Image-driven cross-platform e-commerce commodity price comparison processing method and system

Abstract

The application discloses an image-driven cross-platform e-commerce commodity price comparison processing method and system, and relates to the technical field of price comparison processing. The method comprises the following steps: performing structure slice partition processing on each commodity image, constructing each region as an encoding node, and generating a structure matrix of the commodity image; performing corresponding matching on the structure matrices of multiple platform commodity images, and identifying image commodity pairs with comparison potential based on the combination relationship between the region structure and the semantic features; extracting cross feature groups from the historical price sequences and attributes of each platform commodity, performing price-attribute-structure joint evaluation, and hierarchically screening effective price comparison nodes through cross combination judgment rules; generating a multi-path price comparison set based on the effective price comparison nodes, and outputting a recommended result in stages according to the target commodity state. The application not only improves the accuracy and adaptability of the price comparison processing, but also guarantees the stability and intelligent response capability of the price comparison logic under the condition of complex e-commerce image data.

Description

Technical Field

[0001] This invention relates to the field of price comparison processing technology, and in particular to an image-driven cross-platform e-commerce product price comparison processing method and system. Background Technology

[0002] Currently, with the booming development of e-commerce platforms, the multi-source heterogeneity of product information is becoming increasingly prominent, and users' demand for comparing product prices, attributes, and image information across different platforms is also constantly growing. Traditional keyword-driven price comparison methods often rely on product titles or descriptions for retrieval and matching. However, in practical applications, inconsistent product naming rules, redundant descriptions, or semantic biases lead to inaccurate or seriously incomplete price comparison results. To address this, some studies have attempted to introduce image recognition and visual feature extraction techniques to compensate for the shortcomings of pure text information. However, image-driven price comparison technology is still in its early stages of development, and existing methods still have significant limitations in structural modeling, cross-platform image matching, image semantic fusion, and price comparison accuracy. Furthermore, effective mechanisms for modeling structural hierarchical information and joint reasoning of multi-dimensional information such as images, prices, and attributes have not yet been established, necessitating more targeted and robust image price comparison processing solutions.

[0003] CN109118325A discloses a method for collecting product information and comparing prices in real time across multiple e-commerce platforms. Its main technical features include: allowing users to input product keywords or images, extracting product names based on image recognition, performing searches on various platforms using these names, calculating costs based on user account information, and generating price comparison charts. While this method does achieve automated information acquisition and price comparison output, it has two shortcomings: first, it only uses images as auxiliary information, failing to systematically model the structure, semantics, and regional information of the images themselves, resulting in insufficient image-driven characteristics for price comparison; second, it does not achieve corresponding matching of image structures across products on different platforms, lacking the ability to visually identify potential matching product pairs. Therefore, this technology has low recognition accuracy when handling products with diverse image semantics, homogeneous brands, or ambiguous names.

[0004] CN103412938B provides a product price comparison method based on interactive multi-target image extraction. Users select image regions on the client side, extract image features, and match them against a pre-built visual feature database in an index file, thus returning similar products and price comparison rankings. This technology attempts to construct a product image feature index and achieve interactive user recognition. Its shortcomings are mainly reflected in: image features are limited to global or shallow visual attributes, lacking partitioning modeling and node encoding of image structure; furthermore, its price comparison logic is limited to simple price ranking driven by image similarity, failing to introduce the cross-correlation between product price sequences, attribute information, and image structure, making it difficult to support multi-dimensional price comparison judgments in complex product scenarios. Summary of the Invention

[0005] In view of the problems existing in the information retrieval and price comparison methods based on product images, this invention is proposed.

[0006] Therefore, the problem this invention aims to solve is how to improve the accuracy and adaptability of price comparison processing.

[0007] Furthermore, it ensures the stability and intelligent response capabilities of the price comparison logic when faced with complex e-commerce image data.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0009] In a first aspect, the present invention provides an image-driven cross-platform e-commerce product price comparison processing method.

