CN113609319A - Commodity searching method, device and equipment - Google Patents
Commodity searching method, device and equipment Download PDFInfo
- Publication number
- CN113609319A CN113609319A CN202110846166.5A CN202110846166A CN113609319A CN 113609319 A CN113609319 A CN 113609319A CN 202110846166 A CN202110846166 A CN 202110846166A CN 113609319 A CN113609319 A CN 113609319A
- Authority
- CN
- China
- Prior art keywords
- commodity
- image
- database
- target
- searching
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/53—Querying
- G06F16/532—Query formulation, e.g. graphical querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
Abstract
The embodiment of the invention provides a commodity searching method, a commodity searching device and commodity searching equipment, wherein the commodity searching method comprises the following steps: the method comprises the steps of obtaining a real shot image containing a target commodity, determining a reference commodity image matched with the visual characteristics of the real shot image from a commodity database, searching the target commodity image matched with the reference commodity image from the commodity database, and outputting commodity information corresponding to the target commodity image. The reference commodity image matched with the visual characteristics of the real shot image is determined from the commodity database, and the reference commodity image is used as an input condition for commodity search instead of the real shot image, so that the influence of the quality of the real shot image on the commodity search result can be reduced, and the accuracy of the commodity search result is improved.
Description
Technical Field
The invention relates to the technical field of internet, in particular to a commodity searching method, a commodity searching device and commodity searching equipment.
Background
In an e-commerce scenario, there are various ways for a user to search for a commodity, such as searching through some category options provided on a search interface, searching through input of keywords, and searching through input of an image containing a commodity.
For example, when a user wants to purchase a certain commodity, the user can directly shoot a real shot image containing the commodity and search for the same commodity through the real shot image.
However, the quality of the real shot image taken by the user is affected by the shooting angle, the shooting environment, the light, and the like, and the quality may be poor, so that the accuracy of the search result is poor when the product search is directly performed through the real shot image.
Disclosure of Invention
The embodiment of the invention provides a commodity searching method, a commodity searching device and commodity searching equipment, which are used for improving the accuracy of a commodity searching result.
In a first aspect, an embodiment of the present invention provides a method for searching for a commodity, where the method includes:
acquiring a real shot image containing a target commodity;
determining a reference commodity image matched with the visual characteristics of the real shot image from a commodity database;
searching the commodity database for a target commodity image matching the reference commodity image, and
and outputting the commodity information corresponding to the target commodity image.
In a second aspect, an embodiment of the present invention provides an article search device, including:
the acquisition module is used for acquiring a real shot image containing a target commodity;
the determining module is used for determining a reference commodity image matched with the visual characteristics of the real shooting image from a commodity database;
and the searching module is used for searching the target commodity image matched with the reference commodity image in the commodity database and outputting the commodity information corresponding to the target commodity image.
In a third aspect, an embodiment of the present invention provides an electronic device, including: a memory, a processor; wherein the memory has stored thereon executable code which, when executed by the processor, causes the processor to perform the item search method of the first aspect.
In a fourth aspect, an embodiment of the present invention provides a non-transitory machine-readable storage medium having stored thereon executable code, which, when executed by a processor of an electronic device, causes the processor to execute the item search method according to the first aspect.
According to the scheme provided by the embodiment of the invention, the real shot image containing the target commodity is firstly obtained, then the reference commodity image matched with the visual characteristics of the real shot image is determined from the commodity database, the target commodity image matched with the reference commodity image is searched in the commodity database, and finally the commodity information corresponding to the target commodity image is output. When the merchant uploads the commodity related information, the commodity image is provided for the commodity database, the reference commodity image matched with the visual feature of the real shot image is determined from the commodity database, and the reference commodity image is used as an input condition for commodity search instead of the real shot image, so that the influence of the quality of the real shot image on the commodity search result can be reduced, and the accuracy of the commodity search result is improved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a flowchart of a method for searching for a commodity according to an embodiment of the present invention;
fig. 2 is a schematic application diagram of a commodity search method according to an embodiment of the present invention;
FIG. 3 is a flowchart of another method for searching for merchandise according to an embodiment of the present invention;
fig. 4 is a schematic view of an operation interface in a commodity search process according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a commodity searching apparatus according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device corresponding to the article searching apparatus provided in the embodiment shown in fig. 5.
