Marketing picture generation method, system and device, medium and program product
By automatically filtering and cropping marketing images and using image processing technology to generate images that meet the requirements for placement, the problem of high labor costs and low efficiency in existing technologies is solved, and efficient and automated marketing image generation and placement are achieved.
Patent Information
- Application Number
- CN202410995345.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-15
AI Technical Summary
Current methods for acquiring marketing images require manual processing and screening, resulting in high costs, low efficiency, and difficulty in meeting the needs of large-scale deployment.
By acquiring product information of the products to be marketed, the system automatically filters and crops target product images, uses image processing technologies such as ControlNet and deep learning algorithms to generate marketing images that meet the requirements for ad placement, and optimizes content delivery and data collection.
It enables the automated generation of marketing images, improving production efficiency, reducing labor costs, ensuring the uniqueness and authenticity of images, and meeting online marketing needs.
Smart Images

Figure CN118967223A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of image processing technology, and in particular to a method, system, device, medium and program product for generating marketing images. Background Art
[0002] In the tourism scenario, a large amount of marketing materials are needed to introduce different tourist destinations and attractions to users. Posting pictures of destinations and attractions in online communities (such as company communities, Xiaohongshu, etc.) and displaying exquisite pictures directly to users can allow users to intuitively feel the tourist destinations and the corresponding scenery, thereby attracting users to learn more. By posting and hanging goods in online communities, it is possible to attract traffic to online products, thereby achieving the purpose of marketing and promoting tourism products. However, the source and manpower of picture materials are limited, and there are certain limitations to mass production. Currently, the more common means of obtaining pictures are to purchase copyrighted pictures and obtain pictures through crawlers. These methods have the following problems: (1) Copyrighted pictures are limited, and when a large number of marketing materials are needed, there will be a high purchase fee to pay; (2) Pictures obtained through crawlers may have infringement issues; (3) The source pictures obtained cannot directly meet the delivery needs. For example, when marketing online tourism products, it is difficult to ensure the relevance of pictures and products and the compliance of picture sizes with regulations. These require manual processing and screening, so when faced with the need for large-scale delivery, the corresponding labor costs will be very high. Summary of the invention
[0003] The technical problem to be solved by the present disclosure is to overcome the defects of the existing method of acquiring marketing pictures, which requires manual processing and screening, resulting in high cost and low acquisition efficiency, and to provide a method, system, device, medium and program product for generating marketing pictures.
[0004] The present invention solves the above technical problems through the following technical solutions:
[0005] A first aspect of the present disclosure provides a method for generating a marketing picture, the method comprising:
[0006] Obtain product information of products to be marketed;
[0007] Filter out target product images according to the product information;
[0008] The target product image is cropped to obtain a marketing image of the product to be marketed.
[0009] Preferably, before the step of obtaining the commodity information of the commodity to be marketed, the generation method further includes:
[0010] Obtain the product identification number of the product to be marketed;
[0011] The step of obtaining the commodity information of the commodity to be marketed comprises:
[0012] The commodity information of the commodity to be marketed is obtained according to the commodity identification number.
[0013] Preferably, after the step of filtering out target product images according to the product information, the generation method further comprises:
[0014] Performing image-to-image processing on the target product image to obtain the target product image after image-to-image processing;
[0015] The step of cropping the target product image to obtain the marketing image of the product to be marketed includes:
[0016] The target product image after the raw image processing is cropped to obtain a marketing image of the product to be marketed.
[0017] Preferably, after the step of cropping the target product image to obtain the marketing image of the product to be marketed, the generation method further comprises:
[0018] Using the marketing pictures for content delivery;
[0019] Obtaining delivery data of the marketing image;
[0020] Obtaining a target marketing image according to the delivery data;
[0021] The target marketing picture is stored in a product picture library.
[0022] Preferably, the step of filtering out target product images according to the product information includes:
[0023] The product images are quality evaluated using a picture quality model according to the product information to select the target product images.
[0024] Preferably, the step of cropping the target product image to obtain the marketing image of the product to be marketed includes:
[0025] Get the preselected cropping frame;
[0026] Inputting the target product image and the pre-selected cropping frame into a target detection algorithm to obtain a target area in the target product image;
[0027] Input the target product image into a convolutional network to obtain a feature map corresponding to the target product image;
[0028] Performing pooling processing on the feature map to obtain regional features corresponding to the pre-selected cropping frame area and the target area;
[0029] Using the regional features as network nodes of an adaptive attention graph convolutional network to obtain visual semantic information and spatial composition relationships of different image nodes in the target product image;
[0030] Sorting the pre-selected cropping boxes based on the visual semantic information and the spatial composition relationship to obtain a target pre-selected cropping box;
[0031] The target product image is cropped based on the target pre-selected cropping frame to obtain a marketing image of the product to be marketed.
