A periodically updated picture management system and management method

Abstract

The present invention relates to the technical field of commercial image management, and in particular to a periodically updated image management system and management method. The management system comprises an image library, an acquisition module, a feature extraction module, a feature matching module, a determination module and an update module. The feature extraction module is configured to obtain image features of display images of similar commodities; the feature matching module is configured to determine a candidate image set of a current period from the image library according to the number of clicks and image features of display images of similar commodities in the previous period; the determination module is configured to determine display images of commodities in the current period from the candidate image set based on the number of collections of commodities in the previous period; by accurately matching multiple features in the image library, the candidate image that best meets the current browsing trend can be selected from the image library, and the similarity is further selected for the misrecognition of the image, thereby effectively improving the image display effect under the premise of ensuring the discernibility of the displayed image.

Description

Technical Field

[0001] The present invention relates to the field of commercial picture management, and in particular to a periodically updated picture management system and management method. Background Art

[0002] With the rapid development of e-commerce platforms, display images play a vital role in attracting user attention and increasing product click-through rates. However, existing display image management methods usually rely on manual selection or static display, lacking the ability to dynamically adjust based on user browsing preferences and behavior data, which can easily lead to the display effect being unable to adapt to changes in user interests in a timely manner, thereby reducing the exposure rate and click-through rate of the corresponding items.

[0003] For example, the prior art discloses an automatic image generation and update management method, which specifically includes: obtaining a product for which promotional images are to be generated, and extracting keywords based on the description information of the product, determining the associated image module of the product and the generation of the promotional image based on the keywords, treating other products on the platform that are of the same type as the product as products of the same type, and determining whether the promotional image of the product needs to be verified and updated based on the number of products of the same type and the similarity between the description information and the product information, the online time of the promotional image of the product, and the comprehensive similarity, thereby avoiding the risk of infringement due to the promotional image being too similar to the promotional images of other products.

[0004] However, the above picture management solution lacks dynamic updating and optimization for picture updating, and cannot effectively ensure that the display effect of the picture matches the current browsing trend, which leads to poor display effect of the displayed pictures. Summary of the invention

[0005] The purpose of the present invention is to provide a periodically updated picture management system and management method to solve the problem that the existing technology lacks dynamic updating and optimization for picture updates, cannot effectively ensure that the display effect of pictures matches the current browsing trend, and thus leads to poor display effect of displayed pictures.

[0006] To this end, on one hand, the present invention provides a periodically updated picture management system, the periodically updated picture management system comprising:

[0007] An image library, which stores several candidate images of products;

[0008] A collection module is configured to collect display images of similar products on the platform and the corresponding click counts and favorite counts;

[0009] A feature extraction module is configured to obtain image features of display images of similar products;

[0010] A feature matching module is configured to determine a candidate image set for the current period from the image library according to the number of clicks and image features of display images of similar products in the previous period;

[0011] A determination module is configured to determine, from the candidate image set, a display image of a product in a current period based on the number of collections of the product in a previous period;

[0012] An updating module is configured to determine whether to update the image library according to the number of times the display image is changed every preset number of consecutive cycles, and to update the image library based on image features of similar products in each cycle;

[0013] The image features include fonts and font colors, as well as contrast, brightness, saturation and border styles.

[0014] As a preferred technical solution for a periodically updated picture management system, the process of the feature matching module determining the candidate image set of the current period includes:

[0015] Display images of similar products with preset rankings before determining the number of clicks;

[0016] Determine similar image features of display images of similar products;

[0017] Extract candidate images matching any one of the similar image features from the image library and aggregate them into a candidate image set.

[0018] As an optimal technical solution for a periodically updated image management system, the determination module obtains the number of collections of a product in the previous period, compares the number of collections with a threshold, and if it is not greater than the threshold, selects the alternative image with the largest number of matching items with the similar image features as the display image of the product in the current period.

[0019] As an optimal technical solution for a periodically updated image management system, the determination module compares the number of collections with a threshold. If the number is greater than the threshold, the candidate image with the highest cosine similarity to the display image of the previous cycle is selected from the candidate image set as the display image of the product in the current cycle.

[0020] As a preferred technical solution for a periodically updated image management system, the update module compares the number of replacements with a preset number. If it is less than the preset number, the original image library is maintained; if it is not less than the preset number, it is determined that the image library needs to be updated.

[0021] As a preferred technical solution for a periodically updated picture management system, the process of the updating module updating the image library includes:

[0022] The similar image features of each period are counted, the common feature items among the similar image features are determined, and the corresponding items of the candidate images in the image library are updated to be the same as the common feature items.

[0023] As a preferred technical solution for a periodically updated picture management system, the specific process of the update module determining the common feature items in the features of each similar image includes:

[0024] Extract feature items from similar image features in each cycle, including font, font color, contrast, brightness, saturation and border style;

[0025] Count the extracted feature items and record the frequency of occurrence of each feature item in each cycle;

[0026] The feature items whose occurrence frequency in each cycle exceeds the frequency threshold are determined as common feature items.

[0027] On the other hand, the present invention provides a periodically updated picture management method, which is applied to the periodically updated picture management system described in any of the above schemes, and the periodically updated picture management method includes:

[0028] Collect the display images of similar products on the platform and the corresponding click and favorite times;

[0029] Extracting image features of display images of the same type of products;

[0030] Based on the number of clicks and image features of display images of similar products in the previous cycle, select and determine the candidate image set for the current cycle from the image library;

[0031] According to the number of collections of the product in the previous cycle, the display image of the product in the current cycle is determined from the candidate image set.

[0032] As a preferred technical solution of the periodically updated picture management method, the periodically updated picture management method further includes:

[0033] Every a preset number of cycles, determining whether to update the image library according to the number of times the displayed image is changed;

[0034] In response to the determination result of updating the image library, the image library is updated based on the image features of the same type of products in each period.

[0035] The beneficial effects of the present invention are:

[0036] The present invention can make the display effect of the displayed image more in line with the current browsing trend by periodically updating the displayed image. By accurately matching multiple features in the image library: contrast, brightness, saturation, border style, font and font color, the alternative image in the image library that best meets the current browsing trend can be selected. In addition, considering the user's memory inertia for the original displayed image after collection, if the number of collections is large, it is necessary to further select the similarity according to the misrecognition of the image, so as to ensure the discernibility of the displayed image under the premise of ensuring the updating effect, thereby effectively improving the feasibility of image display and forming a virtuous circle.

[0037] Furthermore, the present invention can accurately capture the current user browsing trends and ensure that the displayed images have high appeal and click-through rate by using a collection module to collect real-time display images of similar products on the platform and the corresponding number of clicks, and combining the feature extraction module to analyze image features such as font, font color, contrast, brightness, saturation and border style.

[0038] Furthermore, the feature matching module of the present invention screens and generates a candidate image set from the image library based on the click data and image features of the previous cycle. Combined with the display images with preset rankings before the number of clicks and the matching items of similar image features, this process ensures the accuracy of the candidate images and lays the foundation for subsequent judgment and display optimization.

[0039] Furthermore, the determination module of the present invention analyzes the number of times the product was collected in the previous period, and selects the candidate image with the most matching items when the number of collections is lower than the threshold; when the number of collections is higher, the candidate image closest to the original display image is selected based on the cosine similarity. This mechanism fully considers the user's memory inertia of the original display image, and effectively avoids the problem of users misrecognizing product images due to frequent updates. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a structural block diagram of a periodically updated picture management system in an embodiment of the present invention;

[0041] Figure 2 The process of the periodically updated picture management method in the embodiment of the present invention is as follows Figure 1 ;

[0042] Figure 3 The process of the periodically updated picture management method in the embodiment of the present invention is as follows Figure 2 . DETAILED DESCRIPTION

[0043] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] In the description of the present invention, it should be noted that the orientations or positional relationships indicated by the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. Among them, the terms "first position" and "second position" are two different positions, and the first feature "above", "above" and "above" the second feature include the first feature being directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. The first feature "below", "below" and "below" the second feature include the first feature being directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0045] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0046] Embodiments of the present invention are described in detail below, examples of which 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 only used to explain the present invention, and cannot be understood as limiting the present invention.

[0047] like Figure 1 As shown, this embodiment provides a periodically updated picture management system, including: an image library, a collection module, a feature extraction module, a feature matching module, a determination module and an update module;

[0048] Among them, the image library has several candidate images of the product;

[0049] The collection module is configured to collect display images of similar products on the platform and the corresponding click counts and favorite counts;

[0050] The feature extraction module is configured to obtain image features of display images of similar products;

[0051] The feature matching module is configured to determine a candidate image set for the current period from the image library according to the number of clicks and image features of display images of similar products in the previous period;

[0052] The determination module is configured to determine the display image of the product in the current cycle from the candidate image set based on the number of collections of the product in the previous cycle;

[0053] The updating module is configured to determine whether to update the image library according to the number of times the display image is changed every preset number of consecutive cycles, and to update the image library based on the image features of the same type of products in each cycle;

[0054] Among them, image features include fonts and font colors, as well as contrast, brightness, saturation and border styles.

[0055] In the above embodiment, by periodically updating the displayed image, the display effect of the displayed image can be made more in line with the current browsing trend. By accurately matching multiple features in the image library: contrast, brightness, saturation, border style, font and font color, the alternative image in the image library that best suits the current browsing trend can be selected. Taking into account the user's memory inertia for the original displayed image after collection, if the number of collections is large, further selection of similarity is required for the misrecognition of the image, so as to ensure the recognizability of the displayed image while ensuring the updating effect, thereby effectively improving the feasibility of image display and forming a virtuous circle.

[0056] In detail, the feature extraction module receives the display image data of similar products from the acquisition module through the data interface, and performs standardization processing on the received image data, including unified resolution, noise removal, and grayscale processing, to provide consistent input conditions for subsequent feature extraction. The module classifies and extracts the features of the display image, uses optical character recognition technology to analyze the text part in the image, extracts the font type, size, and color information, and combines the color distribution detection algorithm to obtain the RGB value or HSV value of the font color; uses the brightness gradient statistical method to calculate the local contrast characteristics of the image and extract the global average contrast value; uses the pixel value distribution to calculate the mean and standard deviation of the overall brightness; extracts the saturation component in the HSV color space and statistically normalizes it into a feature vector; uses the edge detection algorithm to identify the shape, width, and color of the border in the image, and converts it into a structured feature descriptor. The various extracted image features are converted into vector representations and stored in the feature image library for subsequent calls. At the same time, consistency and integrity checks are performed to ensure that the features meet the standards and can accurately reflect the visual features of the display image. The feature vectors generated are finally pushed to the feature matching module to provide accurate data support for the system's screening and matching.

[0057] In order to reduce the system computing power, the candidate images are initially screened through the feature matching module. The process of the feature matching module determining the candidate image set of the current cycle includes:

[0058] Determine the display images of similar products with a preset ranking (exemplarily the top 30, which can be adjusted according to the actual scenario) before the number of clicks;

[0059] Determine similar image features of display images of similar products (contrast, brightness or saturation are in the same data range, or, the same font, font color, border style);

[0060] Extract candidate images that match any of the similar image features in the image library and summarize them into a candidate image set. In detail, first, by counting the display images of similar products with the top preset rankings in the number of clicks in the previous cycle, extract the performance of these display images in user behavior data as the preliminary screening basis to ensure that the screened images can reflect the user's browsing preferences; then, based on the extracted display images, analyze their corresponding image features, including fonts, font colors, contrast, brightness, saturation and border styles, and construct a similar image feature set to comprehensively describe the visual attributes of the display images. Subsequently, by searching the candidate images in the image library one by one, the images that match any of the similar image features are taken as candidate targets, and at the same time, they are prioritized based on the multi-dimensional matching effects to screen out a candidate image set to support subsequent display image judgment and screening optimization.

[0061] Considering the visual impact of display image changes on users who have collected the product, the determination module obtains the number of collections of the product in the previous cycle, compares the number of collections with a threshold (which can be adjusted according to actual scenarios such as cycle length and historical click count, and is selected between 1000 and 5000 times in this embodiment). If it is not greater than the threshold, the candidate image with the largest number of matching items with similar image features is selected as the display image of the product in the current cycle;

[0062] If it is greater than the threshold, the alternative image with the highest cosine similarity to the display image of the previous cycle is selected from the alternative image set as the display image of the product in the current cycle. The above embodiment takes into account the situation that after the user collects the product, the display image may be mistaken for a non-original product due to the change of the display image. In this case, it is necessary to select an alternative image that is similar to the original display image and relatively in line with the current browsing trend from the alternative images for display to improve recognizability. By analyzing the number of collections of the product in the previous cycle, when the number of collections is lower than the threshold, the alternative image with the largest number of matches is selected; when the number of collections is higher, the alternative image that is closest to the original display image is selected based on the cosine similarity. This mechanism fully takes into account the user's memory inertia of the original display image, and effectively avoids the problem of users misrecognizing product images due to frequent updates.

[0063] This embodiment also configures an update module to intermittently and continuously update the image library. The update module compares the number of changes with the preset number. If it is less than the preset number, the original image library is maintained. If it is not less than the preset number, it is determined that the image library needs to be updated. It can be understood that the preset number needs to be set in combination with the cycle length and the cycle interval of continuous analysis. In this embodiment, it is 5 times. In detail, if the display image is frequently changed in each cycle, it means that the candidate images in the image library cannot effectively match the similar image features. The candidate images need to be optimized for various features to reduce the number of updates of the candidate images and improve the display stability.

[0064] Based on the above embodiment, the process of updating the image library by the update module includes:

[0065] The similar image features of each period are counted, the common feature items among the similar image features are determined, and the corresponding items of the candidate images in the image library are updated to be the same as the common feature items.