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 some, but not all, embodiments of the present invention. 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 addition, the sequence of steps in each method embodiment described below is only an example and is not strictly limited.
In practical applications, more and more users prefer to perform commodity search by using a real shot image of a commodity, and the user refers to a user who needs to perform commodity search. For example, when a user sees a commodity and needs to purchase the commodity, the user can take a picture of the commodity by using a mobile phone to obtain a real shot image, and the real shot image is used for searching the corresponding commodity. However, because the real-shot image is greatly influenced by the shooting angle, the light, the pixels of the equipment and the like in the environment, and because of marketing, eye absorption, sales promotion and other reasons of many merchants, the angle, the detail characteristics, the background and the like displayed by the commodity image can be correspondingly designed and packaged, so that a large difference exists between the commodity image and the real-shot image, and the factors cause that the accuracy of the commodity result searched by the user directly using the real-shot image as the search condition is not high, so that the experience of the user is influenced.
The commodity searching method provided by the embodiment of the invention can be executed by an electronic device, and the electronic device can be a terminal device such as a PC (personal computer), a notebook computer, a smart phone and the like, and can also be a server. The server may be a physical server including an independent host, or may also be a virtual server, or may also be a server or a server cluster in the cloud.
Fig. 1 is a flowchart of a method for searching for a product according to an embodiment of the present invention, as shown in fig. 1, the method may include the following steps:
101. and acquiring a real shooting image containing the target commodity.
102. And determining a reference commodity image matched with the visual characteristics of the real shot image from the commodity database.
103. And searching the target commodity image matched with the reference commodity image in the commodity database, and outputting the commodity information corresponding to the target commodity image.
In the embodiment of the invention, the searched commodity can be any kind of externally packaged article such as milk, snack, fresh food, clothing, cosmetics, daily necessities and the like. The purpose of searching for a product may be to search for various product information such as the price of a certain product and various components contained in a certain product.
When a user has a search demand for a certain target commodity, the user can take a picture of the target commodity and search by using the obtained real shot image containing the target commodity. For example, when a user sees an advertisement of milk powder in a subway and wants to obtain specific information of the milk powder, a photo is taken of a milk powder packaging box in the advertisement, and the photo is used as a real image.
In the embodiment of the invention, the real shooting image shot by the user is not directly used as the final search condition of the commodity search, but only the real shooting image of the user is used as the initial search condition.
Specifically, in a mode that a merchant sells goods online through an e-commerce platform, a database containing information of a plurality of goods is maintained in the e-commerce platform, and the database is called a goods database. The product database may include a plurality of kinds of product information such as a product image, a product title, and a price corresponding to one product. The commodity image is generally an image which is uploaded by a merchant and has a certain design and better image quality. In the embodiment of the invention, the photographed image of the user is used as an initial search condition, the commodity image matched with the visual characteristics of the photographed image is firstly searched in the commodity database and is called as a reference commodity image, and then the reference commodity image is used as a final search condition of commodity search, and the commodity image matched with the reference commodity image is continuously searched in the commodity database and is called as a target commodity image. Finally, the product information corresponding to the target product image is output.
In the embodiment of the invention, in the process of determining the reference commodity image, the visual similarity between the commodity image in the commodity database and the real shot image can be determined according to the visual characteristics of the real shot image, so that the commodity image of which the visual similarity with the real shot image meets the set conditions is determined from the commodity database as the reference commodity image.
The visual features (or appearance features) of the image may include color, outline, texture, shape, and other features. It should be noted that the visual features described in the embodiments of the present invention are understood to be what a long-term object can be seen from an image, and the meaning of the object should not be understood. For example, assuming a line of purple text in the live image, the visual feature is a purple stroke and should not be understood as meaning what the line of text is.
In practical application, a neural network model may be trained in advance, and the neural network model may be configured to implement functions of extracting visual features of images and calculating the similarity of the visual features between different images. Based on this, it is possible to extract visual features of the captured image and each product image and calculate visual similarity between the captured image and each product image by using the model for one captured image captured by the user and each product image included in the product database. And then, determining the commodity image with the visual similarity meeting the set conditions with the real shot image from the commodity database as a reference commodity image. Wherein, optionally, the setting condition may be: the visual similarity is greater than a set threshold. Optionally, when the number of the commodity images with the visual similarity between the determined visual features and the visual features of the real shot images is more than 1, determining one commodity image with the largest visual similarity as a reference commodity image; or, the reference commodity image can be screened out by combining other factors, such as the attention degree of the corresponding commodity.
