Industrial product commodity picture generation method and system based on large model picture processing method

Abstract

The invention provides an industrial product commodity picture generation method and system based on a large model picture processing method. The industrial product commodity picture generation method comprises the steps of obtaining an original industrial product commodity picture; preprocessing the original industrial product commodity picture stored in the object storage server to obtain the preprocessed original industrial product commodity picture; capturing the preprocessed original industrial goods pictures in the object storage server, storing the captured pictures to a kafka queue service, and asynchronously calling a large model to carry out parallel processing on the pictures in the kafka queue service to obtain processed pictures; whether the processed picture meets a preset quality requirement or not is judged, and when the preset quality requirement is met, the currently processed picture is regarded as an available finished picture and stored in an object storage server; and when the picture does not meet the preset requirement, calling the large model to perform repeated processing until the picture meeting the requirement is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of picture generation, and in particular to a method and system for generating pictures of industrial products based on a large model. Background Art

[0002] In the process of digitizing industrial products, the original images provided by manufacturers and suppliers have quality problems and need to be manually cut out, cropped, and beautified. For a large number of industrial products, this process requires a lot of manpower and cost.

[0003] Patent document CN117593083A (application number: 202311615150.9) discloses a product image generation method, device, electronic device and storage medium, including: obtaining an original image of the product and obtaining an original description text, the original description text is used to describe the target style; generating a target description text based on the original description text; inputting the original image and the target description text into a preset Wensheng graph model to obtain a candidate image including the target style; counting the number of candidate images and determining whether the number is greater than or equal to a preset number threshold; if so, displaying multiple candidate images; if not, returning to the step of inputting the original image and the target description text into the preset Wensheng graph model.

[0004] The present invention introduces the image processing service of the large model and encapsulates and refines the large model image processing service into an AI image generation tool through system design, which can automatically cut out, crop, and beautify the image to generate a new finished product image. Summary of the invention

[0005] In view of the defects in the prior art, the purpose of the present invention is to provide a method and system for generating industrial product commodity images based on a large model image processing method.

[0006] According to the present invention, a method for generating an industrial product commodity image based on a large model image processing method comprises:

[0007] Step S1: Obtaining original industrial product images and storing them in an object storage server;

[0008] Step S2: preprocessing the original industrial product commodity image stored in the object storage server to obtain the preprocessed original industrial product commodity image, and storing the preprocessed original industrial product commodity image in the object storage server;

[0009] Step S3: grab the original industrial product images pre-processed in the object storage server, store the captured images in the Kafka queue service, and asynchronously call the large model to process the images in the Kafka queue service in parallel to obtain the processed images;

[0010] Step S4: Determine whether the processed image meets the preset quality requirements. If it does, the currently processed image is regarded as a usable finished image and stored in the object storage server. If it does not meet the preset requirements, the large model is called for repeated processing until a picture that meets the requirements is obtained.

[0011] Preferably, the step S1 comprises: obtaining the original industrial product image by uploading the image or capturing the product data.

[0012] Preferably, step S2 includes: using a large model to identify the quality of the original industrial product commodity image, including: compression ratio, brightness, contrast and clarity; and preprocessing the current original industrial product commodity image according to preset quality parameters to obtain a preprocessed original industrial product commodity image, and storing the preprocessed original industrial product commodity image in an object storage server.

[0013] Preferably, step S3 comprises:

[0014] Step S3.1: Build a Kafka queue service and configure partitions, consumer groups, and consumer instances that meet the preset requirements to achieve parallel consumption;

[0015] Step S3.2: Based on the pre-processed original industrial product images in the object storage server, use the task scheduling tool to periodically trigger the crawling operation;

[0016] Step S3.2: The captured pre-processed original industrial product images are stored in the corresponding partition of the Kafka queue service, and the large model is asynchronously called to perform parallel processing on the images in the Kafka queue service to obtain the processed images.