[0032] A second aspect of the present disclosure provides a system for generating marketing pictures, the system comprising:
[0033] A first acquisition module is used to acquire product information of the product to be marketed;
[0034] A screening module, used to screen out target product images according to the product information;
[0035] The cropping module is used to crop the target product image to obtain a marketing image of the product to be marketed.
[0036] Preferably, the generation system further comprises:
[0037] The second acquisition module is used to obtain the commodity identification number of the commodity to be marketed;
[0038] The first acquisition module is used to acquire the commodity information of the commodity to be marketed according to the commodity identification number.
[0039] Preferably, the generation system further comprises:
[0040] A raw image processing module, used to perform raw image processing on the target product image to obtain the target product image after raw image processing;
[0041] The cropping module is used to crop the target product image after the image processing to obtain a marketing image of the product to be marketed.
[0042] Preferably, the generation system further comprises:
[0043] A content delivery module, used for delivering the marketing image to the content;
[0044] A third acquisition module is used to obtain the delivery data of the marketing image;
[0045] A fourth acquisition module, used to acquire a target marketing image according to the delivery data;
[0046] The storage module is used to store the target marketing picture in a product gallery.
[0047] Preferably, the screening module is used to perform quality assessment on product images through a picture quality model according to the product information, so as to screen out the target product images.
[0048] Preferably, the cutting module comprises:
[0049] A first acquisition unit, used to acquire a pre-selected cropping frame;
[0050] A second acquisition unit is used to input the target product image and the pre-selected cropping frame into a target detection algorithm to obtain a target area in the target product image;
[0051] A third acquisition unit is used to input the target product image into a convolutional network to obtain a feature map corresponding to the target product image;
[0052] A pooling processing unit, used to perform pooling processing on the feature map to obtain regional features corresponding to the pre-selected cropping frame area and the target area;
[0053] A fourth acquisition unit is used to use the regional features as network nodes of an adaptive attention graph convolutional network to obtain visual semantic information and spatial composition relationships of different image nodes in the target product image;
[0054] A sorting unit, configured to sort the pre-selected cropping boxes based on the visual semantic information and the spatial composition relationship to obtain a target pre-selected cropping box;
[0055] A cropping unit is used to crop the target product image based on the target pre-selected cropping frame to obtain a marketing image of the product to be marketed.
[0056] A third aspect of the present disclosure provides an electronic device, including a memory, a processor, and a computer program stored in the memory and used to run on the processor, wherein the processor implements the method for generating marketing pictures described in the first aspect when executing the computer program.
[0057] A fourth aspect of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for generating a marketing picture described in the first aspect is implemented.
[0058] A fifth aspect of the present disclosure provides a computer program product, including a computer program, which, when executed by a processor, implements the method for generating a marketing picture as described in the first aspect.
[0059] On the basis of being in accordance with the common sense in the art, the above-mentioned preferred conditions can be arbitrarily combined to obtain the preferred embodiments of the present disclosure.
[0060] The positive and progressive effects of this disclosure are:
[0061] The present invention automatically screens out target product images according to product information, crops the target product images, and obtains marketing images of the products to be marketed; and puts the marketing images in the community to attract traffic to online travel products to achieve the marketing purpose, thereby realizing automatic generation of marketing images, improving production efficiency, and reducing labor costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0062] Figure 1 A flowchart of the method for generating marketing pictures provided in Example 1 of the present disclosure.
[0063] Figure 2 A module diagram of a system for generating marketing images provided in Example 2 of the present disclosure.
[0064] Figure 3 This is a schematic diagram of the structure of an electronic device that implements the method for generating marketing pictures according to Embodiment 3 of the present disclosure. DETAILED DESCRIPTION
[0065] The present disclosure is further described below by way of examples, but the present disclosure is not limited to the scope of the examples.