That is, alternatively, a product image, in which the visual similarity with the photographed image meets the set condition and the corresponding product attention meets the set condition, may also be determined from the product database as the reference product image. For example, the setting condition that the visual similarity meets may be that the visual similarity is greater than a set threshold, and the setting condition that the commodity attention meets may be the attention ranking topN.
In practical application, the visual similarity may be used as a screening condition, and when the number of the commodity images determined from the commodity database and having the visual similarity with the live image greater than a set threshold reaches a set number (for example, exceeds 2), the commodity attention may be further combined for screening.
Specifically, the commodity attention corresponding to the commodity image may be measured by any one of the following indexes: the transaction rate of the product corresponding to the product image, the click rate corresponding to the product image, and the like. The trading rate of the commodity corresponding to the commodity image can be determined by counting the trading volume of the commodity in a set time range; the click rate corresponding to the product image can be determined by counting the number of times the product image is clicked within a set time range. The measure of the attention of the commodity is not limited to these two exemplary measures.
For example, if 10 commodity images with a visual similarity greater than a set threshold (e.g., 80%) are selected from the commodity database according to the visual similarity, then 3 commodity images are selected as the reference commodity images according to the commodity attention degrees corresponding to the 10 commodity images, wherein the commodity attention degree of the 3 commodity images is higher than that of the unselected 7 commodity images, that is, top 3.
In addition, in practical application, no matter how many reference commodity images are finally obtained, optionally, after the reference commodity images are obtained, the reference commodity images can be output firstly for the user to confirm whether the reference commodity images are adopted or not. If the user confirms to adopt the output reference commodity image, optionally, the reference commodity image can replace the initial photographed image, and the target commodity image matched with the reference commodity image is searched in the commodity database; alternatively, both the reference product image and the initial photographed image may be used as search conditions, and the target product image matched with the reference product image and the photographed image may be searched in the product database.
Of course, it is understood that when the number of the determined reference commodity images is plural, the user may select one or several reference commodity images that are closest to the search intention thereof. In addition, on the basis that the user confirms to use all or part of the reference commodity images, the live-shot images can still be used as one of the search conditions, because the commodity images with the visual similarity larger than the set threshold value with the live-shot images can be searched in the commodity database, which often indicates that the shooting quality of the live-shot images is good, and at this time, the live-shot images of the user are added as recall conditions (namely the search conditions) of the target commodities to be searched, so that the recall rate of the target commodities can be improved.
Assuming that the user confirms the search for the target commodity using the reference commodity image, the target commodity image matching the reference commodity image may be searched in the commodity database with the reference commodity image as a search condition. In this case, a visual similarity threshold may be set, and the commodity image whose visual similarity with the reference commodity image is greater than the threshold may be used as the target commodity image. Finally, product information corresponding to the target product image, which is considered to be product information related to the target product, is output to the user. In fact, at least one commodity list interface may be output to the user, each commodity list interface includes a plurality of commodity link information, and the user enters a corresponding commodity introduction interface by clicking the link information, where various detailed information of the corresponding commodity is included in the commodity introduction interface.
In summary, the real-shot image of the commodity of the user and the commodity image of the commodity in the commodity database of the e-commerce platform often have some differences, and these differences result in poor accuracy of the search result of directly searching for the commodity by using the real-shot image. The reference commodity image matched with the visual characteristics of the real shooting image shot by the user at present is determined from the commodity database, and the reference commodity image is used as an input condition for commodity search instead of the real shooting image, so that the influence of the quality of the real shooting image on the commodity search result can be reduced, and the accuracy of the commodity search result is improved.
For ease of understanding, a scenario in which a user searches for a commodity is illustrated in conjunction with fig. 2.
In fig. 2, it is assumed that the product that the user needs to search for is a bottle of milk, and the user wants to obtain the information related to the price of the bottle of milk. The user photographs the bottle of milk to be searched to obtain a real photographed image as illustrated in fig. 2, then extracts visual features of the real photographed milk image and the commodity image in the commodity database, compares the extracted visual features of the real photographed milk image with the visual features of the commodity image in the commodity database to determine a reference commodity image matched with the visual features of the real photographed milk image, then searches the commodity database for a target commodity image matched with the reference commodity image, and finally outputs commodity information corresponding to the target commodity image.