[0017] Preferably, step S4 comprises: judging through a large model whether the compression ratio, brightness, contrast and clarity of the processed image meet preset quality requirements.

[0018] According to the present invention, a system for generating an industrial product image based on a large model image processing method includes:

[0019] Module M1: Obtain original industrial product images and store them in the object storage server;

[0020] Module M2: preprocessing the original industrial product commodity pictures stored in the object storage server to obtain the preprocessed original industrial product commodity pictures, and storing the preprocessed original industrial product commodity pictures in the object storage server;

[0021] Module M3: Capture the pre-processed original industrial product images in the object storage server, store the captured images in the Kafka queue service, and asynchronously call the large model to process the images in the Kafka queue service in parallel to obtain the processed images;

[0022] Module M4: Determine whether the processed image meets the preset quality requirements. If it does, the currently processed image is regarded as a usable finished image and stored in the object storage server. If it does not meet the preset requirements, the large model is called for repeated processing until a picture that meets the requirements is obtained.

[0023] Preferably, the module M1 includes: obtaining original industrial product images by uploading images or capturing product data.

[0024] Preferably, the module M2 includes: using a large model to identify the quality of the original industrial product commodity picture, including: compression ratio, brightness, contrast and clarity; and preprocessing the current original industrial product commodity picture according to preset quality parameters to obtain a preprocessed original industrial product commodity picture, and storing the preprocessed original industrial product commodity picture in an object storage server.

[0025] Preferably, the module M3 comprises:

[0026] Module M3.1: Build a Kafka queue service and configure partitions, consumer groups, and consumer instances that meet the preset requirements to achieve parallel consumption;

[0027] Module M3.2: Based on the pre-processed original industrial product images in the object storage server, use the task scheduling tool to regularly trigger the crawling operation;

[0028] Module M3.2: Store the captured pre-processed original industrial product images in the corresponding partition of the Kafka queue service, and asynchronously call the large model to perform parallel processing on the images in the Kafka queue service to obtain the processed images.

[0029] Preferably, the module M4 comprises: judging through a large model whether the compression ratio, brightness, contrast and clarity of the processed image meet preset quality requirements.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. The traditional image processing process includes manual cutout, cropping and beautification, which requires a lot of manpower and time, especially when processing massive images, the efficiency is extremely low. The present invention significantly improves the image processing efficiency by integrating the Kafka queue service with the automated processing of large models, and can complete high-quality image processing tasks in batches in a short time;

[0032] 2. Manual processing can easily lead to inconsistent image quality due to differences in personal technical level and working status, affecting user experience and brand image. The present invention uses a deep learning model to ensure the consistency and high quality of image processing, reducing quality fluctuations caused by human factors.

[0033] 3. Existing automated tools often have poor effects when processing complex backgrounds and require manual correction. The large model of the present invention has been trained with a large number of complex background images and has stronger background segmentation and processing capabilities. It can accurately identify and extract the main body of the product, and the effect of processing complex background images is significantly improved.

[0034] 4. Existing image processing tools require manual quality inspection after processing, which is inefficient and prone to omissions. The present invention integrates a quality inspection module that can automatically inspect and provide feedback on processed images to ensure that the clarity, color and composition of the output images meet the requirements, thereby improving the overall quality control level. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0036] Figure 1 The flowchart of the method for generating industrial product images based on large models. DETAILED DESCRIPTION

[0037] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0038] Example 1

[0039] A method and system for generating industrial product commodity images based on a large model provided in accordance with the present invention is intended to solve the image quality problem in the process of digitizing industrial products and reduce manpower and cost investment.

[0040] The method for generating industrial product images based on a large model is as follows: Figure 1 As shown, including:

[0041] Step S1: Obtain the original industrial product image, store it in the object storage server, and record the storage number of the image;

[0042] In this embodiment, multiple formats of picture uploads are supported, and the original picture library is refined through uploading and product data crawling.