[0066] Prefixes such as "first" and "second" are used in the embodiments of the present disclosure only to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. The use of prefixes such as ordinal numbers to distinguish description objects in the embodiments of the present disclosure does not constitute a limitation on the described objects. For the statement of the described objects, please refer to the description in the context of the claims or embodiments, and no unnecessary limitation should be constituted due to the use of such prefixes. In addition, in the description of the present embodiment, unless otherwise specified, the meaning of "plurality" is two or more.
[0067] In the embodiments of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.
[0068] Example 1
[0069] Figure 1 A flowchart of a method for generating a marketing picture provided in Example 1 of the present disclosure, such as Figure 1 As shown, the generation method includes:
[0070] S1. Obtain product information of the product to be marketed;
[0071] In this embodiment, product information includes but is not limited to product pictures, review pictures, product destinations, etc.;
[0072] S2. Filter out target product images based on product information;
[0073] S3. Crop the target product image to obtain a marketing image of the product to be marketed.
[0074] This embodiment automatically filters out target product images based on product information, crops the target product images, and obtains marketing images of the products to be marketed; and places the marketing images in the community to attract traffic to online travel products to achieve the marketing purpose, thereby realizing automatic generation of marketing images, improving production efficiency, and reducing labor costs.
[0075] In an optional embodiment, before step S1, the generating method further includes:
[0076] S0. Obtain the product identification number of the product to be marketed;
[0077] S1 includes:
[0078] S11. Obtain product information of the product to be marketed according to the product identification number.
[0079] In the specific implementation process, weighted considerations are made from aspects such as current hot spots and time, product labels, diversity, and product quality to select products that need to be marketed and obtain their product IDs (identification numbers); product information is obtained based on the product IDs; further, each online travel product maintains a corresponding product image and review library, and after scoring and screening the image clarity, composition, aesthetics, and other aspects, high-quality target product images are obtained for image processing.
[0080] In an optional embodiment, after step S2, the generating method further includes:
[0081] S21, performing image-to-image processing on the target product image to obtain the target product image after image-to-image processing;
[0082] Step S3 includes:
[0083] S31. Crop the target product image after the image processing to obtain a marketing image of the product to be marketed.
[0084] In this embodiment, online delivery requires that the pictures are not repeated and that the pictures are similar to the real scenic spots. This embodiment uses Controlnet to generate pictures to meet the above two points. First, each picture generated by the picture generation is different. In addition, the authenticity can be guaranteed by controlling the changes based on the selected real pictures using Controlnet.
[0085] Specifically, ControlNet locks the model parameters of the original neural network block (a model parameter) and copies a trainable copy, and finally connects them together using zero convolution. The locked part of the weights remains unchanged, so that it maintains the original image generation capability of stable diffusion. The other part is a trainable copy, which is used to learn to add conditions. The condition here is to learn the contour of the original image so that when it is generated, there is no obvious difference between the contour of the generated image and the original image, that is, to maintain its authenticity.
[0086] In an optional embodiment, after step S3, the generating method further includes:
[0087] S4. Place marketing images for content delivery;
[0088] S5. Obtaining the delivery data of the marketing pictures;
[0089] S6. Obtain target marketing pictures based on delivery data;
[0090] S7. Store the target marketing image in the product gallery.
[0091] In this embodiment, the online community where the produced and cropped marketing images are placed is used, and the online placement effect is recovered, and the target marketing images with good effects are maintained in the product gallery. Specifically, after manual review by the operator, the image quality and article content are guaranteed to meet the requirements before being placed, and the corresponding marketing product’s business card will be attached when the image is placed. Effect recovery: data such as online exposure, clicks, and transaction amounts of the article are recovered; iterative optimization: based on the online effect data, feedback is given to the marketing effect, and images with good marketing effects are maintained.
[0092] In an optional embodiment, step S2 includes:
[0093] The product images are evaluated for quality using an image quality model based on the product information to select target product images.
[0094] In this embodiment, the product images are evaluated for quality through the image quality model based on the product information, and product images with better quality (i.e., target product images) are screened out. Then, through the AIGC (Artificial Intelligence Generative Content) method, image changes are performed based on the screened high-quality product images to prevent repeated use of images.