[0010] This process includes: collecting product images of the target product from multiple platforms; performing structural slicing and partitioning on each product image, constructing each region as an encoding node, and generating a structural matrix of the product image; performing correspondence matching on the structural matrices of product images from multiple platforms, and identifying image-product pairs with comparison potential based on the combination relationship between regional structure and semantic features; extracting cross-feature groups from the historical price sequences and attributes of products on each platform based on the image-product pairs, performing joint evaluation of price-attribute-structure, and hierarchically filtering effective price comparison nodes through cross-combination judgment rules; generating a multi-path price comparison set based on the effective price comparison nodes, and outputting recommendation results in stages according to the status of the target product.

[0011] As a preferred embodiment of the image-driven cross-platform e-commerce product price comparison processing method of the present invention, the method of corresponding matching of the structure matrices of product images from multiple platforms includes: adding image source tags to the structure matrices corresponding to the product images collected from all platforms, classifying the structure matrices according to the source tags, and assigning the structure matrices of the same product generated on different platforms to different subsets to ensure that there is no pairing of images with the same source when combining them; pairing the structure matrix in each platform with the structure matrix of the same product in other platforms to generate cross-platform structure matrix pairs, each pair containing structure matrices from two different source platforms; and combining each pair of structure matrices in two directions to generate a bidirectional structure matching group.

[0012] As a preferred embodiment of the image-driven cross-platform e-commerce product price comparison processing method of the present invention, the identification of image product pairs with comparison potential includes: for each set of structure matrices, extracting the set of structure encoding nodes, and constructing an intersection node set and a union node set based on the node position index and label semantics in the structure encoding nodes; calculating the ratio of the number of nodes in the intersection node set and the union node set as the original intersection-union ratio index; calculating the coverage ratio of the node sets in the two structure matrices in the entire image structure as a compensation coefficient, and multiplying it by the initial intersection-union ratio to form an adjusted intersection-union ratio; if the adjusted intersection-union ratio is higher than the judgment threshold T... s Then the structure matching group will be labeled as an image item pair.

[0013] As a preferred embodiment of the image-driven cross-platform e-commerce product price comparison processing method of the present invention, the price-attribute-structure joint evaluation includes: extracting the time evolution structure matrix and corresponding price data of products from each platform in the confirmed image product pairs, and constructing a structure-price coupling vector of structure change and price fluctuation; based on the structure matching group, constructing a structure pair bitmap and binding the structure-price coupling vector to form a structure-price fusion feature group.

[0014] As a preferred embodiment of the image-driven cross-platform e-commerce product price comparison processing method of the present invention, the cross-combination judgment rule includes: constructing a cross-stability factor graph, where each factor node corresponds to a joint stability index of three dimensions; wherein, the three dimensions include time slice overlap, structural coding semantic consistency, and structural alignment density; based on the joint judgment of the three dimensions, each node in the graph is assigned the following hierarchical labels: if the structural coding similarity corresponding to the factor node is higher than a preset first threshold, the corresponding structural coding node falls in a continuous region in the structural alignment graph, and the dynamic time regularization distance between the corresponding product price curves within a preset time window is less than a preset second threshold, then it is marked as a high-confidence node; if a certain dimension index fluctuates but other dimensions can be compensated, it is marked as a conditional reference node; otherwise, it is marked as a secondary support node; starting from the high-confidence node, multiple collaborative judgment paths are constructed in combination with the connected conditional reference nodes.

[0015] As a preferred embodiment of the image-driven cross-platform e-commerce product price comparison processing method of the present invention, the hierarchical screening of effective price comparison nodes includes: during the construction of multiple collaborative judgment paths, path classification is formed according to the following rules: if the structural encoding sequence change of any two factor nodes in the path precedes the change of the product price sequence, and the overlap of their time windows exceeds the overlap threshold, the path is marked as a synchronous response channel; if more than 50% of the factor nodes in the path have a label semantic similarity higher than the first threshold, and the price change slope and structural change slope are consistent within a preset tolerance range, the path is marked as a structure-driven channel; if the structural matching degree between factor nodes in the path is lower than the second threshold, but their spatial positions are clustered in adjacent areas of the structure matrix, the path is marked as a regional focus channel; based on the collaborative path type and node level, three types of effective price comparison nodes are output hierarchically from the cross-stable factor map: high-confidence nodes in the synchronous response channel are used for accurate price comparison reference; stable reference nodes in the structure-driven channel are used for trend-based price comparison judgment; and conditional reference nodes in the regional focus channel are used as regional comparison support nodes.