In the above embodiment, in the process of performing the reference product image search, it is assumed that a product image having a visual similarity greater than a set threshold value with respect to the photographed image can be searched in the product database. However, in practice, there is a possibility that a product image having a visual similarity greater than a set threshold value with respect to the photographed image is not searched for in the product database. For this situation, the scheme provided by the embodiment shown in fig. 3 as follows can be adopted.
Fig. 3 is a flowchart of another method for searching for a product according to an embodiment of the present invention, as shown in fig. 3, the method includes the following steps:
301. and acquiring a real shooting image containing the target commodity.
302. And determining the visual similarity between the commodity image in the commodity database and the real image according to the visual characteristics of the real image.
303. If the commodity images with the visual similarity larger than the set threshold value are not determined from the commodity database, determining the commodity images with the highest visual similarity with the photographed images in the preset number from the commodity database as the reference commodity images.
304. And extracting semantic information contained in the photographed image.
305. And searching the commodity database for a target commodity image matched with the reference commodity image and the semantic information.
306. And outputting the commodity information corresponding to the target commodity image.
In this embodiment, in the process of determining the reference commodity image matched with the visual feature of the photographed image in the commodity database, if the visual similarity between each commodity image in the commodity database and the photographed image is lower than the set threshold, it indicates that there is no commodity image in the commodity database that is perfectly matched with the photographed image in terms of visual feature. This may be due to poor quality of the photographed image, or may be that the target product photographed by the user is not actually included in the product database.
In this case, the visual similarity between each commodity image and the photographed image may be sorted from large to small, and a preset number of commodity images having the highest visual similarity with the photographed image may be determined from the commodity database as the reference commodity images. The preset number of commodity images may include only one commodity image with the highest visual similarity to the real image, or may include K commodity images with the visual similarity to the real image ranked at topK, where K is a preset integer greater than 1.
In addition, because the visual similarity between the selected reference commodity image and the photographed image is not high, the commodity search can be assisted by further combining information of other dimensions. In this embodiment, the information of the other dimensions may be: and actually shooting semantic information of the image. Therefore, the semantic features of the real shooting image can be extracted based on the pre-trained semantic recognition model, and the semantic information corresponding to the real shooting image is extracted from the real shooting image. For example, the semantic information may be: semantic information such as brand information, category information, scene information, and the like identified from the photographed image. In practical applications, the outer package of the target product often has characters or figures such as brand marks, names, product names and the like, and the brand and category information can be obtained by identifying and understanding the characters and figures. In addition, for example, if the user sees a certain piece of clothes when shopping mall, and the taken image obtained by taking a picture of the certain piece of clothes in the mall, the taken image may further include related elements of the scene of the mall, such as a show window, a clothes rack, and the like, and these scene elements can be recognized from the taken image so as to know that the user is the target commodity, namely a certain piece of clothes, seen in the scene of the mall.
After the reference commodity image is determined from the commodity database based on the visual features and the semantic information of the photographed image is extracted, the target commodity can be searched in the commodity database by combining the reference commodity image and the semantic information.
Specifically, the target commodity image matching the reference commodity image and the semantic information may be searched in the commodity database, that is, the input search condition for finally performing the target commodity search may include: and reference to the commodity image and semantic information.
Optionally, in the process of searching for the target commodity, a commodity image irrelevant to the semantic information in the reference commodity image may be filtered according to the semantic information, and then the target commodity image matched with the remaining reference commodity image is searched in the commodity database. The target commodity image matched with the remaining reference commodity images may be a commodity image in the commodity database, which meets the requirement of similarity between the visual features of the commodity image and the remaining reference commodity images. When the commodity related information corresponding to each reference commodity image is uploaded to the commodity database, the commodity related information can be associated with related semantic information, such as a brand, a category, a sales location, and the like. Therefore, the reference commodity image irrelevant to the semantic information extracted from the real shooting image can be filtered. Wherein, independently, may be different or dissimilar.
In the examples of categories, it can be considered that the categories belonging to the same class are similar, and the categories belonging to different classes are not similar.
