Cross-border e-commerce electronic data management system based on big data

Abstract

The invention discloses a cross-border e-commerce electronic data management system based on big data, and relates to the technical field of e-commerce data management. Comprising a commodity data acquisition module, a calibrated commodity acquisition module, an image data acquisition module, an image analysis module, an image block vector acquisition module, a vector deviation value acquisition module, a deviation coefficient acquisition module and a delivery identifier generation module. Obtaining a deviation coefficient of the image of the to-be-delivered main image by comparing and analyzing the vector deviation value with a preset deviation threshold value, measuring a difference degree between the to-be-delivered main image and the calibrated commodity main image, comparing and analyzing the deviation coefficient with a preset value, generating a suggested delivery identifier or a non-suggested delivery identifier according to a result, and sending the suggested delivery identifier or the non-suggested delivery identifier to the calibrated commodity main image. According to the method, the exposure rate and the click rate of the commodities are improved, the increase of the sales volume is promoted, higher benefits are brought to merchants, the delivery risk is reduced, the success rate of the commodities on a cross-border e-commerce platform is improved, and meanwhile resource waste caused by unreasonable delivery is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce data management, and particularly to a cross-border e-commerce electronic data management system based on big data. Background Art

[0002] Under the background of the Internet, the scale of domestic e-commerce in China has been continuously increasing, and the e-commerce logistics industry has also achieved certain development. Currently, a logistics development network has gradually taken shape. In the era of big data, the development of cross-border e-commerce will be gradually upgraded. Big data technology has had a profound impact on the management of cross-border e-commerce enterprises, especially in improving the quality of decision-making. With the help of big data technology, cross-border e-commerce enterprises can collect and process a large amount of information, including various data such as customer behavior, market trends, and supply chain status, providing strong support for decision-making.

[0003] However, in the process of product placement, the importance of product main images in attracting consumers' attention is ignored. The visual effects of product main images cannot be evaluated, or only simple manual review is used to evaluate the quality of main images, which cannot ensure the visual consistency of main images with those of well-performing products in the market, affecting the attractiveness and conversion rate of the placed products. Based on this, a cross-border e-commerce electronic data management system based on big data is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a cross-border e-commerce electronic data management system based on big data, which solves the technical problem that in the process of product placement, the importance of product main images in attracting consumers' attention is ignored, the visual effects of product main images cannot be evaluated, or only simple manual review is used to evaluate the quality of main images, which cannot ensure the visual consistency of main images with those of well-performing products in the market, affecting the attractiveness and conversion rate of the placed products.

[0005] The purpose of the present invention can be achieved through the following technical solutions: A cross-border e-commerce electronic data management system based on big data, comprising: A product data acquisition module, which acquires the click volume and sales volume data of each product of the same type as the product to be placed in the area to be placed; A calibrated product acquisition module, which analyzes the click volume and corresponding sales volume data respectively corresponding to each product of the same type as the product to be placed in the area to be placed, and marks the calibrated product of the product to be placed according to the analysis results; An image analysis module, which evenly divides the main image of the calibrated product into multiple image region blocks, analyzes the red, green, and blue channel values of each pixel point in each image region block, and obtains the red, green, and blue channel reference values of each image region block; The tile vector acquisition module vectorizes the red, green, and blue channel reference values of each image region tile to obtain the reference tile vectors corresponding to each image region tile respectively; The vector deviation value acquisition module obtains the red, green, and blue channel values of each region tile of the main image to be placed, performs vectorization processing on them to obtain the to-be-placed tile vectors corresponding to each image region tile in the main image to be placed, and then calculates the vector deviation values between each to-be-placed tile vector and each reference tile vector respectively; The deviation coefficient acquisition module compares and analyzes the vector deviation values with a preset deviation threshold to obtain the deviation coefficient of the main image to be placed; The placement identification generation module compares and analyzes the deviation coefficient with a preset value, and generates a recommended placement identification or a non-recommended placement identification according to the result.

[0006] As a further solution of the present invention: The specific way to mark the calibrated products of the products to be placed is as follows: The sum of the products of the click-through volume DAi and the sales volume value DBi of the same type of products in the region to be placed and the fixed parameter coefficients θ1 and θ2 respectively is marked as the preferred coefficient Xi of each product. If the value of the preferred coefficient Xi is greater than the preset preferred threshold Y1, then the product is marked as the calibrated product of the product to be placed; otherwise, no processing is performed, where i represents different products, i is a positive integer, i≥1, 1 = θ1 + θ2 and θ2>θ1.

[0007] As a further solution of the present invention: The specific way to obtain the red, green, and blue channel reference values of each image region tile is as follows: A1: Randomly select one from multiple image region tiles as the target region tile; A2: Randomly select one from each product main image as the target image; Mark the red channel values corresponding to each pixel point in the target region tile of the target image as He respectively, analyze the discrete value W of the red channel value He, and obtain the red channel calculated value corresponding to the target region tile of the target image, where e represents different pixel points in the target region tile, e = 1, 2, ……, a, where a represents the number of pixel points in the target region tile, a is a positive integer, a≥1; A3: Repeat step A2 to obtain the red channel calculated values Ei corresponding to the target region tiles in each product main image respectively, and analyze the red channel calculated values Ei to obtain the red channel reference values MAg corresponding to each image region tile respectively; A4: Repeat steps A1 - A3 to obtain the green channel reference values MBg and the blue channel reference values MCg corresponding to each image region tile respectively, where g represents different image region tiles, g = 1, 2, ……, v, v represents the number of image region tiles, and v is a positive integer.

[0008] As a further solution of the present invention: The specific method for obtaining the red-channel calculation value corresponding to the target region block of the target image is as follows: Obtain the discrete value W of the red-channel value He corresponding to each pixel point in the target region block. When the discrete value W is less than the preset value Y2, the average value Hp of He is used as the red-channel calculation value corresponding to the target region block of the target image. When W is greater than or equal to the preset value Y2, the values with a large deviation from the average value Hp are deleted in ascending order according to the values of He, and the number r of the deleted values is recorded. At the same time, the discrete value W is recalculated each time a value is deleted until the discrete value W satisfies being less than the preset value Y2. When the deleted number r is greater than the preset value Y3, the average value of the maximum and minimum values in He is used as the red-channel calculation value corresponding to the target region block of the target image. When the deleted number r is less than or equal to the preset value Y3, the average value Hp of He is used as the red-channel calculation value corresponding to the target region block of the target image.

[0009] As a further solution of the present invention: The specific method for obtaining the red-channel reference value corresponding to each image region block is as follows: Obtain the value Ez in the red-channel calculation value Ei that satisfies |Ei - Ep|≥Y5, where z is the number of values in Ei that satisfy the preset screening condition L, and z≥1. Compare the number z with the preset threshold Y4. When the number z is greater than the preset threshold Y4, the average value of Ei is defined as the red-channel reference value corresponding to the target region block. When the number z is less than or equal to the preset threshold Y4, the average value of the maximum and minimum values in Ei is defined as the red-channel reference value corresponding to the target region block.

[0010] As a further solution of the present invention: The specific method for calculating the vector deviation values between each to-be-delivered tile vector and each reference tile vector is as follows: Perform vectorization processing on the red-channel reference value MAg, green-channel reference value MBg, and blue-channel reference value MCg corresponding to each image region block to obtain the reference tile vector Tg(MAg, MBg, MCg) corresponding to each image region block; Input the main image of the to-be-delivered product into the image data acquisition module to obtain the to-be-delivered red-channel value FAg, to-be-delivered green-channel value FBg, and to-be-delivered blue-channel value FCg corresponding to each image region block in the main image of the to-be-delivered product, and perform vectorization processing to obtain the to-be-delivered red tile vector Dg(FAg, FBg, FCg) corresponding to each image region block in the main image of the to-be-delivered product; Through the formula: ; Calculate the vector deviation value Xg between the to-be-delivered block vectors corresponding to each image region block in the to-be-delivered main image and the reference block vectors corresponding to each image region block respectively.

[0011] As a further solution of the present invention: The specific way to obtain the deviation coefficient of the to-be-delivered main image is as follows: Obtain the number f of the vector deviation values Xg greater than the preset deviation threshold Y5, and define the ratio between f and the number v of image region blocks as the deviation coefficient K of the to-be-delivered main image.

[0012] As a further solution of the present invention: The specific way to generate a recommended delivery identifier or a non-recommended delivery identifier according to the result is as follows: When K ≤ Y6, generate a recommended delivery identifier; when K > Y6, generate a non-recommended delivery identifier, where Y6 is a preset value.

[0013] Advantages of the present invention: In the present invention, by inputting the to-be-delivered product main image into the image data acquisition module, obtaining the red, green, and blue channel values of each region block of the to-be-delivered main image and vectorizing them to obtain the to-be-delivered block vectors, then calculating the vector deviation value between it and the reference block vector, and comparing and analyzing the vector deviation value with the preset deviation threshold, the deviation coefficient of the to-be-delivered main image is obtained, which is used to measure the difference degree between the to-be-delivered main image and the calibrated product main image. Comparing and analyzing the deviation coefficient with the preset value, generating a recommended delivery identifier or a non-recommended delivery identifier according to the result, providing a scientific basis for the delivery decision of the product main image, not only improving the exposure rate and click-through rate of the product, but also promoting the growth of sales volume, bringing higher benefits to merchants, reducing the delivery risk, increasing the success rate of the product on the cross-border e-commerce platform, and at the same time reducing the waste of resources caused by unreasonable delivery. Description of the Drawings

[0014] The present invention will be further described below with reference to the accompanying drawings.

[0015] Figure 1 is a schematic diagram of the system framework structure of the present invention; Figure 2 is a schematic diagram of the method framework structure of the present invention. Detailed Embodiments

[0016] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0017] Embodiment 1

[0018] Please refer to Figure 1 - Figure 2 As shown, the present invention is a cross-border e-commerce electronic data management system based on big data, including; A commodity data acquisition module, which is used to acquire the click volume and corresponding sales volume data respectively corresponding to each commodity of the same type as the commodity to be launched in the area to be launched; A calibrated commodity acquisition module analyzes the click volume and corresponding sales volume data respectively corresponding to each commodity of the same type as the commodity to be launched in the area to be launched, and marks the calibrated commodity of the commodity to be launched according to the analysis result. The specific method is as follows: Mark the click volume and corresponding sales volume values respectively corresponding to each commodity of the same type as the commodity to be launched in the area to be launched as DAi and DBi, where i represents different commodities, i is a positive integer, and i≥1; Mark the sum of the products of the click volume DAi and sales volume value DBi of each commodity and the fixed parameter coefficients θ1 and θ2 respectively as the preferred coefficient Xi corresponding to each commodity, that is, through the formula: Xi = DAi×θ1 + DBi×θ2, calculate the preferred coefficient Xi corresponding to each commodity. The specific values of the fixed parameter coefficients θ1 and θ2 are determined by relevant personnel according to actual needs, satisfying 1 = θ1 + θ2 and θ2>θ1; Mark the commodity with the preferred coefficient Xi value greater than the preset preferred threshold Y1 as the calibrated commodity of the commodity to be launched. Otherwise, no processing is performed. The specific value of the preset preferred threshold Y1 is determined by relevant personnel according to actual needs; By comprehensively considering the click volume and sales volume data of the same type of commodities in the area to be launched, and combining the fixed parameter coefficients to calculate the preferred coefficient, it is possible to more accurately mark the calibrated commodities of the commodities to be launched, improve the accuracy of the selection of the calibrated commodities for the commodities to be launched, provide a more reliable basis for subsequent main image analysis and launch decision-making, solve the problem of inaccurate calibration caused by relying only on a single indicator to determine popular commodities, enable the commodities to be launched to have a more targeted reference object, and improve the accuracy of market positioning.

[0019] An image data acquisition module acquires the image data of the main images of the calibrated commodities respectively corresponding to the area to be launched; An image analysis module is used to segment the image data of the main images of the calibrated commodities respectively corresponding to the area to be launched, obtain the image region blocks of each main image, obtain the channel values corresponding to the red, green, and blue channels of each pixel point in each image region block through the RGB color model and analyze them, and then obtain the red, green, and blue channel reference values of each image region block; Through the image segmentation unit, the main product images corresponding to each calibrated product are evenly segmented into multiple image region blocks; Through the image region data analysis unit, the channel values corresponding to the red, green, and blue channels respectively corresponding to each pixel point in the image region blocks of each main product image are analyzed, and then the red channel reference value, green channel reference value, and blue channel reference value corresponding to each image region block are obtained. The specific method is as follows: A1: Randomly select one from multiple image region blocks as the target region block; A2: Randomly select one from each main product image as the target image; For each pixel point in the target region block of the obtained target image, through the RGB color model, the channel values corresponding to the red, green, and blue channels respectively corresponding to each pixel point in the target region block are obtained; Select the red channel from the red, green, and blue channels as the analysis channel, obtain the red channel values corresponding to each pixel point in the target region block, and mark them as He respectively, where e represents different pixel points in the target region block, e = 1, 2,..., a, where a represents the number of pixel points in the target region block, a is a positive integer, a ≥ 1; Through the formula: , obtain the discrete value W of the red channel value He corresponding to each pixel point in the target region block, where Hj is any one of He, Hp is the mean value of He, a ≥ j ≥ 1; And analyze the discrete value W. When the discrete value W is less than the preset value Y2, then take the mean value Hp of He as the red channel calculation value E1 corresponding to the target region block of the target image. When W is greater than or equal to the preset value Y2, then delete the values with larger deviations from the mean value Hp among them in ascending order according to the values of He, and record the number of deleted values as r. At the same time, recalculate the discrete value W each time a value is deleted until the discrete value W satisfies being less than the preset value Y2. When the number of deletions r is greater than the preset value Y3, then take the mean value of the maximum and minimum values in He as the red channel calculation value E1 corresponding to the target region block of the target image. When the number of deletions r is less than or equal to the preset value Y3, then take the mean value Hp of He as the red channel calculation value E1 corresponding to the target region block of the target image, where the specific values of Y2 and Y3 are set by relevant personnel according to specific application scenarios and requirements; A3: Repeat step A2 to calculate and obtain the red channel calculation values Ei corresponding to the target region blocks in each main product image; Obtain the value Ez in the red channel calculation value Ei that satisfies the preset screening condition G, where z is the number of values in Ei that satisfy the preset screening condition L, z ≥ 1. Compare the quantity z with the preset threshold Y4. When the quantity z is greater than the preset threshold Y4, it indicates that the number of values in Ei that satisfy the preset screening condition L is relatively large, and the mean value of Ei is representative. Furthermore, define the mean value of Ei as the red channel reference value MA1 corresponding to the target region block. When the quantity z is less than or equal to the preset threshold Y4, it indicates that the number of values in Ei that satisfy the preset screening condition L is relatively small, and the mean value of Ei is not representative. Furthermore, define the mean value of the maximum and minimum values in Ei as the red channel reference value MA1 corresponding to the target region block, that is, MA1 = (MA min + MA max ) / 2, where MA min and MA max are the maximum and minimum values in Ei respectively; Here, the preset condition L is specifically: |Ei - Ep| ≥ Y5, where both Y4 and Y5 are preset values, and the specific values of Y4 and Y5 are determined by relevant personnel according to actual needs; A4: Repeat steps A1 - A3 to obtain the red channel reference values MAg corresponding to each image region block respectively, where g represents different image region blocks, g = 1, 2, ……, v, and v represents the number of image region blocks, and v is a positive integer; A5: In the same way as obtaining the red channel reference values MAg corresponding to each image region block respectively, analyze the green channel values and blue channel values corresponding to each pixel point in the image region block of each product main image, and then obtain the green channel reference value MBg and blue channel reference value MCg corresponding to each image region block respectively; The tile vector acquisition module performs vectorization processing on the red channel reference value MAg, green channel reference value MBg, and blue channel reference value MCg corresponding to each image region block respectively to obtain the reference tile vector Tg(MAg, MBg, MCg) corresponding to each image region block; The vector deviation value acquisition module inputs the to-be-delivered main image of the product to be delivered into the image data acquisition module, and then obtains the to-be-delivered red channel value FAg, to-be-delivered green - red channel value FBg, and to-be-delivered red - blue channel value FCg corresponding to each image region block in the to-be-delivered main image of the product to be delivered. Perform vectorization processing on the to-be-delivered red channel value FAg, to-be-delivered green - red channel value FBg, and to-be-delivered red - blue channel value FCg corresponding to each image region block in the to-be-delivered main image of the product to be delivered to obtain the to-be-delivered tile vector Dg(FAg, FBg, FCg) corresponding to each image region block in the to-be-delivered main image; Through the formula: ; Calculate the vector deviation value Xg between the to-be-delivered red map block vectors Dg(FAg, FBg, FCg) corresponding to each image region block in the to-be-delivered main map image and the reference map block vectors Tg(MAg, MBg, MCg) corresponding to each image region block respectively; The deviation coefficient acquisition module analyzes and compares the vector deviation value with a preset deviation threshold Y5, and obtains the deviation coefficient of the to-be-delivered main map image according to the analysis and comparison result. The specific method is as follows: Obtain the number f of vector deviation values Xg greater than the preset deviation threshold Y5, and define the ratio of f to the number v of image region blocks as the deviation coefficient K of the to-be-delivered main map image. The specific value of the preset deviation threshold Y5 is determined by relevant personnel according to actual needs and market research; The delivery flag generation module compares and analyzes the deviation coefficient K with a preset value Y6, and generates a recommended delivery flag or a non-recommended delivery flag according to the analysis result. The specific method is as follows: When K ≤ Y6, it indicates that the main map of the to-be-delivered product is visually similar enough to the main maps of the same type of products of the to-be-delivered product, and this main map can be accepted as the image for market delivery. Furthermore, a recommended delivery flag is generated; When K > Y6, it indicates that there are significant visual differences between the main map of the to-be-delivered product and the main maps of the same type of products of the to-be-delivered product. It is necessary to adjust or optimize the main map to better attract the target market, and a non-recommended delivery flag is generated; By performing image segmentation on the product main map and through complex red, green, and blue channel value analysis, the reference values and vectors of each image region block are obtained. The red, green, and blue channel reference values of each image region block are vectorized to obtain the reference map block vectors corresponding to each image region block respectively. The to-be-delivered product main map image is input into the image data acquisition module to obtain the red, green, and blue channel values of each region block of the to-be-delivered main map image and vectorize them to obtain the to-be-delivered map block vectors. Then, calculate the vector deviation value between it and the reference map block vector, and compare and analyze the vector deviation value with the preset deviation threshold to obtain the deviation coefficient of the to-be-delivered main map image, which is used to measure the difference degree between the to-be-delivered main map and the calibrated product main map. Compare and analyze the deviation coefficient with the preset value, and generate a recommended delivery flag or a non-recommended delivery flag according to the result, providing a scientific basis for the delivery decision of the product main map. It not only improves the exposure rate and click-through rate of the product, but also promotes the growth of sales volume, brings higher benefits to merchants, reduces the delivery risk, improves the success rate of products on cross-border e-commerce platforms, and at the same time reduces the waste of resources caused by unreasonable delivery.

[0020] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0021] As described above, it is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights. Similarity analysis was carried out on the main pictures before launch, and launch suggestions were given in advance, enabling merchants to adjust and optimize in a timely manner, avoiding unnecessary losses, ensuring the visual similarity between the product main pictures and those of well-performing products in the market, thereby enhancing the attractiveness and click-through rate of the main pictures, ultimately promoting the growth of sales, ensuring that the product main pictures to be launched have a high enough visual similarity to the best-selling products in the target market to increase the attractiveness and sales volume of the products.

Claims

1. A cross-border e-commerce electronic data management system based on big data, characterized by: include: The product data acquisition module obtains the click volume and sales volume data of each product of the same type as the product to be launched in the area to be launched; The module for acquiring the marked products analyzes the click volume and sales volume of each product of the same type as the product to be released in the area to be released, and marks the marked products of the product to be released according to the analysis results; The image analysis module evenly divides the main image of the calibrated product into multiple image area blocks, analyzes the red, green, and blue channel values ​​of each pixel in each image area block, and obtains the red, green, and blue channel reference values ​​of each image area block; The image block vector acquisition module performs vector processing on the red, green and blue channel reference values ​​of each image area block to obtain the reference image block vector corresponding to each image area block; The vector deviation value acquisition module obtains the red, green and blue channel values ​​of each area block of the main image to be placed and performs vector processing to obtain the image block vectors to be placed corresponding to each image area block in the main image to be placed, and then calculates the vector deviation values ​​between each image block vector to be placed and each reference image block vector; The deviation coefficient acquisition module compares and analyzes the vector deviation value with the preset deviation threshold to obtain the deviation coefficient of the main image to be placed; The delivery mark generation module compares and analyzes the deviation coefficient with the preset value, and generates a recommended delivery mark or a non-recommended delivery mark based on the result.

2. The cross-border e-commerce electronic data management system based on big data according to claim 1 is characterized in that: The specific method for marking the designated products of the products to be put on the market is as follows: The sum of the products of the click volume DAi and sales volume DBi of the same type of products in the area to be launched and the fixed parameter coefficients θ1 and θ2 are marked as the preferred coefficient Xi of each product. If the value of the preferred coefficient Xi is greater than the preset preferred threshold Y1, the product is marked as the calibration product for the products to be launched. Otherwise, no processing is done, where i refers to different products, i is a positive integer, i≥1, 1=θ1+θ2 and θ2>θ1.

3. The cross-border e-commerce electronic data management system based on big data according to claim 2 is characterized in that: The specific method of obtaining the red, green and blue channel reference values ​​of each image area block is: A1: Randomly select one of multiple image area blocks as the target area block; A2: Randomly select one of the main images of each product as the target image; The red channel values ​​corresponding to each pixel in the target area block of the target image are marked as He, and the discrete value W of the red channel value He is analyzed to obtain the red channel calculation value corresponding to the target area block of the target image, where e refers to different pixels in the target area block, e=1, 2, ..., a, where a refers to the number of pixels in the target area block, a is a positive integer, a≥1; A3: Repeat step A2 to obtain the red channel calculation value Ei corresponding to each target area block in each product main image, and analyze the red channel calculation value Ei to obtain the red channel reference value MAg corresponding to each image area block; A4: Repeat steps A1-A3 to obtain the green channel reference value MBg and the blue channel reference value MCg corresponding to each image area block, where g refers to different image area blocks, g=1, 2, ..., v, v refers to the number of image area blocks, and v is a positive integer.

4. The cross-border e-commerce electronic data management system based on big data according to claim 3 is characterized in that: The specific method of obtaining the red channel calculation value corresponding to the target area block of the target image is: The discrete value W of the red channel value He corresponding to each pixel point in the target area block is obtained. When the discrete value W is less than the preset value Y2, the mean value Hp of He is used as the red channel calculation value corresponding to the target area block of the target image. When W is greater than or equal to the preset value Y2, the values ​​with larger deviations from the mean value Hp are deleted in order from small to large according to the values ​​of He, and the number of deleted values ​​is recorded as r. At the same time, the discrete value W is recalculated at each deletion until the discrete value W satisfies the preset value Y2. When the number of deletions r is greater than the preset value Y3, the mean of the maximum and minimum values ​​in He is used as the red channel calculation value corresponding to the target area block of the target image. When the number of deletions r is less than or equal to the preset value Y3, the mean value Hp of He is used as the red channel calculation value corresponding to the target area block of the target image.

5. The cross-border e-commerce electronic data management system based on big data according to claim 4 is characterized in that: The specific method of obtaining the red channel reference value corresponding to each image area block is as follows: Get the value Ez that satisfies |Ei-Ep|≥Y5 in the red channel calculation value Ei, where z is the number of values ​​in Ei that meet the preset screening condition L, z≥1, and compare the number z with the preset threshold value Y4. When the number z is greater than the preset threshold value Y4, the mean value of Ei is defined as the red channel reference value corresponding to the target area block. When the number z is less than or equal to the preset threshold value Y4, the mean value of the maximum and minimum values ​​in Ei is defined as the red channel reference value corresponding to the target area block.

6. The cross-border e-commerce electronic data management system based on big data according to claim 5 is characterized in that: The specific method of calculating the vector deviation value between each to-be-placed tile vector and each reference tile vector is as follows: Vectorization is performed on the red channel reference value MAg, the green channel reference value MBg and the blue channel reference value MCg corresponding to each image area block, so as to obtain the reference block vector Tg (MAg, MBg, MCg) corresponding to each image area block; Input the main image of the product to be placed into the image data acquisition module, obtain the red channel value FAg to be placed, the green channel value FBg to be placed and the blue channel value FCg to be placed corresponding to each image area block in the main image of the product to be placed, and perform vector processing to obtain the red image block vector Dg (FAg, FBg, FCg) to be placed corresponding to each image area block in the main image of the product to be placed; By formula: ; Calculate the vector deviation value Xg between the block vectors to be placed corresponding to each image area block in the main image to be placed and the reference block vectors corresponding to each image area block.

7. The cross-border e-commerce electronic data management system based on big data according to claim 6 is characterized in that: The specific method of obtaining the deviation coefficient of the main image to be placed is: The number f of each vector deviation value Xg that is greater than a preset deviation threshold value Y5 is obtained, and the ratio between f and the number of image area blocks v is defined as the deviation coefficient K of the main image to be placed.

8. The cross-border e-commerce electronic data management system based on big data according to claim 7 is characterized in that: The specific method of generating a recommended placement flag or a non-recommended placement flag based on the results is as follows: When K≤Y6, a recommended placement flag is generated. When K>Y6, a non-recommended placement flag is generated. Y6 is a preset value.

Images