[0016] As a preferred embodiment of the image-driven cross-platform e-commerce product price comparison processing method of the present invention, the step of outputting recommendation results in stages based on the target product status includes: dividing the output three types of effective price comparison nodes into node sets according to their respective collaborative path types, and constructing three types of multi-path price comparison sets; determining the latest evolution state of the target product in the structure matrix based on the change frequency and spatial clustering of the structure encoding nodes, identifying the structural change trend, and outputting the multi-path price comparison set according to the structural change trend.

[0017] Secondly, the present invention provides an image-driven cross-platform e-commerce product price comparison processing system, which includes: an image acquisition module, which acquires product images of the target product on multiple platforms, performs structural slicing and partitioning processing on each product image, constructs each region as an encoding node, and generates a structural matrix of the product image;

[0018] The matching and recognition module is used to perform correspondence matching on the structure matrix of product images from multiple platforms, and to identify image-product pairs with comparison potential based on the combination relationship between regional structure and semantic features.

[0019] The feature cross module is used to extract cross feature groups from the historical price series and attributes of products on various platforms based on image product pairs, to perform joint evaluation of price-attribute-structure, and to filter effective price comparison nodes in a hierarchical manner through cross combination judgment rules.

[0020] The recommendation output module is used to generate multi-path price comparison sets based on effective price comparison nodes, and output recommendation results in stages according to the status of the target product.

[0021] Thirdly, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the image-driven cross-platform e-commerce product price comparison processing method as described in the first aspect of the present invention are implemented.

[0022] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the image-driven cross-platform e-commerce product price comparison processing method as described in the first aspect of the present invention are implemented.

[0023] The beneficial effects of this invention are as follows: By constructing a structural matrix and using a joint modeling mechanism of structural-semantic features, this invention significantly improves the image comparison accuracy between products across platforms. Through node hierarchical division and cross-feature evaluation, it can achieve deep-level correlation mining between structural changes, price evolution, and attribute information, thereby ensuring higher timeliness and consistency of price comparison references. Simultaneously, relying on multi-path collaborative judgment and node-level hierarchical mechanisms, this invention can dynamically output personalized price comparison result recommendations based on the structural evolution trends of different target products. In summary, this invention not only improves the accuracy and adaptability of price comparison processing but also ensures the stability and intelligent response capability of price comparison logic under the condition of complex e-commerce image data. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A flowchart illustrating an image-driven method for cross-platform e-commerce product price comparison.

[0026] Figure 2 This is a structural diagram of an image-driven cross-platform e-commerce product price comparison processing system. Detailed Implementation

[0027] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0029] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0030] As mentioned in the background section, traditional keyword-driven price comparison methods often rely on product titles or descriptions for retrieval and matching. However, in practical applications, inconsistent product naming rules, redundant descriptions, or semantic biases lead to inaccurate or incomplete price comparison results. To address this, some studies have attempted to incorporate image recognition and visual feature extraction to compensate for the limitations of pure text information. However, image-driven price comparison technology is still in its early stages of development, and existing methods have significant limitations in structural modeling, cross-platform image matching, image semantic fusion, and price comparison accuracy. Furthermore, effective mechanisms for modeling structural hierarchical information and joint reasoning based on multi-dimensional image-price-attribute information have not yet been established, necessitating more targeted and robust image price comparison processing solutions.

[0031] Figure 1 This is a flowchart of an image-driven cross-platform e-commerce product price comparison processing method according to an embodiment of the present invention. Figure 1 As shown, the image-driven cross-platform e-commerce product price comparison processing method includes:

[0032] S1: Collect product images of the target product on multiple platforms, perform structural slicing and partitioning processing on each product image, construct each region as an encoding node, and generate a structural matrix of the product image.

[0033] Specifically, the first step is to collect product display images of the same target product on multiple mainstream e-commerce platforms through cross-platform data interfaces or web crawling. Priority is given to obtaining clear image samples with the least background interference, such as the main image, front view, and images with the least background interference, to ensure that the images have a good structural expression foundation.

[0034] Subsequently, structural slicing partitioning is performed on each product image. This partitioning process divides the entire image into several structural unit regions with fixed spatial sizes based on the image pixel size and the distribution patterns of visually salient regions. The partitioning method can employ a regular grid partitioning method or combine edge detection and region segmentation algorithms. This embodiment of the invention is not limited to a single method, aiming to achieve a slicing effect that better conforms to the physical configuration.

[0035] After slicing and partitioning, image feature descriptors are extracted from each region. These descriptors, combined with the region's spatial location, texture attributes, and edge features, are used to construct a structural encoding node for that region. Each structural encoding node includes a node location index, label semantics, and feature value summary.

[0036] Finally, all structural coding nodes are combined according to their spatial arrangement in the image to form a structural matrix corresponding to the target product image, which serves as the basic input structure for subsequent comparison and evaluation. This structural matrix not only preserves the spatial construction information of the image but also provides coding support for semantic comparison.

[0037] S2: Perform correspondence matching on the structure matrix of product images from multiple platforms, and identify image-product pairs with comparison potential based on the combination relationship between regional structure and semantic features.

[0038] S2.1: Based on the structure matrix corresponding to the product images of multiple platforms, the structure matrix is ​​bidirectionally paired and combined according to the image source to construct a structure matching group. Each group includes the structure matrix from two platforms.

[0039] Specifically, image source tags are added to the structure matrices corresponding to product images collected from all platforms. The structure matrices are then categorized according to the source tags, and the structure matrices of the same product generated on different platforms are assigned to different subsets to ensure that there are no pairings of images from the same source when combining them.

[0040] For each platform's structure matrix, pair it with structure matrices belonging to the same product on other platforms to generate cross-platform structure matrix pairs. Each pair contains structure matrices from two different source platforms.

[0041] For each pair of structure matrices, bidirectional structure matching groups are generated by combining them in two directions. For example, structure matrix A and structure matrix B will form two combinations: A→B and B→A, serving as independent structure matching perspectives.

[0042] In conventional methods, multi-platform image comparison typically uses image-level global features, such as SIFT and SURF keypoint descriptors, for coarse matching. This approach struggles to guarantee matching accuracy at the structural level, especially after product images have been edited by the platform (by adding borders, occlusions, etc.), making traditional methods prone to misjudgment. Our invention abandons traditional overall image feature matching and shifts to a region combination strategy at the structural matrix level, emphasizing region correspondence and semantic coupling features, thereby better reflecting the consistency of the product ontology.

[0043] S2.2: For each set of structure matrices, extract the set of structure encoding nodes, and construct the intersection node set and the union node set based on the position index and label semantics of the structure encoding nodes.

[0044] In practice, for any pair of structure matrices in the structure matching group, the node set is first extracted. Then, based on the spatial location index and label semantics of the nodes, the intersection node set and the union node set between the two sets are constructed. The intersection node set refers to the set of nodes in a pair of structure matrices that have the same label semantics and appear in similar position regions, emphasizing the structural consistency of local regions; while the union node set is the union of all node sets of the two structure matrices, representing the overall structural range.

[0045] In traditional structural matching methods, relying solely on spatial location (such as IoU or pixel overlap) often neglects semantic consistency. Our invention combines spatial location with label semantics as a judgment criterion, which significantly improves matching accuracy, effectively suppresses erroneous matching caused by visual confusion or image occlusion, and enhances the discriminativeness of the matching node set.

[0046] S2.3: Calculate the intersection-union ratio (IUR) index, adjust the IUR index according to the node coverage compensation coefficient, and determine whether the structure matching group meets the preset judgment threshold T. s The structural matching groups that meet the conditions are labeled as image product pairs.

[0047] Specifically, based on the intersection node set and the union node set, the ratio of the number of nodes between the two is calculated as the original intersection-union ratio index. This index can measure the degree of matching between two structure matrices in semantic space and structural region. The higher the value, the more similar the structures are.

[0048] The coverage ratio of the node sets in each of the two structure matrices in the entire image structure is statistically analyzed and used as a compensation coefficient to reflect the spatial coverage balance of the structure matching group. The compensation coefficient is calculated by merging the areas corresponding to all nodes in the image coordinate system and then calculating the ratio of the area to the total area of ​​the image.

[0049] The node coverage compensation coefficient is multiplied by the initial cross-union ratio to form the adjusted cross-union ratio, which is used to reduce the misjudgment situation where local areas are highly overlapping but the overall structure is inconsistent. It should be noted that if only the cross-union ratio is used, it is easy to misjudge products with highly matched structures in several small areas of the image as the same product, while ignoring the differences in the overall framework. This invention can effectively improve the global rationality of matching judgment by adding compensation coefficient constraints.

[0050] Based on the preset judgment threshold T s The adjusted crossover ratio (CRR) of each structural matching group is evaluated. If it exceeds the judgment threshold T, the result is considered positive. s If the matching group is found to be a valid product image pair, it means that the product images displayed on the two platforms are very likely different platform versions of the same product, and they are identified as image product pairs.

[0051] S3: Based on image-product pairs, extract cross-feature groups from the historical price sequences and attributes of products on various platforms, conduct joint evaluation of price-attribute-structure, and use cross-combination judgment rules to hierarchically filter effective price comparison nodes.

[0052] S3.1: In the confirmed image product pairs, extract the time evolution structure matrix and corresponding price data of products from each platform, and construct a structure-price coupling vector of structural changes and price fluctuations.

[0053] First, for each product image, from the established platform data samples, its historical image version sequence is obtained according to the time dimension. Priority is given to collecting product main images with clear structure and stable main viewpoints to ensure the continuity and accuracy of structural representation. Each historical image undergoes structural slicing and encoding processing according to the procedure in S1 to construct the structural matrix at that point in time.

[0054] After obtaining the structure matrix in the time series, node-level feature difference processing is performed on the corresponding structure coding nodes in each matrix (aligned one-to-one based on position index), that is, the coding value change of the same structure coding node at different time points is calculated.

[0055] These node evolution vectors are aggregated along the time dimension to form an overall structural evolution trajectory matrix. Simultaneously, the commodity price at each time point is extracted, constructing a one-to-one mapping between the structural encoding change sequence and the price change sequence.

[0056] Based on structural stability trajectories, time slices with sharp price jumps but no significant structural changes are identified and removed, while time slices with synchronous structural and price changes are retained, thus constructing a set of structure-price coupling vectors.

[0057] S3.2: Based on the structure matching group, construct the structure pair bitmap and bind the structure-price coupling vector to form a structure-price fusion feature group.

[0058] In the image product pair, based on the structure matching group selected in S2.3, the positions of the structure encoding nodes in the structure matrix of the two platforms are precisely aligned to construct a structure alignment map. In the structure alignment map, the time slices in the structure-price coupling vector in S3.1 are bound to the node pairs in the alignment map, thereby forming a structure-price fusion feature group, which provides a basis for subsequent price comparison node selection.

[0059] S3.3: Based on the structure-pair plot and time series data, construct a cross-stability factor map. In the cross-stability factor map, each factor node corresponds to a joint stability index in the following three dimensions:

[0060] Time slice overlap: Indicates the degree of temporal overlap in the evolution trajectory of the product structure in images on two platforms;

[0061] Structural encoding semantic consistency: Calculates the degree of semantic consistency between the labels of encoded nodes in the bitmap based on the structure;

[0062] Structural alignment density: represents the proportion and distribution concentration of alignment nodes in the entire structural matrix.

[0063] Based on the joint determination of the three dimensions, the following hierarchical labels are assigned to each factor node in the graph:

[0064] If the structural encoding similarity of a node is higher than a preset first threshold, the corresponding structural node falls within a continuous region in the structural alignment diagram, and the dynamic time warping distance between the corresponding commodity price curves within a preset time window is less than a preset second threshold, then it is marked as a high-confidence node. If a certain dimension of the indicator fluctuates but other dimensions can be compensated for, it is marked as a conditional reference node; otherwise, it is marked as a secondary support node. Here, structural encoding similarity refers to the cosine similarity of the encoding vectors of corresponding structural nodes in two platform commodities; the preset first threshold is a structural semantic similarity judgment threshold set by experience or training samples, such as 0.85; a continuous region refers to the spatial distance between two nodes in the structural alignment diagram being within a set range (such as a 3×3 neighborhood); the dynamic time warping distance of the price curves is used to measure the similarity of two imperfectly aligned time series in the time dimension; the preset second threshold is a set upper limit for the DTW distance, such as 1.5, used to determine the similarity of price fluctuation patterns.

[0065] S3.4: Starting from a highly reliable node, and combining it with the condition reference node of the connection, construct multiple collaborative judgment paths.

[0066] During the path construction process, path categories are formed according to the following rules:

[0067] If the structural coding sequence changes of any two factor nodes in the path precede the changes of the commodity price sequence, and the overlap of their time windows exceeds the overlap threshold, then the path is marked as a synchronous response channel.

[0068] If more than 50% of the factor nodes in a path have a semantic similarity to the label that is higher than the first threshold, and the price change slope and the structural change slope are consistent within a preset tolerance range, then the path is marked as a structure-driven channel.

[0069] If the structural matching degree (such as the cosine similarity of the encoded vectors) between structural encoding nodes in the path is lower than the second threshold, but the spatial positions of the nodes are clustered in adjacent regions (such as significant edge regions or central regions) in the structural matrix, then the path is marked as a region-focused channel.

[0070] Finally, based on the type of collaborative path and the level of nodes, three types of effective price comparison nodes are output from the cross-stability factor map: high-reliability nodes in the synchronous response channel are used for accurate price comparison reference; stable reference nodes in the structure-driven channel are used for trend-based price comparison judgment; and conditional reference nodes in the regional focus channel are used as regional comparison support nodes.

[0071] S4: Generate a multi-path price comparison set based on effective price comparison nodes, and output recommendation results in stages according to the status of the target product.

[0072] S4.1: The three types of effective price comparison nodes—synchronous response channel, structure-driven channel, and regional focus channel—output by layer are categorized and organized to form corresponding path-type price comparison sets.

[0073] Each set maintains the original correspondence between the node's structural position, platform, time slice, and price information, forming a data foundation that supports multi-angle price comparison analysis.

[0074] S4.2: Based on the frequency of change and spatial clustering of structural coding nodes, determine the latest evolution state of the target commodity in the structural matrix, identify the structural change trend, and output the multi-path price comparison set according to the structural change trend.

[0075] Specifically, the current structure matrix of the target product is compared with the structure matrix of the previous period, and the newly added, deleted, or labeled structure coding nodes are extracted, and the changes in position index are recorded. The total number and distribution area of ​​the structure coding nodes that have changed in the structure matrix are counted.

[0076] In the variable structure encoding nodes, it is detected whether they are concentrated in a certain sub-region or multiple adjacent regions of the structure matrix (the structure matrix can be divided into 3x3 grids). If the structural changes show obvious clustering (such as more than 70% of the variable nodes falling within a certain block), it is judged as a local evolution trend; if they are evenly distributed and span multiple sub-regions, it is considered as an overall evolution trend.

[0077] Based on the results of the above two steps, the current structural change trend of the target product can be divided into three categories, for example:

[0078] For the overall rapid evolution type: the number of node changes is large and widely distributed, and the synchronous response channel node set is recommended; for the trend-maintaining type: the number of node changes is small and evenly distributed, and the structure-driven channel node set is recommended; for the local update type: the number of node changes is concentrated in a local area, and the region-focused channel node set is recommended.

[0079] Furthermore, such as Figure 2 As shown, this embodiment also provides an image-driven cross-platform e-commerce product price comparison processing system, including:

[0080] The image acquisition module acquires product images of the target product on multiple platforms, performs structural slicing and partitioning processing on each product image, constructs each region as an encoding node, and generates a structural matrix of the product image.

[0081] The matching and recognition module is used to perform correspondence matching on the structure matrix of product images from multiple platforms, and to identify image-product pairs with comparison potential based on the combination relationship between regional structure and semantic features.

[0082] The feature cross module is used to extract cross feature groups from the historical price series and attributes of products on various platforms based on image product pairs, to perform joint evaluation of price-attribute-structure, and to filter effective price comparison nodes in a hierarchical manner through cross combination judgment rules.

[0083] The recommendation output module is used to generate multi-path price comparison sets based on effective price comparison nodes, and output recommendation results in stages according to the status of the target product.

[0084] This embodiment also provides a computer device applicable to the image-driven cross-platform e-commerce product price comparison processing method, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the image-driven cross-platform e-commerce product price comparison processing method proposed in the above embodiment.

[0085] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0086] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the image-driven cross-platform e-commerce product price comparison processing method proposed in the above embodiments.

[0087] In summary, this invention significantly improves the image comparison accuracy between products across platforms through the construction of a structure matrix and a joint modeling mechanism of structure-semantic features. By dividing nodes into hierarchical levels and evaluating cross-features, it enables deep-level correlation mining between structural changes, price evolution, and attribute information, thereby ensuring higher timeliness and consistency in price comparison references. Simultaneously, relying on multi-path collaborative judgment and node-level hierarchical mechanisms, this invention can dynamically output personalized price comparison recommendations based on the structural evolution trends of different target products. In conclusion, this invention not only improves the accuracy and adaptability of price comparison processing but also ensures the stability and intelligent response capability of the price comparison logic under the condition of complex e-commerce image data.

[0088] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An image-driven cross-platform e-commerce product price comparison processing method, characterized in that: include: Product images of the target product are collected from multiple platforms. Structural slicing and partitioning are performed on each product image, dividing the entire product image into several structural unit regions with fixed spatial sizes. After slicing and partitioning, image feature descriptors are extracted for each structural unit region. Combining the region's spatial location, texture attributes, and edge features, each structural unit region is constructed as a structural encoding node. Each structural encoding node includes a node position index, label semantics, and feature value summary. All structural encoding nodes are combined according to their spatial arrangement in the product image to form a structural matrix corresponding to the target product image. The structure matrix of product images from multiple platforms is matched for correspondence, and image-product pairs with comparison potential are identified based on the combination relationship between region structure and semantic features. Based on image-product pairs, cross-feature groups are extracted from the historical price sequences, attributes and structure matrices of products on various platforms to conduct joint evaluation of price, attributes and structure. Effective price comparison nodes are then selected hierarchically through cross-combination judgment rules. A multi-path price comparison set is generated based on effective price comparison nodes, and recommendation results are output in stages according to the status of the target product. The hierarchical screening of effective price comparison nodes includes: during the construction of multiple collaborative judgment paths, path classification is formed according to the following rules: if the structural coding sequence change of any two factor nodes in the path precedes the change of the commodity price sequence, and the overlap of their time windows exceeds the overlap threshold, the path is marked as a synchronous response channel; if more than 50% of the factor nodes in the path have a label semantic similarity higher than the first threshold, and the price change slope and structural change slope are consistent within a preset tolerance range, the path is marked as a structure-driven channel; if the structural matching degree between factor nodes in the path is lower than the second threshold, but their spatial locations are clustered in adjacent areas of the structure matrix, the path is marked as a regional focus channel; based on the collaborative path type and node level, three types of effective price comparison nodes are output hierarchically from the cross-stable factor map: high-confidence nodes in the synchronous response channel are used for accurate price comparison reference; stable reference nodes in the structure-driven channel are used for trend-based price comparison judgment; and conditional reference nodes in the regional focus channel are used as regional comparison support nodes. The multi-path price comparison set includes: classifying and organizing the three types of effective price comparison nodes—synchronous response channel, structure-driven channel, and regional focus channel—outputted in a hierarchical manner, and forming corresponding path-type price comparison sets respectively; The cross-combination judgment rule includes: constructing a cross-stability factor graph, where each factor node corresponds to a joint stability index in three dimensions; wherein, the three dimensions include time slice overlap, structural coding semantic consistency, and structural alignment density; the cross-stability factor graph is a graph structure constructed based on structural alignment graphs and time series data, where each factor node corresponds to a pair of structural coding nodes and price series associations, and carries a joint stability index in three dimensions; the structural alignment density represents the proportion and distribution concentration of alignment nodes in the entire structural matrix; based on the joint judgment of the three dimensions, each factor node in the graph is assigned the following hierarchical labels: if the structural coding similarity of the node is higher than a preset first threshold, the corresponding structural coding node falls in a continuous region in the structural alignment graph, and the dynamic time regularization distance between the corresponding commodity price curves within a preset time window is less than a preset second threshold, then it is marked as a high-confidence node; if a certain dimension index fluctuates but other dimensions can be compensated, it is marked as a conditional reference node; otherwise, it is marked as a secondary support node; starting from the high-confidence node, multiple collaborative judgment paths are constructed by combining the connected conditional reference nodes.

2. The image-driven cross-platform e-commerce product price comparison processing method as described in claim 1, characterized in that: The corresponding matching of the structure matrix of product images from multiple platforms includes: Add image source tags to the structure matrices corresponding to product images collected from all platforms, categorize the structure matrices according to the source tags, and assign the structure matrices of the same product generated on different platforms to different subsets to ensure that there are no pairings of images from the same source when combining them. For each platform's structure matrix, pair it with structure matrices belonging to the same product on other platforms to generate cross-platform structure matrix pairs. Each pair contains structure matrices from two different source platforms. For each pair of structure matrices, a bidirectional structure matching group is generated by combining them in two directions.

3. The image-driven cross-platform e-commerce product price comparison processing method as described in claim 2, characterized in that: The identified image-item pairs with comparison potential include: For each set of structure matrices, extract the set of structure encoding nodes, and construct the intersection node set and the union node set based on the node position index and label semantics in the structure encoding nodes; Based on the intersection node set and the union node set, calculate the ratio of the number of nodes between the two sets, and use it as the original intersection-union ratio index; The coverage ratio of the node sets in each of the two structure matrices in the entire image structure is calculated as a compensation coefficient and multiplied by the initial cross-union ratio to form the adjusted cross-union ratio. If the adjusted cross-union ratio is higher than the judgment threshold Then the structure matching group will be labeled as an image item pair.

4. The image-driven cross-platform e-commerce product price comparison processing method as described in claim 1, characterized in that: The joint evaluation of price, attribute, and structure includes: In the confirmed image product pairs, extract the structure matrix of the time evolution of products on each platform and the corresponding price data, and construct a structure-price coupling vector of structural changes and price fluctuations. Based on the structure matching group, a structure pair bitmap is constructed and a structure-price coupling vector is bound to form a structure-price fusion feature group.

5. The image-driven cross-platform e-commerce product price comparison processing method as described in claim 1, characterized in that: The step-by-step output of recommendation results based on the target product's status includes: The output of the three types of effective price comparison nodes are divided into node sets according to their respective collaborative path types to construct three types of multi-path price comparison sets; Based on the frequency of change and spatial clustering of structural coding nodes, the latest evolutionary state of the target commodity in the structural matrix is ​​determined, the structural change trend is identified, and a multi-path price comparison set is output according to the structural change trend.

6. An image-driven cross-platform e-commerce product price comparison processing system, based on the image-driven cross-platform e-commerce product price comparison processing method according to any one of claims 1 to 5, characterized in that: Also includes: The image acquisition module acquires product images of the target product from multiple platforms. It performs structural slicing and partitioning on each product image, dividing the entire image into several structural unit regions with fixed spatial sizes. After slicing and partitioning, it extracts image feature descriptors for each structural unit region and, combined with the region's spatial location, texture attributes, and edge features, constructs a structural encoding node for each structural unit region. Each structural encoding node includes a node position index, label semantics, and feature value summary. All structural encoding nodes are combined according to their spatial arrangement in the product image to form a structural matrix corresponding to the target product image. The matching and recognition module is used to perform correspondence matching on the structure matrix of product images from multiple platforms, and to identify image-product pairs with comparison potential based on the combination relationship between regional structure and semantic features. The feature cross module is used to extract cross feature groups from the historical price sequence, attribute and structure matrix of products on various platforms based on image product pairs, to perform joint evaluation of price-attribute-structure, and to filter effective price comparison nodes in a hierarchical manner through cross combination judgment rules. The recommendation output module is used to generate multi-path price comparison sets based on effective price comparison nodes, and output recommendation results in stages according to the status of the target product.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the image-driven cross-platform e-commerce product price comparison processing method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the image-driven cross-platform e-commerce product price comparison processing method according to any one of claims 1 to 5.

Images