In another optional embodiment, in the process of searching for the target commodity, the commodity images matched with the semantic information may be searched in the commodity database according to the semantic information, then the commodity images unmatched with the reference commodity image are filtered out, and the remaining commodity images are the target commodity images.
In another optional embodiment, if it is found that there is no reference commodity image related to the semantic information after filtering the reference commodity image based on the semantic information, optionally, a commodity image matched with the semantic information may be directly searched in the commodity database as the target commodity image, or another semantic information may be determined according to the semantic information, which is called similar semantic information for distinction, and a commodity image corresponding to the similar semantic information is searched in the commodity database as the target commodity image.
The similar semantic information may be semantic information corresponding to a commodity which can be used in combination with the target commodity. The semantic information extracted from the photographed image is semantic information corresponding to the target product. At this time, the similar semantic information may be determined according to a preset matching relationship between category information, for example, the category information extracted from the live-shot image is: dress, categories that have a match with this category of dress such as: the belt, bag and hat, the similar semantic information is the belt, bag and hat.
For the specific implementation process of some steps of the scheme provided in this embodiment, reference may be made to the related descriptions in the foregoing other embodiments, which are not described herein again.
Based on the scheme provided by the embodiment, the visual characteristics and the semantic information of the real shot image are integrated to search the target commodity, so that the user can obtain a commodity search result more meeting the expectation of the user.
In practical application, the commodity searching scheme provided by the embodiment can be realized by the cooperation of the user equipment and the server. Specifically, the user obtains a live image by starting a photographing function of the user equipment to photograph the target commodity, and the live image is loaded to a search interface of a shopping application program started in the user equipment, so that the shopping application program uploads the live image to the server. The server determines a reference commodity image matched with the visual features of the photographed image in a commodity database, then searches a target commodity image matched with the reference commodity image in the commodity database to obtain commodity information corresponding to the target commodity image, sends the commodity information to the user equipment, and displays the commodity information in a search result interface of a shopping application program.
Optionally, to complete the commodity search process, the shopping application program may be provided with a relevant operation interface, and the user completes relevant operations of commodity search and views the commodity search result through the operation interface. The composition and related operations of the operation interface are exemplarily described with reference to fig. 4.
As shown in fig. 4, when a user wants to search for a certain product, the shopping application described above in the user device is started, and an operation interface 401A is displayed on the user device. In practical applications, the operation interface 401A may be a search interface of a shopping application. An image acquisition button 402A, a search start button 402B, and an output unit 402C can be displayed on the operation interface 401A.
In practical application, the buttons and the components can be arranged in the same interface, or arranged in different interfaces in consideration of convenience of user operation, or some buttons and components can be always arranged on top for display. The display position and the display mode are not strictly limited in this embodiment, and it is emphasized that the shopping application may include these buttons and components in this embodiment. In addition, all or part of the above buttons and components may be included in the operation interface 401A. Of course, other related components may also be included.
When the user clicks the image acquisition button 402A, the shopping application program may be controlled to invoke a photographing function of the user equipment, and the target product is photographed to obtain a real-time photographed image, which may be displayed on the operation interface 401A. In practical application, if the user finds that the quality of the photographed image is not good, corresponding adjustment can be made in time, and the photographed image of the target commodity is collected again.
After the user confirms that the real shooting image is collected, the real shooting image can be sent to the server, and the server determines a reference commodity image matched with the visual characteristics of the real shooting image in the commodity database and feeds the reference commodity image back to the user equipment. The user device may display the reference commodity image on the operation interface 401A for the user to confirm whether or not to adopt.
When the user confirms that the reference commodity image is adopted, after the reference commodity image is selected, the search starting button 402B is clicked, the process that the server searches the target commodity image matched with the reference commodity image in the commodity database and obtains the commodity information corresponding to the target commodity image is triggered, the server feeds the obtained commodity information back to the user equipment, and the user equipment displays the commodity information through the output component 402C in the operation interface 401A.
Hereinafter, an article search apparatus according to one or more embodiments of the present invention will be described in detail. Those skilled in the art will appreciate that these means can each be constructed using commercially available hardware components and by performing the steps taught in this disclosure.
Fig. 5 is a schematic structural diagram of a product searching apparatus according to an embodiment of the present invention, and as shown in fig. 5, the apparatus includes: the device comprises an acquisition module 11, a determination module 12 and a search module 13.
The acquiring module 11 is configured to acquire a real shot image containing a target commodity.
