Picture searching method, system, equipment and medium

Abstract

The invention provides a picture searching method, system and device and a medium, and relates to the technical field of image processing, and the method comprises the steps: obtaining a plurality of basic pictures and description information corresponding to each basic picture; based on each basic picture and the description information corresponding to each basic picture, generating a tag set corresponding to each basic picture, the tag set comprising a plurality of different types of picture tags; acquiring search information, wherein the search information comprises user demand information; performing natural language processing on the search information to obtain at least one intention tag; and according to the at least one intention tag, matching from the plurality of basic pictures to obtain at least one target picture, wherein a tag set corresponding to the target picture comprises the at least one intention tag. According to the method and the device, multi-source generation of the picture labels is realized, and the picture search efficiency and the search accuracy are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to an image search method, system, device and medium. Background Art

[0002] Existing text-based image search technologies often suffer from low search accuracy and a single label generation method. When designers are looking for inspirational images, it is difficult to quickly and accurately obtain images that meet their needs through simple text descriptions. Therefore, improving image search accuracy and efficiency has become a technical problem that needs to be urgently addressed. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to overcome the deficiencies in the prior art and provide an image search method, system, device, and medium. The present invention provides the following technical solutions: In a first aspect, the present application provides an image search method, the method comprising: obtaining a plurality of basic images and description information corresponding to each of the basic images; generating a tag set corresponding to each of the basic images based on each of the basic images and the description information corresponding to each of the basic images, the tag set including a plurality of different types of image tags; Acquire search information, the search information including: user demand information; perform natural language processing on the search information to obtain at least one intent tag; obtain at least one target image from a plurality of the base images according to the at least one intent tag, the tag set corresponding to the target image including at least one intent tag.

[0004] In one embodiment, the label set includes: a first label subset and a second label subset, and the label set corresponding to each basic picture is generated based on each basic picture and the description information corresponding to each basic picture, including: performing feature recognition on each basic picture to obtain a basic feature set corresponding to each basic picture; determining the first label subset corresponding to each basic picture based on each basic feature set; performing natural language processing on each description information to obtain a keyword set corresponding to each basic picture, the keyword set including at least one keyword; and generating the second label subset corresponding to each basic picture based on the corresponding at least one keyword.

[0005] In one embodiment, the basic feature set includes at least one basic feature, and based on each of the basic feature sets, the first label subset corresponding to each of the basic images is determined respectively, including: traversing a preset label library, and determining the set of each of the image labels corresponding to each of the basic features as the corresponding first label subset of the basic image.

[0006] In one embodiment, after obtaining at least one target image from multiple base images according to at least one of the intent tags, the method further includes: sorting each of the target images according to the number of matching tags, where the number of matching tags is the number of the intent tags included in the tag set.

[0007] In one embodiment, the method further includes: sorting the at least two target images having the same number of matching tags according to the tag weights corresponding to the target images.

[0008] In one embodiment, obtaining the label weight corresponding to the target image includes: if the label set corresponding to the target image includes: at least one of the intention labels, then determining at least one of the intention labels as the matching label of the target image; performing weighted calculation on each of the matching labels to obtain the label weight corresponding to the target image.

[0009] In a second aspect, the present application provides an image search system, the system comprising: A first acquisition module is used to acquire a plurality of basic images and description information corresponding to each of the basic images; a label generation module, configured to generate a label set corresponding to each of the basic images based on each of the basic images and the description information corresponding to each of the basic images, wherein the label set includes a plurality of different types of image labels; The second acquisition module is used to acquire search information, wherein the search information includes: user demand information; a processing module, configured to perform natural language processing on the search information to obtain at least one intent tag; A matching module is configured to obtain at least one target image from the plurality of base images according to at least one of the intent labels, wherein the label set corresponding to the target image includes at least one of the intent labels.

[0010] In one embodiment, the label generation module includes: A feature recognition submodule is used to perform feature recognition on each of the basic images to obtain a basic feature set corresponding to each of the basic images; A first generating submodule, configured to determine, based on each of the basic feature sets, a first label subset corresponding to each of the basic images; A keyword extraction submodule, configured to perform natural language processing on each of the description information to obtain a keyword set corresponding to each of the basic images, wherein the keyword set includes at least one keyword; The second generating submodule is configured to generate a second label subset corresponding to each of the basic images according to the corresponding at least one keyword.

[0011] In a third aspect, the present application provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program runs on the processor, the image search method described in the first aspect is executed.

[0012] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the image search method described in the first aspect is implemented.

[0013] The image search method, system, device and medium provided in the embodiments of the present application obtain multiple basic images and descriptive information corresponding to each of the basic images; generate a tag set corresponding to each of the basic images based on the basic images and the descriptive information corresponding to each of the basic images, and the tag set includes multiple different types of image tags; obtain search information, and the search information includes: user demand information; perform natural language processing on the search information to obtain at least one intent tag; match at least one target image from the multiple basic images according to at least one intent tag, and the tag set corresponding to the target image includes at least one intent tag, thereby realizing multi-source generation of image tags and effectively improving image search efficiency and search accuracy.

[0014] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 A schematic diagram of a process of an image search method provided by an embodiment of the present application is shown; Figure 2 An example diagram of a basic image provided by an embodiment of the present application is shown; Figure 3 Another schematic diagram of the process of searching for an image provided by an embodiment of the present application is shown; Figure 4 A schematic diagram of the structure of the image search system provided by an embodiment of the present application is shown; Figure 5 A structural schematic diagram of an electronic device provided in an embodiment of the present application is shown.

[0017] Description of main component symbols: 400 - image search system; 410 - first acquisition module; 420 - label generation module; 430 - second acquisition module; 440 - processing module; 450 - matching module; 500 - electronic device; 501 - transceiver; 502 - processor; 503 - memory. DETAILED DESCRIPTION

[0018] The following describes embodiments of the present invention in detail. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only to explain the present invention and are not to be construed as limiting the present invention.

[0019] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature identified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used in the template description herein are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0021] Example 1 The existing image tag generation method is single. When designers are looking for inspiration images, it is difficult to quickly and accurately obtain images that meet their needs through simple text descriptions. Figure 1 , an embodiment of the present application provides an image search method, including: steps S110~S150.

[0022] Step S110: Acquire multiple basic pictures and description information corresponding to each basic picture.

[0023] In this embodiment, multiple basic images and descriptive information associated with each basic image are obtained, wherein the descriptive information includes a text description corresponding to the basic image, metadata attributes of the image (such as shooting time, scene description, etc.), or other text content related to the image content, purpose, and creation background. Its role is to provide original data support for the subsequent generation of a label set to ensure that the label can accurately reflect the characteristics of the image.

[0024] Step S120 : Based on each of the basic images and the description information corresponding to each of the basic images, a tag set corresponding to each of the basic images is generated, where the tag set includes a plurality of different types of image tags.

[0025] In this embodiment, after obtaining the basic image and the descriptive information corresponding to the basic image, the background processing system and the AI ​​visual big model collaborative working mechanism are used to analyze each basic image and its descriptive information in different dimensions.

[0026] On the one hand, the background processing system generates image tags based on the description information of the basic image, such as author information, function information and other non-visual attribute information; on the other hand, the visual model combines the preset tag library to identify and analyze the visual features of the basic image itself, such as the elements, colors, composition, style, etc. in the basic image, and thus generates image tags for each basic image. Figure 2 The picture shown may generate multiple different image labels such as "lipstick", "luxury", "gold rim", and "female". The image label set corresponding to the basic picture can comprehensively and accurately describe the attributes and characteristics of the basic picture.

[0027] In one embodiment, the tag set includes: a first tag subset and a second tag subset, see Figure 3 , step S120 includes: steps S121~S124.

[0028] In step S121 , feature recognition is performed on each of the basic images to obtain a basic feature set corresponding to each of the basic images.

[0029] In this embodiment, visual analysis is performed on each base image, and image recognition technology is used to extract various visual elements within the base image. Specifically, this involves detecting and identifying features such as the shape, color, texture, composition, and style of objects in the base image, such as product appearance, packaging design style, and color matching. Ultimately, a basic feature set corresponding to each base image is formed.

[0030] Step S122: Determine the first label subset corresponding to each basic image according to each basic feature set.

[0031] The basic feature set is matched and mapped to the preset label library. Based on each basic feature in the basic feature set, the corresponding image labels are filtered from the preset label library to generate the first label subset. For example, if the basic feature set includes the following features: lipstick packaging has a gold rim and an embossed design, then the first label subset may include image labels directly related to the visual features, such as "luxury," "gold rim," and "embossed."

[0032] In one embodiment, the basic feature set includes at least one basic feature, and based on each of the basic feature sets, the first label subset corresponding to each of the basic images is determined respectively, including: traversing a preset label library, and determining the set of each of the image labels corresponding to each of the basic features as the corresponding first label subset of the basic image.

[0033] It can be understood that by traversing the preset label library, the image labels corresponding to each basic feature in the basic feature set of each basic image in the label library are extracted, and the set of these extracted image labels is determined to be the first label subset of the basic image. For example, if the basic feature set of a basic image includes basic features such as "lipstick", "gold rim", and "retro style", when traversing the label library, the "cosmetics" label corresponding to "lipstick", the "luxury feeling" and "material・gold rim" labels corresponding to "gold rim", and the "packaging design style・retro style" label corresponding to "retro style" will be found. These image labels are combined to form the first label subset of the basic image. In this way, by mapping the basic features with the label library, a label set that accurately describes the visual attributes of the image is generated.

[0034] Step S123 : performing natural language processing on each of the description information to obtain a keyword set corresponding to each of the basic images, wherein the keyword set includes at least one keyword.

[0035] In this embodiment, the description information of each basic image is parsed using natural language processing (NLP). Specifically, through operations such as word segmentation, part-of-speech tagging, and semantic understanding, keywords that reflect the core content of the description information are extracted to form a set containing at least one keyword. For example, if the description information is "New Christmas limited edition bright red lipstick, metallic packaging", after processing, keywords such as "Christmas limited edition," "bright red," and "metallic texture" may be obtained.

[0036] Step S124: Generate the second label subset corresponding to each of the basic images according to the corresponding at least one keyword.

[0037] In this embodiment, keywords are converted into tags that conform to the tag library specifications, thereby generating a second tag subset. This second tag subset primarily includes semantic tags extracted from the descriptive information. For example, the keyword "Christmas limited edition" can be mapped to "holiday limited edition" and "Christmas" in the tag library, "red" can be mapped to "red series", and "metallic texture" can be mapped to "main packaging color - metallic color". This complements the non-visual attribute image tags of the base image, making the tag set more comprehensive in covering the base image information.

[0038] Step S130: Acquire search information, where the search information includes user demand information.

[0039] In this embodiment, the search information input by the user is obtained through the user interaction interface, such as "looking for macaron-colored skin care product packaging design suitable for summer" or "needing luxurious lipstick pictures for dinner party scenes". Step S140: Perform natural language processing on the search information to obtain at least one intent tag.

[0040] In this embodiment, natural language processing technology is used to convert search information into structured intent tags. For example, if the search information is input as "summer refreshing sunscreen packaging design", after processing, the intent tags such as "summer", "refreshing", "sunscreen", and "packaging design" are obtained. Step S150 : obtaining at least one target image from the plurality of base images according to at least one of the intent labels, wherein the label set corresponding to the target image includes at least one of the intent labels.

[0041] As you can understand, each intent tag is used as a search criterion to perform a matching search within a database of multiple base images and their tag sets. The specific logic is: traverse the tag set of each base image. If an image's tag set contains at least one tag that matches the user's intended tag, then that image is identified as the target image. For example, if the user's intended tags include "summer" and "sunscreen," all base images with the tags containing "summer" and "sunscreen" will be matched. The final set of target images returned will be sorted by the number of matching intent tags, ensuring that users prioritize the images that best meet their needs.

[0042] In one embodiment, after obtaining at least one target image from multiple base images according to at least one of the intent tags, the method further includes: sorting each of the target images according to the number of matching tags, where the number of matching tags is the number of the intent tags included in the tag set.

[0043] For example, target images with a large number of matching tags are ranked in front, and those with a small number are ranked in the back, and finally the sorted target images are returned to the user in this order, so that the user can prioritize the images that match the search intent most highly, thereby improving the accuracy of search results and user experience.

[0044] In one embodiment, the method further includes: sorting the at least two target images having the same number of matching tags according to the tag weights corresponding to the target images.

[0045] If there are at least two target images with the same number of matching tags, a secondary sorting is performed based on the tag weights corresponding to the two target images. This ensures that, under the premise of the same number of matching tags, the degree of fit between the target images and the user's search intent can be accurately distinguished, avoiding sorting ambiguity and further improving the rationality of the search result priority.

[0046] In one embodiment, obtaining the label weight corresponding to the target image includes: if the label set corresponding to the target image includes: at least one of the intention labels, then determining at least one of the intention labels as the matching label of the target image; performing weighted calculation on each of the matching labels to obtain the label weight corresponding to the target image.

[0047] In this embodiment, a determination is first made as to whether the target image's tag set contains at least one intent tag. If so, these intent tags are identified as matching tags. Next, a weighted calculation is performed on each matching tag. Different types of tags (such as product, style, and scene) are assigned different weights based on their importance in the search intent (for example, core product tags are weighted higher than auxiliary style tags). This weighted calculation ultimately yields a tag weight for the target image. This weight serves as the key basis for sorting when the number of matches is the same, ensuring that images with tags more closely aligned with the user's intent are ranked higher.

[0048] The image search method provided in the embodiment of the present application obtains multiple basic images and descriptive information corresponding to each of the basic images; generates a tag set corresponding to each of the basic images based on the basic images and the descriptive information corresponding to each of the basic images, and the tag set includes multiple different types of image tags; obtains search information, and the search information includes: user demand information; performs natural language processing on the search information to obtain at least one intent tag; and obtains at least one target image from the multiple basic images based on at least one intent tag, and the tag set corresponding to the target image includes at least one intent tag, thereby realizing multi-source generation of image tags and effectively improving image search efficiency and search accuracy.

[0049] Example 2 Also, see Figure 4 , the present application also provides an image search system 400, comprising: A first acquisition module 410 is configured to acquire a plurality of basic images and description information corresponding to each of the basic images; a label generation module 420 for generating a label set corresponding to each basic image based on each basic image and the description information corresponding to each basic image, wherein the label set includes multiple different types of image labels; The second acquisition module 430 is used to acquire search information, wherein the search information includes: user demand information; A processing module 440 is configured to perform natural language processing on the search information to obtain at least one intent tag; The matching module 450 is configured to obtain at least one target image from the plurality of base images according to at least one of the intent labels, wherein the label set corresponding to the target image includes at least one of the intent labels.

[0050] The label generation module includes: A feature recognition submodule is used to perform feature recognition on each of the basic images to obtain a basic feature set corresponding to each of the basic images; A first generating submodule, configured to determine, based on each of the basic feature sets, a first label subset corresponding to each of the basic images; A keyword extraction submodule, configured to perform natural language processing on each of the description information to obtain a keyword set corresponding to each of the basic images, wherein the keyword set includes at least one keyword; The second generating submodule is configured to generate a second label subset corresponding to each of the basic images according to the corresponding at least one keyword.

[0051] The image search system 400 provided in the embodiment of the present application can execute the image search method provided in the above-mentioned method embodiment 1, and will not be described again here to avoid repetition.

[0052] The image search system provided by the embodiment of the present application obtains multiple basic images and descriptive information corresponding to each of the basic images through a first acquisition module; the label generation module generates a label set corresponding to each of the basic images based on the basic images and the descriptive information corresponding to each of the basic images, and the label set includes multiple different types of image labels; the second acquisition module obtains search information, and the search information includes: user demand information; the processing module performs natural language processing on the search information to obtain at least one intent label; the matching module matches at least one target image from the multiple basic images according to at least one intent label, and the label set corresponding to the target image includes at least one intent label, thereby realizing multi-source generation of image labels and effectively improving image search efficiency and search accuracy.

[0053] Example 3 In addition, an embodiment of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program runs on the processor, the image search method provided in Example 1 is executed.

[0054] For details, see Figure 5 The electronic device 500 includes: a transceiver 501, a bus interface and a processor 502, wherein the processor 502 is used to obtain multiple basic pictures and description information corresponding to each of the basic pictures; based on each of the basic pictures and the description information corresponding to each of the basic pictures, generate a tag set corresponding to each of the basic pictures, wherein the tag set includes multiple different types of picture tags; obtain search information, wherein the search information includes: user demand information; perform natural language processing on the search information to obtain at least one intent tag; match at least one target picture from the multiple basic pictures according to at least one intent tag, wherein the tag set corresponding to the target picture includes at least one intent tag.

[0055] In one embodiment, the label set includes: a first label subset and a second label subset, and the label set corresponding to each basic picture is generated based on each basic picture and the description information corresponding to each basic picture, including: performing feature recognition on each basic picture to obtain a basic feature set corresponding to each basic picture; determining the first label subset corresponding to each basic picture based on each basic feature set; performing natural language processing on each description information to obtain a keyword set corresponding to each basic picture, the keyword set including at least one keyword; and generating the second label subset corresponding to each basic picture based on the corresponding at least one keyword.

[0056] In one embodiment, the basic feature set includes at least one basic feature, and based on each of the basic feature sets, the first label subset corresponding to each of the basic images is determined respectively, including: traversing a preset label library, and determining the set of each of the image labels corresponding to each of the basic features as the corresponding first label subset of the basic image.

[0057] In one embodiment, after obtaining at least one target image from multiple base images according to at least one of the intent tags, the method further includes: sorting each of the target images according to the number of matching tags, where the number of matching tags is the number of the intent tags included in the tag set.

[0058] In one embodiment, the method further includes: sorting the at least two target images having the same number of matching tags according to the tag weights corresponding to the target images.

[0059] In one embodiment, obtaining the label weight corresponding to the target image includes: if the label set corresponding to the target image includes: at least one of the intention labels, then determining at least one of the intention labels as the matching label of the target image; performing weighted calculation on each of the matching labels to obtain the label weight corresponding to the target image.