[0095] In an optional embodiment, step S3 includes:
[0096] S3-1, obtaining a pre-selected cropping frame;
[0097] S3-2. Input the target product image and the preselected cropping frame into the target detection algorithm to obtain the target area in the target product image;
[0098] S3-3. Input the target product image into the convolutional network to obtain the feature map corresponding to the target product image;
[0099] S3-4. Perform pooling processing on the feature map to obtain the regional features corresponding to the preselected cropping frame area and the target area;
[0100] S3-5. Use the regional features as the network nodes of the adaptive attention map convolutional network to obtain the visual semantic information and spatial composition relationship of different image nodes in the target product image;
[0101] S3-6. Sort the preselected cropping frames based on the visual semantic information and spatial composition relationship to obtain the target preselected cropping frame;
[0102] S3-7. Crop the target product image based on the target preselected cropping frame to obtain the marketing image of the product to be marketed.
[0103] In this embodiment, the target product image is cropped by the cropping model to meet the size requirements of the target placement position; specifically, since the placement image has specific size requirements and the image generation from image cannot customize the generated image size, it is necessary to implement the cropping of the generated image through the cropping algorithm. And the cropping algorithm should尽可能保留图片信息以及考虑到美观性。Given that the length and height of an image to be cropped are w and h respectively, it is necessary to crop it into the target size with length and height of x and y respectively.
[0104] 1. Generation of preselected cropping frames: In order to尽可能保证保留的图片能有最大的信息量,therefore, the cropping is based on the situation of fixed size to retain the maximum amount of information.
[0105] If h < w * y / x: Then, with the step length x_step being w / 30, fix the height unchanged from left to right and sample the length.
[0106] If h > w * y / x: Then, with the step length y_step being h / 30, fix the length unchanged from top to bottom and sample the height.
[0107] Up to 30 preselected cropping frames are generated thus far.
[0108] 2. Input the corresponding target product image and the preselected cropping frame into YoloV6 (target detection algorithm) to detect the target area in the target product image.
[0109] 3. The target product image is passed into the convolutional network to obtain the corresponding feature map, and the pooling operation is implemented through RoIAlign (a regional feature aggregation method) and RoDAlign (a regional feature discarding method) to obtain the regional features of the cropping box area and the detected target area.
[0110] 4. The regional features corresponding to the cropping frame area and the detected target area are used as nodes of the AAG (Adaptive Attention Graph) network to obtain the visual semantic information and spatial composition relationship of different image nodes in the image. Finally, several pre-selected cropping frames are aesthetically sorted, and the cropping frame with the highest score is output as the target pre-selected cropping frame selected by the model; the target product image is cropped based on the target pre-selected cropping frame to obtain the marketing image of the product to be marketed.
[0111] This embodiment uses AIGC and deep learning related technologies to achieve the generation of marketing pictures for products. By placing marketing pictures and texts in the community, online travel products are directed to achieve the purpose of marketing. The marketing pictures generated by this embodiment can ensure uniqueness and authenticity. In addition, the picture production process proposed in this embodiment is highly automated, which not only improves the efficiency of generation, but also reduces labor costs. In terms of efficiency, the community article production capacity can reach 1,000 articles / week, and in terms of quality, the overall excellence rate of articles reaches more than 90%.
[0112] Furthermore, this embodiment realizes the generation of original materials for marketing pictures by utilizing related technologies such as controllable generation of stable diffusion and intelligent cropping, solves the problem of the source and processing of picture materials when conducting marketing in tourism scenarios, and also solves the problem of exhaustion of marketing materials. In addition, the cropping algorithm realizes the automatic processing of original materials to meet online requirements, greatly improving the production efficiency of materials and reducing manpower consumption; finally, by recycling the online delivery effect, the subsequent picture generation is guided; a highly automated and efficient picture production and processing process is realized. Through this technology, product information can be imported in batches, and then the marketing pictures of the corresponding products can be obtained in batches for online promotion and delivery and finally the effect is recycled. In addition, the generation method of marketing pictures proposed in this embodiment can ensure the strong correlation between pictures and product content; ensure the diversity of pictures when the same product is produced multiple times; and improve the authenticity of marketing pictures.
[0113] Example 2
[0114] Corresponding to the aforementioned embodiment of a method for generating a marketing picture, the present disclosure also provides an embodiment of a system for generating a marketing picture.
[0115] Figure 2A schematic diagram of a module of a marketing picture generation system provided in Embodiment 2 of the present disclosure, such as Figure 2 As shown, the generation system includes: a first acquisition module 21, a screening module 22, and a cutting module 23;
[0116] A first acquisition module 21 is used to acquire product information of the product to be marketed;
[0117] In this embodiment, product information includes but is not limited to product pictures, review pictures, product destinations, etc.;
[0118] A screening module 22 is used to screen out target product images according to product information;
[0119] The cropping module 23 is used to crop the target product image to obtain a marketing image of the product to be marketed.