And the determining module 12 is used for determining a reference commodity image matched with the visual characteristics of the real shooting image from the commodity database.
And the searching module 13 is configured to search the commodity database for a target commodity image matched with the reference commodity image, and output commodity information corresponding to the target commodity image.
Optionally, the search module 13 is specifically configured to: outputting the reference commodity image; and if an instruction that the user confirms to use the reference commodity image is received, searching a target commodity image matched with the reference commodity image in the commodity database.
Optionally, the determining module 12 is specifically configured to: determining the visual similarity between the commodity image in the commodity database and the real shot image according to the visual characteristics of the real shot image; and determining the commodity image with the visual similarity meeting set conditions from the commodity database as the reference commodity image.
Optionally, the determining module 12 is specifically configured to: and determining the commodity image with the visual similarity larger than a set threshold value from the commodity database as the reference commodity image.
Optionally, the determining module 12 is specifically configured to: and determining the visual similarity between the reference commodity image and the photographed image from the commodity database to meet a set condition, wherein the corresponding commodity image with the commodity attention meeting the set condition is the reference commodity image.
Optionally, the commodity attention degree corresponding to the commodity image is measured by any one of the following indexes:
the commodity transaction rate corresponding to the commodity image and the click rate corresponding to the commodity image.
Optionally, the determining module 12 is specifically configured to: if the commodity images with the visual similarity larger than the set threshold value are not determined from the commodity database, determining a preset number of commodity images with the highest visual similarity with the photographed images from the commodity database as the reference commodity images.
Optionally, the search module 13 is further configured to: extracting semantic information contained in the real shot image; and searching the commodity database for a target commodity image matched with the reference commodity image and the semantic information.
Optionally, the search module 13 is specifically configured to: filtering out commodity images which are irrelevant to the semantic information in the reference commodity images according to the semantic information; and searching the commodity database for target commodity images matched with the rest reference commodity images.
The apparatus shown in fig. 5 may perform the commodity searching method provided in the foregoing embodiment, and the detailed performing process and technical effect refer to the description in the foregoing embodiment, which are not described herein again.
In one possible design, the structure of the article searching apparatus shown in fig. 5 may be implemented as an electronic device, as shown in fig. 6, which may include: a processor 21 and a memory 22. Wherein the memory 22 has stored thereon executable code which, when executed by the processor 21, makes the processor 21 at least to implement the item search method as provided in the previous embodiments.
The above-described apparatus embodiments are merely illustrative, wherein the units described as separate components may or may not be physically separate. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by adding a necessary general hardware platform, and of course, can also be implemented by a combination of hardware and software. With this understanding in mind, the above-described aspects and portions of the present technology which contribute substantially or in part to the prior art may be embodied in the form of a computer program product, which may be embodied on one or more computer-usable storage media having computer-usable program code embodied therein, including without limitation disk storage, CD-ROM, optical storage, and the like.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for searching for a commodity, comprising:
acquiring a real shot image containing a target commodity;
determining a reference commodity image matched with the visual characteristics of the real shot image from a commodity database;
searching the commodity database for a target commodity image matched with the reference commodity image, and outputting commodity information corresponding to the target commodity image.
2. The method of claim 1, wherein searching the merchandise database for a target merchandise image that matches the reference merchandise image comprises:
outputting the reference commodity image;
and if an instruction that the user confirms to use the reference commodity image is received, searching a target commodity image matched with the reference commodity image in the commodity database.
3. The method of claim 1, wherein determining a reference merchandise image from a merchandise database that matches the visual characteristics of the live image comprises:
determining the visual similarity between the commodity image in the commodity database and the real shot image according to the visual characteristics of the real shot image;
and determining the commodity image with the visual similarity meeting set conditions from the commodity database as the reference commodity image.
4. The method according to claim 3, wherein the determining, from the product database, the product image with the visual similarity meeting the set condition with the live image as the reference product image comprises:
and determining the commodity image with the visual similarity larger than a set threshold value from the commodity database as the reference commodity image.
5. The method according to claim 3 or 4, wherein the determining, from the commodity database, the commodity image with the visual similarity meeting the set condition with the live image as the reference commodity image comprises:
and determining the visual similarity between the reference commodity image and the photographed image from the commodity database to meet a set condition, wherein the corresponding commodity image with the commodity attention meeting the set condition is the reference commodity image.
