti-counterfeiting image identification method and system based on big data |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230506 Address after: Part of Unit 2 # 7+8, 2nd Floor, No. 10 Huasheng Road, Yuzhong District, Chongqing, 400043 Applicant after: Chongqing Yanglian Data Technology Co.,Ltd. Address before: Room 222, 3-16, floors 2-3, No. 3, Maizidian West Road, Chaoyang District, Beijing 100020 Applicant before: Beijing jidaoer Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |