Rim analysis information acquisition method, device, equipment, storage medium and program product

Abstract

The application relates to a wheel rim analysis information acquisition method and device, computer equipment and a readable storage medium. The method comprises the following steps: inputting wheel rim image data and a first prompt word into a visual language model which has been pre-trained, to obtain constraint information output by the visual language model; the first prompt word is used for guiding the visual language model to analyze and infer the constraint information based on the wheel rim image data; a second prompt word is generated based on the constraint information; the second prompt word is used for guiding the visual language model to analyze and infer analysis information of the wheel rim based on the wheel rim image data; and the wheel rim image data and the second prompt word are input into the visual language model, to obtain the analysis information of the wheel rim output by the visual language model. The method can improve the accuracy of the wheel rim analysis information.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for acquiring wheel rim analysis information. Background Technology

[0002] The wheel rim refers to the outer core component of a car wheel, which supports the tire and is also an important part of the vehicle's exterior decoration.

[0003] In existing technologies, most methods directly analyze and infer information from wheel rim image data to obtain analytical information about the wheel rim. However, these wheel rim analysis methods suffer from poor accuracy. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for acquiring wheel rim analysis information that can improve the accuracy of wheel rim analysis, in order to address the above-mentioned technical problems.

[0005] Firstly, this application provides a method for obtaining wheel rim analysis information, including:

[0006] The wheel rim image data and the first prompt word are input into a pre-trained visual language model to obtain the constraint information output by the visual language model; the first prompt word is used to guide the visual language model to analyze and reason about the constraint information based on the wheel rim image data.

[0007] A second prompt word is generated based on constraint information; the second prompt word is used to guide the visual language model to infer the analytical information of the rim based on the rim image data.

[0008] The wheel rim image data and the second cue word are input into the visual language model to obtain the wheel rim analysis information output by the visual language model.

[0009] In one embodiment, after obtaining the analysis information of the wheel rim output by the visual language model, the method further includes: performing verification on the analysis information to obtain the verification result of the analysis information; if the verification result is that the verification fails, performing adjustment processing on the second prompt word, and inputting the wheel rim image data and the adjusted second prompt word into the visual language model to obtain new analysis information of the wheel rim output by the visual language model.

[0010] In one embodiment, the analysis information includes the color names and area percentages of each major color on the rim. Major colors are those whose area percentage on the rim is greater than a preset threshold. The area percentage indicates the proportion of the area occupied by each major color on the rim to the front area of ​​the rim. The analysis information is validated to obtain validation results, including: performing an integrity check on the analysis information to obtain an integrity check result; performing a comparison check on the number of color names and the number of major colors to obtain a consistency check result; performing a numerical range check on the area percentage of each major color to obtain a numerical range check result; if the integrity check result, consistency check result, and numerical range check result are all passed, the validation result is determined to be passed; otherwise, the validation result is determined to be failed.

[0011] In one embodiment, the analysis information includes the process type and area ratio of each major color on the rim. The major color is the color whose area ratio on the rim is greater than a preset threshold. The process type is used to indicate the processing process type corresponding to the area where each major color is located on the rim. The area ratio is used to indicate the proportion of the area where each major color is located on the rim to the front area of ​​the rim. After obtaining the analysis information of the rim output by the visual language model, the method further includes: determining the processing process resource consumption information corresponding to the area where each major color is located on the rim based on the process type and area ratio of each major color on the rim.

[0012] In one embodiment, based on the process type and area proportion of each major color on the rim, the processing resource consumption information corresponding to the area where each major color is located on the rim is determined, including: for each major color, determining the actual area of ​​the major color based on the area proportion of the major color, and obtaining the processing resource consumption model corresponding to the major color based on the process type of the major color; the actual area is the area of ​​the region where the major color is located on the rim; inputting the actual area into the processing resource consumption model to obtain the processing resource consumption information corresponding to the region where the major color is located on the rim.

[0013] In one embodiment, the analysis information includes the color names of the main colors on the rim. After obtaining the analysis information of the rim output by the visual language model, the method further includes: determining the color values ​​of the main colors on the rim in the target color space; determining the standard color names of the main colors based on the color values; and updating the color names of the main colors to the standard color names.

[0014] In one embodiment, determining the standard color name of each primary color based on color values ​​includes: generating a third prompt word based on the color values, the third prompt word being used to guide the color naming model to output the standard color name of each primary color on the rim based on the color values; and inputting the third prompt word and color values ​​into the color naming model to obtain the standard color name of each primary color.

[0015] In one embodiment, after obtaining the analysis information of the wheel rim output by the visual language model, the method further includes: acquiring a standard processing technology library, which includes multiple standard processing technology types; determining the standard processing technology type corresponding to the processing technology type of each major color based on the standard processing technology library; and updating the processing technology type of each major color to the respective standard processing technology type.

[0016] Secondly, this application also provides a device for acquiring wheel rim analysis information, comprising:

[0017] The first execution module is used to input the rim image data and the first prompt word into the pre-trained visual language model to obtain the constraint information output by the visual language model; the first prompt word is used to guide the visual language model to analyze and reason about the constraint information based on the rim image data.