[0120] This embodiment automatically filters out target product images based on product information, crops the target product images, and obtains marketing images of the products to be marketed; and places the marketing images in the community to attract traffic to online travel products to achieve the marketing purpose, thereby realizing automatic generation of marketing images, improving production efficiency, and reducing labor costs.
[0121] In an optional embodiment, the generating system further includes:
[0122] The second acquisition module is used to obtain the commodity identification number of the commodity to be marketed;
[0123] The first acquisition module is used to acquire the commodity information of the commodity to be marketed according to the commodity identification number.
[0124] In the specific implementation process, weighted considerations are made from aspects such as current hot spots and time, product labels, diversity, and product quality to select products that need to be marketed and obtain their product IDs (identification numbers); product information is obtained based on the product IDs; further, each online travel product maintains a corresponding product image and review library, and after scoring and screening the image clarity, composition, aesthetics, and other aspects, high-quality target product images are obtained for image processing.
[0125] In an optional embodiment, the generating system further includes:
[0126] A raw image processing module is used to perform raw image processing on the target product image to obtain the target product image after raw image processing;
[0127] The cropping module is used to crop the target product image after image processing to obtain a marketing image of the product to be marketed.
[0128] In this embodiment, online delivery requires that the images are not repeated and that the images are similar to the real scenic spots. This embodiment uses Controlnet to generate images to meet the above two points. First, each image generated by the image generation is different. In addition, the authenticity can be guaranteed by controlling the changes based on the selected real images using Controlnet.
[0129] Specifically, ControlNet locks the parameters of the original neural network block model and copies a trainable copy, and finally connects them together using zero convolution. The locked weights remain unchanged, so that it maintains the original image generation capability of stable diffusion. The other part is a trainable copy, which is used to learn to add conditions. The condition here is to learn the contour of the original image so that when it is generated, there is no obvious difference between the contours of the generated image and the original image, that is, to maintain its authenticity.
[0130] In an optional embodiment, the generating system further includes:
[0131] Content delivery module, used to deliver marketing images into content;
[0132] The third acquisition module is used to obtain the delivery data of the marketing pictures;
[0133] The fourth acquisition module is used to acquire target marketing images according to the delivery data;
[0134] The storage module is used to store the target marketing pictures in the product gallery.
[0135] In this embodiment, the online community where the produced and cropped marketing images are placed is used, and the online placement effect is recovered, and the target marketing images with good effects are maintained in the product gallery. Specifically, after manual review by the operator, the image quality and article content are guaranteed to meet the requirements before being placed, and the corresponding marketing product’s business card will be attached when the image is placed. Effect recovery: data such as online exposure, clicks, and transaction amounts of the article are recovered; iterative optimization: based on the online effect data, feedback is given to the marketing effect, and images with good marketing effects are maintained.
[0136] In an optional embodiment, the screening module is used to perform quality assessment on product images through an image quality model according to product information to screen out target product images.
[0137] In this embodiment, according to the product information, the quality of the product images is evaluated through an image quality score model, and the product images with better quality (i.e., target product images) are screened out. Then, through the method of AIGC (Artificial Intelligence Generative Content), image-to-image transformation is performed based on the screened high-quality product images to prevent repeated use of images.
[0138] In an alternative embodiment, the cropping module includes:
[0139] A first acquisition unit for acquiring a preselected cropping frame;
[0140] A second acquisition unit for inputting the target product image and the preselected cropping frame into a target detection algorithm to obtain the target area in the target product image;
[0141] A third acquisition unit for inputting the target product image into a convolutional network to obtain the feature map corresponding to the target product image;
[0142] A pooling processing unit for performing pooling processing on the feature map to obtain the regional features corresponding to the preselected cropping frame area and the target area;
[0143] A fourth acquisition unit for using the regional features as the network nodes of an adaptive attention map convolutional network to obtain the visual semantic information and spatial composition relationship of different image nodes in the target product image;
[0144] A sorting unit for sorting the preselected cropping frames based on the visual semantic information and spatial composition relationship to obtain the target preselected cropping frame;
[0145] A cropping unit for cropping the target product image based on the target preselected cropping frame to obtain the marketing image of the product to be marketed.