[0066] In detail, the specific process of the updating module determining the common feature items in the features of each similar image includes:

[0067] Extract feature items from similar image features in each cycle, including font, font color, contrast, brightness, saturation and border style;

[0068] Count the extracted feature items and record the frequency of occurrence of each feature item in each cycle;

[0069] The feature items whose frequency of appearance in each cycle exceeds the frequency threshold (in this embodiment, the preset number is 8 consecutive cycles, and the frequency threshold is 5 times in 8 consecutive cycles, which can be adjusted according to the actual scenario) are determined as common feature items. The frequent appearance of common feature items in consecutive cycles indicates that the corresponding feature items have the ability to match the current browsing trend. Optimizing the candidate images towards the common feature items can further improve the display stability.

[0070] See also Figure 2 As shown, this embodiment further provides a periodically updated picture management method, which is applied to the periodically updated picture management system described in any of the above solutions. The periodically updated picture management method includes:

[0071] Step S1, collecting display images of similar products on the platform and the corresponding click counts and favorite counts;

[0072] Step S2, extracting image features of display images of similar products;

[0073] Step S3, based on the number of clicks and image features of the display images of similar products in the previous cycle, screen and determine the candidate image set for the current cycle from the image library;

[0074] Step S4, determining the display image of the product in the current cycle from the candidate image set according to the number of collections of the product in the previous cycle.

[0075] To further ensure the image display effect, please refer to Figure 3 As shown, the picture management method of this embodiment also includes the following steps:

[0076] Step S01, every preset number of cycles, determining whether to update the image library according to the number of times the displayed image is changed;

[0077] Step S02 : in response to the determination result of updating the image library, the image library is updated based on the image features of the same type of products in each period.

[0078] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the devices, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based device that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0079] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A periodically updated picture management system, characterized in that: include: An image library, which stores several candidate images of products; A collection module is configured to collect display images of similar products on the platform and the corresponding click counts and favorite counts; A feature extraction module is configured to obtain image features of display images of similar products; A feature matching module is configured to determine a candidate image set for the current period from the image library according to the number of clicks and image features of display images of similar products in the previous period; A determination module is configured to determine, based on the number of collections of the product in the previous cycle, a display image of the product in the current cycle from the candidate image set; An updating module is configured to determine whether to update the image library according to the number of times the display image is changed every preset number of consecutive cycles, and to update the image library based on image features of similar products in each cycle; The image features include fonts and font colors, as well as contrast, brightness, saturation and border styles; The process of determining the candidate image set of the current cycle by the feature matching module includes: Display images of similar products with preset rankings before determining the number of clicks; Determine similar image features of display images of similar products; Extracting candidate images matching any one of the similar image features from the image library and summarizing them into a candidate image set; The determination module obtains the number of times the product was collected in the previous cycle, compares the number of times collected with a threshold, and if the number is not greater than the threshold, selects the candidate image with the largest number of matching items with the similar image features as the display image of the product in the current cycle; The determination module compares the number of collections with a threshold value, and if the number is greater than the threshold value, selects from the candidate image set a candidate image with the highest cosine similarity to the display image of the previous cycle as the display image of the product in the current cycle.

2. The periodically updated picture management system according to claim 1, characterized in that: The updating module compares the number of replacements with a preset number, and if the number is less than the preset number, maintains the original image library; if the number is not less than the preset number, determines that the image library needs to be updated.

3. The periodically updated picture management system according to claim 2, characterized in that: The process of updating the image library by the updating module includes: The similar image features of each period are counted, the common feature items among the similar image features are determined, and the corresponding items of the candidate images in the image library are updated to be the same as the common feature items.

4. The periodically updated picture management system according to claim 3, characterized in that: The specific process of the updating module determining the common feature items in the features of each similar image includes: Extract feature items from similar image features in each cycle, including font, font color, contrast, brightness, saturation and border style; Count the extracted feature items and record the frequency of occurrence of each feature item in each cycle; The feature items whose occurrence frequency in each cycle exceeds the frequency threshold are determined as common feature items.

5. A periodically updated picture management method, characterized in that: Applied to the periodically updated picture management system according to any one of claims 1 to 4, the periodically updated picture management method comprises: Collect the display images of similar products on the platform and the corresponding click and favorite times; Extracting image features of display images of the same type of products; Based on the number of clicks and image features of display images of similar products in the previous cycle, select and determine the candidate image set for the current cycle from the image library; According to the number of collections of the product in the previous cycle, the display image of the product in the current cycle is determined from the candidate image set.

6. The periodically updated picture management method according to claim 5, characterized in that: The periodically updated picture management method further includes: Every a preset number of cycles, determining whether to update the image library according to the number of times the displayed image is changed; In response to the determination result of updating the image library, the image library is updated based on the image features of the same type of products in each period.

Images