[0018] The generation module is used to generate a second prompt word based on constraint information; the second prompt word is used to guide the visual language model to infer the analysis information of the rim based on the rim image data.

[0019] The second execution module is used to input the rim image data and the second prompt word into the visual language model to obtain the rim analysis information output by the visual language model.

[0020] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in any of the embodiments of the first aspect above.

[0021] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the embodiments of the first aspect above.

[0022] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in any of the embodiments of the first aspect above.

[0023] The aforementioned method, apparatus, computer equipment, computer-readable storage medium, and computer program product for acquiring wheel rim analysis information first inputs wheel rim image data and a first prompt word into a visual language model. The visual language model then infers constraint information based on the wheel rim image data. Next, a second prompt word is generated based on the constraint information, and the second prompt word and wheel rim image data are input into the visual language model again. The visual language model then infers analysis information about the wheel rim based on the wheel rim image data. Since the analysis and inference of the wheel rim image data analysis information is carried out under the guidance of the second prompt word generated based on the constraint information, it can effectively avoid inference bias and misidentification problems caused by unconstrained analysis, thereby improving the accuracy of wheel rim analysis information. Attached Figure Description

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

[0025] Figure 1 This is a schematic diagram of an optional process for obtaining wheel rim analysis information in one embodiment;

[0026] Figure 2 This is a schematic diagram of an optional process for obtaining new analytical information about the wheel rim in one embodiment;

[0027] Figure 3 This is a schematic diagram of an optional process for obtaining verification results in one embodiment;

[0028] Figure 4 This is an optional flowchart illustrating one embodiment for determining processing resource consumption information;

[0029] Figure 5 This is a schematic diagram of an optional process for updating color names in one embodiment;

[0030] Figure 6 This is a schematic diagram of an optional process for determining a standard color name in one embodiment;

[0031] Figure 7 This is a schematic diagram of an optional process flow for obtaining the adjusted process type in one embodiment;

[0032] Figure 8 This is a schematic diagram of an optional process for acquiring rim image data in one embodiment;

[0033] Figure 9A schematic diagram of an optional structure of the rim analysis information acquisition device in one embodiment;

[0034] Figure 10 This is a schematic diagram of an optional internal structure of a computer device in one embodiment. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0036] The terms "first," "second," etc., used in this application may be used to describe various elements, but these elements are not limited by these terms. These terms are used only to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0037] In one embodiment, such as Figure 1 As shown, a method for obtaining wheel rim analysis information is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and is implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0038] Step S101: Input the rim image data and the first prompt word into the pre-trained visual language model to obtain the constraint information output by the visual language model.

[0039] The rim image data refers to the rim image encoded data adapted to the input format of the visual language model. The first prompt can be a preset text command or a text command input by the user in real time. It can be used to guide the visual language model to analyze and reason about constraint information based on the rim image data. The constraint information can be the number of main colors of the rim, the effective analysis area of ​​the rim, etc. The main color is the color on the rim whose color area ratio is greater than a preset threshold. The preset threshold can be set in advance according to actual needs, such as 10%, 15%, etc. The effective analysis area is the area on the rim after removing invalid analysis areas such as bolt holes, center cap cutouts, and background debris. The first prompt can be "Please analyze the rim image data and output the number of main colors of the rim" or "Please analyze the rim image data and output the effective analysis area of ​​the rim". The first prompt can also include limiting rules, such as "Please analyze the rim image data, output the number of main colors of the rim, and ignore transition colors, noise, edge effects and contour lines, only count the main color, and output an integer". The visual language model is a multimodal model used to analyze constraint information.

[0040] Specifically, when analyzing a wheel rim, the terminal first acquires the rim image data and a preset first prompt word. Then, it uses the rim image data and the first prompt word as input to a visual language model. The visual language model then analyzes and infers from the rim image data according to the guidance of the first prompt word. Furthermore, the terminal can extract constraint information from the text output by the visual language model using regular expressions or structured parsing.

[0041] Step S102: Generate a second prompt word based on the constraint information.

[0042] The second prompt can be a pre-defined text instruction used to guide the visual language model to analyze and infer the rim's analytical information based on the rim image data, using constraint information as a guide. This second prompt could be something like, "The rim image data contains N main colors; please analyze and infer the analytical information for each color region," or "The effective analysis region in the rim image data is X; please analyze and infer the analytical information for each color region within the effective analysis region," where N is the number of main colors and X is the effective analysis region. The second prompt can also guide the visual language model to output the rim's analytical information in a pre-defined structured data format, such as, "The rim image data contains N main colors; please analyze and infer the analytical information for each color region and output it in a pre-defined structured data format," or "The effective analysis region in the rim image data is X; please analyze and infer the analytical information for each color region within the effective analysis region and output it in a pre-defined structured data format."

[0043] Specifically, after obtaining the number of primary colors, the terminal can retrieve a pre-stored second prompt word template, fill in the constraint information into the quantity placeholder of the second prompt word template, and generate a second prompt word carrying the constraint. For example, if the number of primary colors is 3, the generated second prompt word could be "There are 3 primary colors in the rim image data. Please analyze the rim analysis information."

[0044] Step S103: Input the rim image data and the second prompt word into the visual language model to obtain the rim analysis information output by the visual language model.

[0045] Specifically, after obtaining the second cue word, the rim image data and the second cue word can be used as input to the visual language model. The visual language model can then analyze and reason about the rim image data according to the guidance of the second cue word to obtain analytical information about the rim. When the constraint information is the number of primary colors, the analysis of the rim image data is performed under the constraint of the number of primary colors. Compared to traditional methods, this allows the visual language model to divide color regions based on a defined number of colors, avoiding the visual language model from identifying regions associated with primary color regions (such as lighting changes, metallic reflections, and color transition areas related to primary color regions) as separate color regions. This avoids messy color region division and improves the accuracy of rim analysis information. Similarly, when the constraint information is the effective analysis area, the analysis of the rim image data is performed under the constraint of the effective analysis area. Compared to traditional methods, this allows the visual language model to analyze based on a defined effective analysis area, avoiding the visual language model from analyzing invalid analysis areas in the rim image data (such as bolt holes, central cutouts, and surrounding background debris), thus improving the accuracy of rim analysis information.

[0046] The above-mentioned method for obtaining wheel rim analysis information first inputs wheel rim image data and a first cue word into a visual language model. The visual language model then infers constraint information based on the wheel rim image data. Next, a second cue word is generated based on the constraint information, and the second cue word and wheel rim image data are input into the visual language model again. The visual language model then infers the wheel rim analysis information based on the wheel rim image data. Since the analysis and inference of the wheel rim image data analysis information is carried out under the guidance of the second cue word generated based on the constraint information, the reasoning bias and misidentification problems caused by unconstrained analysis can be effectively avoided, thereby improving the accuracy of wheel rim analysis information.

[0047] In an optional embodiment of this application, the analysis information may include the color name, process type, and area percentage of each major color on the rim. The process type may be used to indicate the processing process type corresponding to the area of ​​each major color on the rim, such as spraying, electroplating, high-gloss injection molding, etc.; the area percentage may be used to indicate the percentage of the area of ​​each major color on the rim to the front area of ​​the rim.

[0048] The second prompt can also be used to guide the visual language model to analyze and infer the color name, process type, and area ratio of each main color on the rim based on the rim image data, as analytical information. For example, if there are 3 main colors, the generated second prompt could be: "There are 3 main colors in the rim image data. Please analyze each color area: 1. Color name (e.g., matte black, bright silver, champagne gold, etc.); 2. Process type corresponding to the color area (e.g., spraying, electroplating, high-gloss injection molding, others); 3. Area ratio of the color area, and return it in JSON format (e.g., [{"color": "matte black", "process": "spraying", "area_ratio": 0.6}, {"color": "bright silver", "process": "electroplating", "area_ratio": 0.3}, {"color": "high-gloss black", "process": "high-gloss injection molding", "area_ratio": 0.1}]").

[0049] In one embodiment, such as Figure 2 As shown, the wheel rim analysis information acquisition method provided in this application may further include the following after step S103:

[0050] Step S201: Perform verification on the analysis information to obtain the verification result of the analysis information.

[0051] The verification result can include verification passed and verification failed. Specifically, the terminal can perform multiple types of verification on the analyzed information, such as integrity verification, accuracy verification, reasonableness verification, and consistency verification. If all types of verification pass, the verification result is determined to be verification passed; if any type of verification fails, the verification result is determined to be verification failed.

[0052] In step S202, if the verification result is that the verification fails, the second prompt word is adjusted, and the rim image data and the adjusted second prompt word are input into the visual language model to obtain the new rim analysis information output by the visual language model.

[0053] If the verification fails, the terminal can adjust the second prompt word according to the type of failure. After adjusting the second prompt word, the terminal can use the rim image data and the adjusted second prompt word as input to the visual language model. The visual language model can then analyze the rim image data according to the guidance of the adjusted second prompt word to obtain new rim analysis information. The terminal can then perform the above verification again on the new rim analysis information until the verification result is successful. If the verification result is still unsuccessful after multiple adjustments to the second prompt word, a "pending review" mark can be output for further review by technical personnel.

[0054] In this embodiment, the analysis information is validated to obtain the validation result. If the validation result is that the validation fails, the second prompt word is adjusted. The rim image data and the adjusted second prompt word are then input into the visual language model to obtain new rim analysis information output by the visual language model. By adjusting the second prompt word, the analysis information output by the visual language model is optimized, thereby improving the accuracy of the analysis information.

[0055] In one embodiment, such as Figure 3 As shown, the analysis information includes the color names and area proportions of each major color on the rim. Step S201 may further include:

[0056] Step S301: Perform an integrity check on the analysis information to obtain the integrity check result of the analysis information.

[0057] The integrity verification result of the analysis information is used to indicate whether the content contained in the analysis information is complete. Specifically, after the terminal receives the analysis information output by the visual language model, it can iterate through the records corresponding to each major color in the analysis information to see if each record corresponding to each major color contains the color name and area percentage. If each record corresponding to each major color contains the color name and area percentage, the integrity verification result is determined to be successful; otherwise, the integrity verification result is determined to be unsuccessful.

[0058] Furthermore, in an optional embodiment, the analysis information may include the color name, process type, and area percentage of each major color on the rim. After receiving the analysis information output by the visual language model, the terminal can iterate through the records corresponding to each major color in the analysis information to see if they all contain the color name, process type, and area percentage. If each record corresponding to each major color contains the color name, process type, and area percentage, the integrity verification result can be determined to be that the integrity verification passed; otherwise, the integrity verification result can be determined to be that the integrity verification failed.

[0059] Step S302: Perform a comparison and verification on the number of color names and the number of main colors for each main color to obtain the consistency verification result of the analysis information.

[0060] The consistency check result of the analyzed information can be used to indicate whether the number of color names of each primary color in the analyzed information is consistent with the number of primary colors. Specifically, after receiving the analyzed information output by the visual language model, the terminal can determine the number of color names of each primary color based on the analyzed information, and compare the number of color names with the number of primary colors. If the two values ​​are the same, the consistency check result is determined to be a pass; otherwise, the consistency check result is determined to be a fail.

[0061] Step S303: Perform a numerical range check on the area ratio of each main color to obtain the numerical range check result of the analysis information.

[0062] The numerical range verification results of the analysis information can be used to indicate whether the area proportion of each major color on the rim is reasonable. Specifically, after receiving the analysis information output by the visual language model, the terminal can iterate through the analysis information to see if the area proportion corresponding to each major color is within the range of 0 to 1. If it is, the numerical range verification result is determined to be passed; otherwise, the numerical range verification result is determined to be failed.

[0063] In step S304, if the integrity check result, the consistency check result, and the numerical range check result are all passed, then the check result is determined to be passed; otherwise, the check result is determined to be failed.

[0064] After obtaining the integrity check result, consistency check result, and value range check result, if any one of the three results fails the check, the check result is determined to be a failed check. If all three results pass the check, the check result is determined to be a passed check.

[0065] Furthermore, the terminal can adjust the second prompt word based on the type of verification failure. For example, if the integrity verification fails, the second prompt word can be adjusted by adding a description requiring each major color to include its color name, process type, and area percentage; if the consistency verification fails, the second prompt word can be adjusted by adding a description requiring the number of color names for each major color to match the number of major colors; if the numerical range verification fails, the second prompt word can be adjusted by adding a description requiring a numerical range for the area percentage.

[0066] In this embodiment, the integrity of the analysis information is checked to obtain the integrity check result. The number of color names and the number of main colors are compared and checked to obtain the consistency check result. The area ratio of each main color is checked to obtain the numerical range check result. If the integrity check result, consistency check result, and numerical range check result are all passed, the check result is determined to be passed. Otherwise, the check result is determined to be failed. By checking the analysis information, the accuracy of the analysis information is improved.

[0067] In one embodiment, the analysis information includes the process type and area ratio of each major color on the rim. After obtaining the analysis information of the rim output by the visual language model, the method further includes: determining the processing resource consumption information corresponding to the area where each major color on the rim is located, based on the process type and area ratio of each major color on the rim.

[0068] Processing resource consumption information is used to indicate the resources required to complete the surface processing steps for the areas where the main colors are located. Specifically, the terminal can determine the processing resource consumption information per unit area for each main color's processing type based on the processing type of each main color, and then determine the processing resource consumption information corresponding to the areas on the rim where each main color is located based on the area proportion of each main color and the processing resource consumption information per unit area for each main color's processing type.

[0069] In this embodiment, based on the process type and area ratio of each major color on the rim, the processing resource consumption information corresponding to the area where each major color is located on the rim is determined, which can improve the efficiency and accuracy of estimating the resources required for rim surface treatment.

[0070] In one embodiment, such as Figure 4 As shown, based on the processing type and area proportion of each main color on the rim, the processing resource consumption information corresponding to the area where each main color is located on the rim further includes:

[0071] Step S401: For each main color, determine the actual area of ​​the main color based on the area ratio of the main color, and obtain the processing technology resource consumption model corresponding to the main color based on the process type of the main color.

[0072] The actual area of ​​the main color refers to the area on the rim where the main color is located, while the processing resource consumption model can be used to determine the processing resource consumption information corresponding to the area where the main color is located.

[0073] Specifically, the terminal can first obtain the total area of ​​the front surface of the rim. For each main color, the area percentage of that main color can be multiplied by the total area of ​​the front surface of the rim to obtain the actual area of ​​that main color. The actual area of ​​the main color can be expressed as:

[0074] ;

[0075] in, The actual area of ​​the main color. This represents the total area of ​​the front surface of the wheel rim. This represents the area percentage of the primary color. When determining the actual area of ​​the primary color, the terminal can also retrieve a matching processing resource consumption model from a preset model library based on the processing type of that primary color.

[0076] Step S402: Input the actual area into the processing technology resource consumption model to obtain the processing technology resource consumption information corresponding to the area where the main color on the rim is located.

[0077] Specifically, the terminal can input the calculated actual area into the processing resource consumption model to obtain the processing resource consumption information corresponding to the area on the rim where the main color is located. For example, in the case of a spray painting process, the processing resource consumption model can be expressed as:

[0078] ;

[0079] in, This provides information on the processing resource consumption corresponding to the area where the main color is located. The actual area of ​​the main color. The resources required per unit area for the spraying process. Resources required for fixed steps in the spraying process (e.g., resources required for pretreatment and tooling).

[0080] In this embodiment, for each main color, the actual area of ​​the main color is determined based on its area ratio. Then, based on the processing type of the main color, a processing resource consumption model corresponding to that main color is obtained. The actual area is the area of ​​the main color on the rim. The actual area is input into the processing resource consumption model to obtain the processing resource consumption information corresponding to the area of ​​the main color on the rim. By matching the corresponding processing resource consumption model to calculate the processing resource consumption information, the processing resource consumption information corresponding to each main color can be directly obtained, improving the efficiency and accuracy of estimating the resources required for rim surface treatment.

[0081] In one embodiment, such as Figure 5As shown, the analysis information includes the color names of the main colors on the rim. The rim analysis information acquisition method provided in this application may further include the following after step S103:

[0082] Step S501: Determine the color values ​​of each major color on the rim in the target color space.

[0083] The color values ​​in the target color space are parameters used to characterize the actual color of each primary color. The target color space can be the RGB color space, and the color values ​​in the target color space can be RGB color values. Specifically, the terminal can determine the region where each primary color is located in the rim image data, and determine the average RGB color value of multiple pixels in that region as the color value of each primary color in the target color space.

[0084] Step S502: Determine the standard color name of each main color based on the color value.

[0085] Standard color names are standardized color names. Specifically, a terminal can use a color value-standard color name mapping list to determine the standard color name corresponding to each primary color based on the color value. The color value-standard color name mapping list can be used to indicate the mapping relationship between different color values ​​and different standard color names.

[0086] Step S503: Update the color names of each primary color to the standard color names.

[0087] After obtaining the standard color names, the terminal can update the color names of each primary color to the standard color names. This update can be done by replacing the original color names of each primary color in the analysis information with the corresponding standard color names.

[0088] In this embodiment, the color values ​​of each main color on the rim in the target color space are determined, the standard color name of each main color is determined based on the color value, and the color name of each main color is updated to the standard color name. By matching the color value with the standard color name and updating it, color naming can be unified, avoiding the difference in expression of the same color with different names, and improving the accuracy of the analysis information.

[0089] Furthermore, such as Figure 6 As shown, step S502 may further include:

[0090] Step S601: Generate a third prompt word based on color values.

[0091] In this embodiment, the third prompt word is used to guide the color naming model to output the standard color names of each major color on the rim based on the color value.

[0092] Specifically, the terminal can retrieve a pre-stored third prompt word template, fill in the color values ​​corresponding to each major color into the quantity placeholder of the third prompt word template, and generate the corresponding third prompt word, such as "Please identify the standard color name corresponding to the following color values, color values: ({R}, {G}, {B})".

[0093] The third prompt can also contain constraint information, such as "Please identify the standard color name corresponding to the following color values: ({R}, {G}, {B}), return the most suitable standard color name, mark the metallic properties for metallic colors, and add dark and light prefixes according to brightness".

[0094] Step S602: Input the third prompt word and color value into the color naming model to obtain the standard color name of each main color.

[0095] The color naming model is a model with color semantic mapping capabilities. It can reuse the aforementioned visual language model or use an independent color recognition model.

[0096] Specifically, after obtaining the third prompt word, the third prompt word and the color value can be used as the model input of the color naming model. The color naming model can then analyze the color value according to the guidance of the third prompt word to obtain the standard color name of each major color.

[0097] In this embodiment, a third prompt word is generated based on the color value. The third prompt word is used to guide the color naming model to output the standard color name of each main color on the rim based on the color value. The third prompt word and the color value are input into the color naming model to obtain the standard color name of each main color. By guiding the color naming model to output the standard color name based on the color value through the third prompt word, the color naming can be unified, the difference in the expression of the same color with different names can be avoided, and the accuracy of the analysis information can be improved.

[0098] In one embodiment, such as Figure 7 As shown, the analysis information includes the process type of each major color on the rim. The rim analysis information acquisition method provided in this application may further include the following after step S103:

[0099] Step S701: Obtain the standard processing technology library.

[0100] The standard processing technology library is a pre-built dataset containing standardized process types. This library includes multiple standard process types. For example, it might contain four standard process types: spraying, electroplating, high-gloss injection molding, and others. Each standard process type corresponds to a unified naming convention. Specifically, the terminal can retrieve the pre-built standard processing technology library from its local storage or the server-side database.

[0101] Step S702: Determine the standard process type corresponding to the process type of each major color based on the standard processing process library.

[0102] Specifically, the terminal can iterate through the analysis information to find the process types corresponding to each major color, and match each process type with the standard process types in the standard processing process library. If the match is successful, the matched standard process type is determined as the standard process type corresponding to the process type. If the match fails, the standard process type of the process type is determined as another standard process type, and a check mark is added.

[0103] Step S703: Update the process type of each major color to the standard process type.

[0104] After obtaining the standard process type, the terminal can update the process type of each major color to the standard process type. This update can be done by replacing the original process type of each major color in the analysis information with the corresponding standard process type.

[0105] In this embodiment, a standard processing technology library is obtained, and the standard process type corresponding to the process type of each major color is determined based on the standard processing technology library. The process type of each major color is updated to the standard process type. By standardizing the process type through the standard processing technology library, the naming of process types can be unified, the difference in expression can be eliminated, and the accuracy of the analysis information can be improved.

[0106] Furthermore, after step S103, the method may include: performing normalization processing on the area proportion of each main color on the rim to obtain the standard area proportion of each main color; and updating the area proportion of each main color to the standard area proportion.

[0107] Normalization can be used to correct deviations in the sum of the area proportions of each primary color. Specifically, the terminal can first calculate the sum of the area proportions of all primary colors. If the sum of the area proportions is not equal to 1, then normalization is performed on the area proportion of each primary color to obtain a standard area proportion. The original area proportions of each primary color in the analysis information are then replaced with the corresponding standard area proportions to ensure that the sum of the area proportions of all primary colors is 1.

[0108] In this embodiment, the area proportions of each major color on the rim are normalized to obtain standard area proportions. The area proportions of each major color are then updated to these standard area proportions. By normalizing the area proportions of each major color, the estimation bias of the visual language model can be corrected, improving the accuracy of the analyzed information. In embodiments with higher area accuracy requirements, the color names and process types output by the visual language model can be used as prior information. A target detection model or image segmentation model can be called to perform pixel-level precise segmentation of each color region in the rim image. Based on the segmentation results, the number of pixels in each region is counted and converted to obtain a more accurate area proportion, replacing the estimation results of the visual language model and further improving the accuracy of the analyzed information.

[0109] In one embodiment, such as Figure 8 As shown, step S101 may further include:

[0110] Step S801: Obtain the original image of the wheel rim.

[0111] In this embodiment, the original rim image is a binary front view image of the rim to be analyzed. Specifically, the terminal can acquire the original rim image by taking a picture with an image acquisition device, or it can obtain the original rim image from local storage or the business system. When the original rim image contains a lot of background, the rim area in the original rim image can be cropped first, or the region of interest can be limited to reduce the interference of background color.

[0112] Step S802: Perform image encoding processing on the original rim image to obtain rim image data.

[0113] Image encoding processing is used to convert the raw rim image into a format compatible with the visual language model. Specifically, the terminal can perform encoding conversion on the raw rim image, converting the binary front view image into Base64 encoded format to obtain rim image data that conforms to the visual language model interface specification.

[0114] In this embodiment, the original rim image is acquired, and the original rim image is image encoded to obtain rim image data. By acquiring the original rim image and performing encoding processing, rim image data that meets the input requirements of the visual language model can be obtained, which facilitates the subsequent processing of the visual language model.

[0115] In one embodiment, this application also provides a method for obtaining wheel rim analysis information. The overall method flow is as follows:

[0116] Step 1: Image encoding processing;

[0117] The original image of the wheel rim is acquired, and image encoding processing is performed on the original wheel rim image to obtain wheel rim image data. This image encoding processing can be to convert the original wheel rim image into Base64 encoding format.

[0118] Step 2: Output constraint information;

[0119] The first cue word is obtained by inputting the wheel rim image data and the first cue word into a pre-trained visual language model, which then outputs constraint information. This first cue word guides the visual language model to analyze and infer constraint information based on the wheel rim image data.

[0120] Step 3: Output the analysis information of the wheel rim;

[0121] A second cue word is generated based on the constraint information. The rim image data and the second cue word are then input into the visual language model to obtain the rim analysis information output by the visual language model. The second cue word guides the visual language model to analyze and infer the rim analysis information based on the rim image data. The analysis information includes the color name, process type, and area percentage of each main color on the rim. The process type indicates the processing type corresponding to the area of ​​each main color on the rim, and the area percentage indicates the proportion of the area of ​​each main color on the rim to the front area of ​​the rim.

[0122] Step 4: Verification of the analyzed information;

[0123] The system performs an integrity check on the analyzed information to obtain the integrity check result. It also performs a comparison check on the number of color names and the number of main colors to obtain the consistency check result. Finally, it performs a numerical range check on the area proportion of each main color to obtain the numerical range check result. If the integrity check, consistency check, and numerical range check all pass, the check result is considered passed; otherwise, it is considered failed. If the check result is failed, the second prompt word is adjusted, and the rim image data and the adjusted second prompt word are input into the visual language model to obtain new rim analysis information output by the visual language model.

[0124] Step 5: Standardize color names;

[0125] Determine the color values ​​of each main color on the rim in the target color space. Generate a third prompt word based on the color values. The third prompt word is used to guide the color naming model to output the standard color names of each main color on the rim based on the color values. Input the third prompt word and color values ​​into the color naming model to obtain the standard color names of each main color. Update the color names of each main color to the standard color names.

[0126] Step Six: Standardization of Process Types;

[0127] Obtain the standard processing technology library, which includes multiple standard process types. Based on the standard processing technology library, determine the standard process type corresponding to the process type of each major color, and update the process type of each major color to the corresponding standard process type.

[0128] Step 7: Area percentage normalization processing;

[0129] Normalize the area proportion of each major color on the rim to obtain the standard area proportion of each major color; update the area of ​​each major color to the standard area proportion.

[0130] Step 8: Determine the resource consumption information for the processing technology;

[0131] For each primary color, the actual area of ​​the primary color is determined based on its area ratio, and the processing resource consumption model corresponding to the processing type of the primary color is obtained. The actual area is the area of ​​the primary color on the rim. The actual area is input into the processing resource consumption model to obtain the processing resource consumption information corresponding to the area of ​​the primary color on the rim.

[0132] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0133] Based on the same inventive concept, this application also provides a rim analysis information acquisition device for implementing the rim analysis information acquisition method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations of one or more rim analysis information acquisition device embodiments provided below can be found in the limitations of the rim analysis information acquisition method described above, and will not be repeated here.

[0134] In one exemplary embodiment, such as Figure 9 As shown, a wheel rim analysis information acquisition device is provided, comprising: a first execution module 901, a generation module 902, and a second execution module 903, wherein:

[0135] The first execution module 901 is used to input the rim image data and the first prompt word into a pre-trained visual language model to obtain the constraint information output by the visual language model; the first prompt word is used to guide the visual language model to analyze and reason about the constraint information based on the rim image data.

[0136] The generation module 902 is used to generate a second prompt word based on constraint information; the second prompt word is used to guide the visual language model to infer the analysis information of the rim based on the rim image data.

[0137] The second execution module 903 is used to input the rim image data and the second prompt word into the visual language model to obtain the rim analysis information output by the visual language model.

[0138] In one embodiment, the second execution module 903 is further configured to perform verification on the analysis information to obtain the verification result of the analysis information; if the verification result is that the verification fails, the second prompt word is adjusted, and the rim image data and the adjusted second prompt word are input into the visual language model to obtain the new rim analysis information output by the visual language model.

[0139] In one embodiment, the analysis information includes the color names and area percentages of each major color on the rim. Major colors are those whose area percentage on the rim is greater than a preset threshold. The area percentage indicates the proportion of the area occupied by each major color on the rim to the front area of ​​the rim. The second execution module 903 is specifically used to perform integrity checks on the analysis information to obtain integrity check results; to perform a comparison check on the number of color names and the number of major colors to obtain consistency check results; and to perform a numerical range check on the area percentages of each major color to obtain numerical range check results. If the integrity check result, consistency check result, and numerical range check result are all passed, the check result is determined to be passed; otherwise, the check result is determined to be failed.

[0140] In one embodiment, the analysis information includes the process type and area ratio of each major color on the rim. The major color is the color whose area ratio on the rim is greater than a preset threshold. The process type is used to indicate the processing process type corresponding to the area where each major color is located on the rim. The area ratio is used to indicate the proportion of the area where each major color is located on the rim to the front area of ​​the rim. The second execution module 903 is also used to determine the processing process resource consumption information corresponding to the area where each major color is located on the rim based on the process type and area ratio of each major color on the rim.

[0141] In one embodiment, the second execution module 903 is specifically used to determine the actual area of ​​each main color based on the area ratio of the main color, and to obtain the processing technology resource consumption model corresponding to the main color based on the process type of the main color; the actual area is the area of ​​the main color on the rim; the actual area is input into the processing technology resource consumption model to obtain the processing technology resource consumption information corresponding to the area of ​​the main color on the rim.

[0142] In one embodiment, the analysis information includes the color names of each major color on the rim. The second execution module 903 is further configured to determine the color values ​​of each major color on the rim in the target color space; determine the standard color names of each major color based on the color values; and update the color names of each major color to the standard color names.

[0143] In one embodiment, the second execution module 903 is specifically used to generate a third prompt word based on the color value. The third prompt word is used to guide the color naming model to output the standard color names of each main color on the rim based on the color value. The third prompt word and the color value are input into the color naming model to obtain the standard color names of each main color.

[0144] In one embodiment, the analysis information includes the process type of each major color on the rim. The process type is used to indicate the processing process type corresponding to the area where each major color is located on the rim. The second execution module 903 is also used to obtain a standard processing process library, which includes multiple standard process types; determine the standard process type corresponding to the process type of each major color based on the standard processing process library; and update the process type of each major color to the standard process type.

[0145] Each module in the aforementioned wheel rim analysis information acquisition device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0146] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 10 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. 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 input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for acquiring wheel rim analysis information. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0147] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0148] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0149] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0150] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0151] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0152] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0153] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0154] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for obtaining wheel rim analysis information, characterized in that, The method includes: The wheel rim image data and the first prompt word are input into a pre-trained visual language model to obtain constraint information output by the visual language model; the first prompt word is used to guide the visual language model to analyze and infer the constraint information based on the wheel rim image data. A second prompt word is generated based on the constraint information; the second prompt word is used to guide the visual language model to analyze and infer the analysis information of the rim based on the rim image data; The wheel rim image data and the second prompt word are input into the visual language model to obtain the analysis information of the wheel rim output by the visual language model.

2. The method according to claim 1, characterized in that, After obtaining the analysis information of the wheel rim output by the visual language model, the method further includes: The analysis information is validated to obtain the validation result. If the verification result is that the verification fails, the second prompt word is adjusted, and the rim image data and the adjusted second prompt word are input into the visual language model to obtain new analysis information of the rim output by the visual language model.

3. The method according to claim 2, characterized in that, The analysis information includes the color names and area percentages of each major color on the wheel rim. The major colors are those whose area percentage on the wheel rim is greater than a preset threshold. The area percentage indicates the proportion of the area occupied by each major color on the wheel rim to the front area of ​​the wheel rim. The analysis information is validated to obtain validation results, including: Perform an integrity check on the analysis information to obtain the integrity check result of the analysis information; A comparison and verification is performed on the number of color names of each of the main colors and the number of the main colors themselves to obtain the consistency verification result of the analysis information; Perform a numerical range check on the area percentage of each of the main colors to obtain the numerical range check result of the analysis information; If the integrity check result, the consistency check result, and the numerical range check result are all passed, the check result is determined to be passed; otherwise, the check result is determined to be failed.

4. The method according to claim 1, characterized in that, The analysis information includes the processing type and area percentage of each major color on the rim. The major colors are those whose area percentage on the rim is greater than a preset threshold. The processing type indicates the processing type corresponding to the area of ​​each major color on the rim. The area percentage indicates the percentage of the area of ​​each major color on the rim to the front area of ​​the rim. After obtaining the analysis information of the rim output by the visual language model, the method further includes: Based on the process type and area ratio of each of the main colors on the rim, determine the processing resource consumption information corresponding to the area where each of the main colors on the rim is located.

5. The method according to claim 4, characterized in that, The step of determining the processing resource consumption information corresponding to the area of ​​each of the main colors on the rim based on the processing type and area proportion of each of the main colors on the rim includes: For each of the main colors, the actual area of ​​the main color is determined according to its area ratio, and the processing technology resource consumption model corresponding to the main color is obtained according to its processing type; the actual area is the area of ​​the region where the main color is located on the wheel rim. By inputting the actual area into the processing resource consumption model, processing resource consumption information corresponding to the area where the main color is located on the rim is obtained.

6. The method according to claim 4, characterized in that, The analysis information includes the color names of the main colors on the wheel rim. After obtaining the analysis information of the wheel rim output by the visual language model, the method further includes: Determine the color values ​​of each of the main colors on the rim in the target color space; Determine the standard color name for each of the primary colors based on the color values; Update the color names of each of the primary colors to the standard color names.

7. The method according to claim 6, characterized in that, The step of determining the standard color name of each of the primary colors based on the color value includes: A third prompt word is generated based on the color value. The third prompt word is used to guide the color naming model to output the standard color name of each main color on the rim based on the color value. The third prompt word and the color value are input into the color naming model to obtain the standard color name of each of the main colors.

8. The method according to claim 4, characterized in that, After obtaining the analysis information of the wheel rim output by the visual language model, the method further includes: Obtain a standard processing technology library, which includes multiple standard process types; Based on the standard processing technology library, determine the standard process type corresponding to the process type of each of the main colors; Update the process type of each of the main colors to the standard process type.

9. A device for acquiring wheel rim analysis information, characterized in that, The device includes: The first execution module is used to input the wheel rim image data and the first prompt word into a pre-trained visual language model to obtain the constraint information output by the visual language model; the first prompt word is used to guide the visual language model to analyze and reason about the constraint information based on the wheel rim image data. The generation module is used to generate a second prompt word based on the constraint information; the second prompt word is used to guide the visual language model to analyze and infer the analysis information of the rim based on the rim image data; The second execution module is used to input the wheel rim image data and the second prompt word into the visual language model to obtain the analysis information of the wheel rim output by the visual language model.

10. 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 method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

Images