6. The method according to claim 3, wherein the determining, from the product database, the product image with the visual similarity meeting the set condition with the live image as the reference product image comprises:
if the commodity images with the visual similarity larger than the set threshold value are not determined from the commodity database, determining a preset number of commodity images with the highest visual similarity with the photographed images from the commodity database as the reference commodity images.
7. The method of claim 6, wherein searching the merchandise database for a target merchandise image that matches the reference merchandise image comprises:
extracting semantic information contained in the real shot image;
and searching the commodity database for a target commodity image matched with the reference commodity image and the semantic information.
8. The method of claim 7, wherein searching the merchandise database for a target merchandise image that matches the reference merchandise image and the semantic information comprises:
filtering out commodity images which are irrelevant to the semantic information in the reference commodity images according to the semantic information;
and searching the commodity database for target commodity images matched with the rest reference commodity images.
9. An article search device, comprising:
the acquisition module is used for acquiring a real shot image containing a target commodity;
the determining module is used for determining a reference commodity image matched with the visual characteristics of the real shooting image from a commodity database;
and the searching module is used for searching the target commodity image matched with the reference commodity image in the commodity database and outputting the commodity information corresponding to the target commodity image.
10. An electronic device, comprising: a memory, a processor; wherein the memory has stored thereon executable code which, when executed by the processor, causes the processor to perform the item search method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110846166.5A CN113609319A (en) | 2021-07-26 | 2021-07-26 | Commodity searching method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110846166.5A CN113609319A (en) | 2021-07-26 | 2021-07-26 | Commodity searching method, device and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113609319A true CN113609319A (en) | 2021-11-05 |
Family
ID=78338376
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110846166.5A Pending CN113609319A (en) | 2021-07-26 | 2021-07-26 | Commodity searching method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113609319A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114897576A (en) * | 2022-05-05 | 2022-08-12 | 深圳市极客智能科技有限公司 | Commodity pushing method based on data analysis |
-
2021
- 2021-07-26 CN CN202110846166.5A patent/CN113609319A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114897576A (en) * | 2022-05-05 | 2022-08-12 | 深圳市极客智能科技有限公司 | Commodity pushing method based on data analysis |
| CN114897576B (en) * | 2022-05-05 | 2024-04-19 | 深圳市极客智能科技有限公司 | Commodity pushing method based on data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11682141B2 (en) | Item recommendations based on image feature data | |
| CN107424043B (en) | Product recommendation method and device and electronic equipment | |
| EP3267362B1 (en) | Machine learning image processing | |
| US9336459B2 (en) | Interactive content generation | |
| US9607010B1 (en) | Techniques for shape-based search of content | |
| JP5787011B2 (en) | Search support system, search support method and program | |
| US10963939B1 (en) | Computer vision based style profiles | |
| US20180096069A1 (en) | Interactive clothes searching in online stores | |
| US20220138831A1 (en) | Method of Providing Fashion Item Recommendation Service Using User's Body Type and Purchase History | |
| CN106598998B (en) | Information acquisition method and information acquisition device | |
| US10769524B1 (en) | Non-binary gender filter | |
| KR20190000397A (en) | Fashion preference analysis | |
| CN106055710A (en) | Video-based commodity recommendation method and device | |
| US20190325497A1 (en) | Server apparatus, terminal apparatus, and information processing method | |
| US11568460B2 (en) | Device, method, and program for commercial product reliability evaluation based on image comparison | |
| KR102295459B1 (en) | A method of providing a fashion item recommendation service to a user using a date | |
| US11232511B1 (en) | Computer vision based tracking of item utilization | |
| CN111028029A (en) | Offline commodity recommendation method and device and electronic equipment | |
| CN109074498A (en) | Visitor's tracking and system for the region POS | |
| CN110858375A (en) | Data, display processing method and device, electronic equipment and storage medium | |
| CN113609319A (en) | Commodity searching method, device and equipment | |
| CN106250541A (en) | The method for pushing of a kind of information and device | |
| Lodkaew et al. | Fashion finder: A system for locating online stores on instagram from product images | |
| KR20200084647A (en) | A method of providing a fashion item recommendation service to a user using a date | |
| JP6152188B1 (en) | Information processing apparatus, information processing method, and program |
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 |