[0060] In the embodiment of the present invention, the electronic device 500 further includes a memory 503. Figure 5 In the embodiment, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits such as one or more processors represented by processor 502 and memory represented by memory 503. The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be further described herein. The bus interface provides an interface. The transceiver 501 can be multiple components, that is, including a transmitter and a receiver, providing a unit for communicating with various other devices over a transmission medium. The processor 502 is responsible for managing the bus architecture and general processing, and the memory 503 can store data used by the processor 502 when performing operations.

[0061] The electronic device 500 provided in the embodiment of the present invention can execute the image search method provided in the above method embodiment 1, which will not be described again here to avoid repetition.

[0062] Example 4 In addition, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the image search method provided in Example 1 is implemented.

[0063] In this embodiment, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0064] The computer-readable storage medium provided in this embodiment can implement the image search method provided in Example 1, and will not be described again here to avoid repetition.

[0065] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not limiting, and thus other examples of the exemplary embodiments may have different values.

[0066] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0067] The above-described embodiments merely illustrate several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that variations and modifications are possible without departing from the scope of the present invention, and such variations and modifications are fully within the scope of protection of the present invention.

Claims

1. A picture search method, characterized in that: The method comprises: Obtaining multiple basic images and description information corresponding to each of the basic images; Based on each of the basic images and the description information corresponding to each of the basic images, generating a label set corresponding to each of the basic images, wherein the label set includes multiple different types of image labels; Acquiring search information, wherein the search information includes: user demand information; Performing natural language processing on the search information to obtain at least one intent tag; At least one target image is obtained by matching the plurality of base images according to at least one of the intent labels, and the label set corresponding to the target image includes at least one of the intent labels.

2. The image search method according to claim 1, wherein: The tag set includes: a first tag subset and a second tag subset, and generating the tag set corresponding to each basic image based on each basic image and the description information corresponding to each basic image includes: Perform feature recognition on each of the basic images to obtain basic feature sets corresponding to each of the basic images; Determining, according to each of the basic feature sets, the first label subset corresponding to each of the basic images; Performing natural language processing on each of the description information to obtain a keyword set corresponding to each of the basic images, wherein the keyword set includes at least one keyword; According to the corresponding at least one keyword, the second tag subset corresponding to each of the basic images is generated respectively.

3. The image search method according to claim 2, wherein: The basic feature set includes at least one basic feature, and determining a first label subset corresponding to each basic image based on each basic feature set includes: The preset label library is traversed, and the sets of the image labels corresponding to the basic features are determined as the first label subset of the corresponding basic image.

4. The image search method according to claim 1, wherein: After obtaining at least one target image from the plurality of base images according to at least one intent tag, the method further includes: The target images are sorted according to the number of matching tags, where the number of matching tags is the number of the intent tags included in the tag set.

5. The image search method according to claim 4, characterized in that: The method further comprises: For at least two target images with the same number of matching tags, the target images are sorted according to the tag weights corresponding to the target images.

6. The image search method according to claim 5, characterized in that: Obtaining the label weight corresponding to the target image includes: If the tag set corresponding to the target image includes: at least one of the intention tags, determining the at least one intention tag as a matching tag for the target image; A weighted calculation is performed on each of the matching tags to obtain the tag weight corresponding to the target image.

7. An image search system, characterized in that: The system comprises: A first acquisition module is used to acquire a plurality of basic images and description information corresponding to each of the basic images; a label generation module, configured to generate a label set corresponding to each of the basic images based on each of the basic images and the description information corresponding to each of the basic images, wherein the label set includes a plurality of different types of image labels; The second acquisition module is used to acquire search information, wherein the search information includes: user demand information; a processing module, configured to perform natural language processing on the search information to obtain at least one intent tag; A matching module is configured to obtain at least one target image from the plurality of base images according to at least one of the intent labels, wherein the label set corresponding to the target image includes at least one of the intent labels.

8. The image search system according to claim 7, characterized in that: The label generation module includes: A feature recognition submodule is used to perform feature recognition on each of the basic images to obtain a basic feature set corresponding to each of the basic images; A first generating submodule, configured to determine, based on each of the basic feature sets, a first label subset corresponding to each of the basic images; A keyword extraction submodule, configured to perform natural language processing on each of the description information to obtain a keyword set corresponding to each of the basic images, wherein the keyword set includes at least one keyword; The second generating submodule is configured to generate a second label subset corresponding to each of the basic images according to the corresponding at least one keyword.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is run on the processor, the image search method according to any one of claims 1 to 6 is executed.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the image search method according to any one of claims 1 to 6 is implemented.

Images