[0146] In this embodiment, the target product image is cropped through a cropping model to meet the size requirements of the target placement position; specifically, since there are specific size requirements for the placement image and image-to-image generation cannot customize the generated image size, it is necessary to implement cropping of the generated image through a cropping algorithm. And the cropping algorithm should尽可能保留图片信息以及考虑到美观性。给定一个待裁剪图像的长和高分别为w和h,需要将其裁剪为目标尺寸的长和高分别为x和y。
[0147] 1. Generation of preselected cropping frame: To ensure that the retained image can have the maximum amount of information as much as possible, the cropping is based on a fixed size to retain the maximum amount of information.
[0148] If h < w * y / x: Then, with a step size x_step of w / 30, sample the length from left to right while keeping the height fixed. It should be noted that the part "尽可能保留图片信息以及考虑到美观性" in the original text seems to be an incomplete or inaccurate expression. The translation is made based on the best understanding of the overall context. If there is a more accurate or complete expression for this part, it can be adjusted accordingly.
[0149] If h>w*y / x: the height is sampled with a step size y_step of h / 30 and a fixed length from top to bottom.
[0150] At most 30 pre-selected cropping boxes are generated.
[0151] 2. Input the corresponding target product image and pre-selected cropping box into YoloV6 to detect the target area in the target product image.
[0152] 3. Pass the target product image into the convolutional network to obtain the corresponding feature map, and implement the pooling operation through RoIAlign and RoDAlign to obtain the regional features of the cropping box area and the detected target area.
[0153] 4. The regional features corresponding to the cropping frame area and the detected target area are used as nodes of the AAG (Adaptive Attention Graph) network to obtain the visual semantic information and spatial composition relationship of different image nodes in the image. Finally, several pre-selected cropping frames are aesthetically sorted, and the cropping frame with the highest score is output as the target pre-selected cropping frame selected by the model; the target product image is cropped based on the target pre-selected cropping frame to obtain the marketing image of the product to be marketed.
[0154] This embodiment uses AIGC and deep learning related technologies to achieve the generation of marketing pictures for products. By placing marketing pictures and texts in the community, online travel products are directed to achieve the purpose of marketing. The marketing pictures generated by this embodiment can ensure uniqueness and authenticity. In addition, the picture production process proposed in this embodiment is highly automated, which not only improves the efficiency of generation, but also reduces labor costs. In terms of efficiency, the community article production capacity can reach 1,000 articles / week, and in terms of quality, the overall excellence rate of articles reaches more than 90%.
[0155] Furthermore, this embodiment realizes the generation of original materials for marketing pictures by utilizing related technologies such as controllable generation of stable diffusion and intelligent cropping, solves the problem of the source and processing of picture materials when conducting marketing in tourism scenarios, and also solves the problem of exhaustion of marketing materials. In addition, the cropping algorithm realizes the automatic processing of original materials to meet online requirements, greatly improving the production efficiency of materials and reducing manpower consumption; finally, by recycling the online delivery effect, the subsequent picture generation is guided; a highly automated and efficient picture production and processing process is realized. Through this technology, product information can be imported in batches, and then the marketing pictures of the corresponding products can be obtained in batches for online promotion and delivery and finally the effect is recycled. In addition, the generation method of marketing pictures proposed in this embodiment can ensure the strong correlation between pictures and product content; ensure the diversity of pictures when the same product is produced multiple times; and improve the authenticity of marketing pictures.
[0156] As for the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The system embodiment described above is only illustrative, wherein the units described as separate components may or may not be physically separated, and the components as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the disclosed solution.
[0157] Example 3
[0158] Figure 3 This is a schematic diagram of the structure of an electronic device shown in Embodiment 3 of the present disclosure, wherein the electronic device includes a memory, a processor, and a computer program stored in the memory and used to run on the processor, and when the processor executes the computer program, the method for generating marketing pictures described in any of the above embodiments is implemented. Figure 3 The electronic device 90 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.
[0159] like Figure 3 As shown, the electronic device 90 may be in the form of a general-purpose computing device, for example, it may be a server device. The components of the electronic device 90 may include, but are not limited to: at least one processor 91, at least one memory 92, and a bus 93 connecting different system components (including the memory 92 and the processor 91).
[0160] The bus 93 includes a data bus, an address bus, and a control bus.
[0161] The memory 92 may include a volatile memory, such as a random access memory (RAM) 921 and / or a cache memory 922 , and may further include a read-only memory (ROM) 923 .
[0162] The memory 92 may also include a program tool 925 (or utility) having a set (at least one) of program modules 924, such program modules 924 including but not limited to: an operating system, one or more application programs, other program modules and program data, each of which or some combination may include an implementation of a network environment.
[0163] The processor 91 executes various functional applications and data processing by running the computer program stored in the memory 92, such as the method for generating a marketing picture provided in any of the above embodiments.
[0164] The electronic device 90 may also communicate with one or more external devices 94 (e.g., keyboards, pointing devices, etc.). Such communication may be performed via an input / output (I / O) interface 95. Furthermore, the electronic device 90 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 96. Figure 3 As shown, the network adapter 96 communicates with other modules of the electronic device 90 via the bus 93. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device 90, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems, etc.
[0165] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided into multiple units / modules to be embodied.
[0166] Example 4
[0167] Embodiment 4 of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the method for generating a marketing picture provided in any of the above embodiments is implemented.
[0168] The readable storage medium may include but is not limited to: a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device or any suitable combination of the above.
[0169] Example 5
[0170] Embodiment 5 of the present disclosure also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for generating marketing pictures.
[0171] Among them, the program code for executing the computer program product of the present disclosure can be written in any combination of one or more programming languages, and the program code can be executed completely on the user device, partially on the user device, as an independent software package, partially on the user device and partially on a remote device, or completely on the remote device.
[0172] Although the specific embodiments of the present disclosure are described above, those skilled in the art should understand that this is only an example, and the protection scope of the present disclosure is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present disclosure, but these changes and modifications all fall within the protection scope of the present disclosure.
Claims
1. A method for generating a marketing picture, characterized in that: The generation method comprises: Obtain product information of products to be marketed; Filter out target product images according to the product information; The target product image is cropped to obtain a marketing image of the product to be marketed.
2. The method for generating a marketing picture according to claim 1, characterized in that: Before the step of obtaining the commodity information of the commodity to be marketed, the generation method further includes: Obtain the product identification number of the product to be marketed; The step of obtaining the commodity information of the commodity to be marketed comprises: The commodity information of the commodity to be marketed is obtained according to the commodity identification number.
3. The method for generating a marketing picture according to claim 1, characterized in that: After the step of filtering out the target product image according to the product information, the generation method further includes: Performing image-to-image processing on the target product image to obtain the target product image after image-to-image processing; The step of cropping the target product image to obtain the marketing image of the product to be marketed includes: The target product image after the raw image processing is cropped to obtain a marketing image of the product to be marketed.
4. The method for generating a marketing picture according to claim 1, characterized in that: After the step of cropping the target product image to obtain the marketing image of the product to be marketed, the generation method further includes: Using the marketing pictures for content delivery; Obtaining delivery data of the marketing image; Obtaining a target marketing image according to the delivery data; The target marketing picture is stored in a product picture library.
5. The method for generating a marketing picture according to claim 1, characterized in that: The step of filtering out target product images according to the product information includes: The product images are quality evaluated using a picture quality model according to the product information to select the target product images.
6. The method for generating a marketing picture according to claim 1, characterized in that: The step of cropping the target product image to obtain the marketing image of the product to be marketed includes: Get the preselected cropping frame; Inputting the target product image and the pre-selected cropping frame into a target detection algorithm to obtain a target area in the target product image; Input the target product image into a convolutional network to obtain a feature map corresponding to the target product image; Performing pooling processing on the feature map to obtain regional features corresponding to the pre-selected cropping frame area and the target area; Using the regional features as network nodes of an adaptive attention graph convolutional network to obtain visual semantic information and spatial composition relationships of different image nodes in the target product image; Sorting the pre-selected cropping boxes based on the visual semantic information and the spatial composition relationship to obtain a target pre-selected cropping box; The target product image is cropped based on the target pre-selected cropping frame to obtain a marketing image of the product to be marketed.
7. A system for generating marketing pictures, characterized in that: The generation system comprises: A first acquisition module is used to acquire product information of the product to be marketed; A screening module, used to screen out target product images according to the product information; The cropping module is used to crop the target product image to obtain a marketing image of the product to be marketed.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and used to run on the processor, characterized in that: When the processor executes the computer program, the method for generating a marketing picture according to any one of claims 1 to 6 is implemented.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for generating a marketing picture according to any one of claims 1 to 6 is implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for generating a marketing picture as described in any one of claims 1 to 6 is implemented.