[0043] Step S2: preprocessing the original industrial product commodity image stored in the object storage server to obtain the preprocessed original industrial product commodity image, and storing the preprocessed original industrial product commodity image in the object storage server;

[0044] In this embodiment, a large model is used to identify the quality of the image, including compression ratio, brightness, contrast and clarity, and the image that exceeds the quality range is converted and processed according to the configured quality parameters, stored in the object storage server, and the storage number of the image is recorded.

[0045] This embodiment can also automatically capture pre-processed original industrial product images according to user needs.

[0046] Step S3: grab the original industrial product images pre-processed in the object storage server, store the captured images in the Kafka queue service, and asynchronously call the large model to process the images in the Kafka queue service in parallel to obtain the processed images;

[0047] In this embodiment, through the scheduled task, the system automatically captures the image and throws it into the Kafka queue service. By reasonably configuring the partitions, consumer groups and consumer instances, parallel consumption is achieved to improve the throughput and efficiency of image processing. In the parallel consumption of image data, the large model image processing service will be called, but the underlying service of the large model will not be processed. By encapsulating the front-end queue service, the efficiency of overall image processing is improved.

[0048] Step S4: Determine whether the processed image meets the preset quality requirements. If it does, the currently processed image is regarded as a usable finished image and stored in the object storage server. If it does not meet the preset requirements, the large model is called for repeated processing until a picture that meets the requirements is obtained.

[0049] In this embodiment, the image processed by the large model is sent back to the quality detection method, which determines whether the image meets the quality requirements by identifying the image compression ratio, brightness, contrast and clarity. The images that meet the quality standards are stored in the finished product library, and the images that do not meet the quality standards continue to be sent to the image processing service.

[0050] This embodiment automatically stores 100% high-quality pictures in the product picture library, continues to label and grade the incompletely recognized pictures, and retrains the pictures that are not processed by the large model, and finally obtains high-quality finished pictures.

[0051] This embodiment repackages the basic large model image processing service, which can receive massive data in advance and then queue it to the underlying large model through asynchronous consumption. The packaged image service can simplify user operations, receive massive image data, optimize complex image information in batches, and then feed it to the large model for processing after packaging.

[0052] The present invention also provides a large-model-based industrial product commodity image generation system, which can be implemented by executing the process steps of the large-model-based industrial product commodity image generation method, that is, those skilled in the art can understand the large-model-based industrial product commodity image generation method as a preferred implementation of the large-model-based industrial product commodity image generation system.

[0053] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0054] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for generating industrial product images based on a large model image processing method, characterized in that: include: Step S1: Obtaining original industrial product images and storing them in an object storage server; Step S2: preprocessing the original industrial product commodity image stored in the object storage server to obtain the preprocessed original industrial product commodity image, and storing the preprocessed original industrial product commodity image in the object storage server; Step S3: grab the original industrial product images pre-processed in the object storage server, store the captured images in the Kafka queue service, and asynchronously call the large model to process the images in the Kafka queue service in parallel to obtain the processed images; Step S4: Determine whether the processed image meets the preset quality requirements. If it does, the currently processed image is regarded as a usable finished image and stored in the object storage server. If it does not meet the preset requirements, the large model is called for repeated processing until a picture that meets the requirements is obtained.

2. The method for generating industrial product images based on a large model image processing method according to claim 1, characterized in that: The step S1 includes: obtaining the original industrial product image by uploading the image or crawling the product data.

3. The method for generating industrial product images based on a large model image processing method according to claim 1, characterized in that: The step S2 includes: using a large model to identify the quality of the original industrial product commodity image, including: compression ratio, brightness, contrast and clarity; and preprocessing the current original industrial product commodity image according to preset quality parameters to obtain a preprocessed original industrial product commodity image, and storing the preprocessed original industrial product commodity image in an object storage server.

4. The method for generating industrial product images based on a large model image processing method according to claim 1, characterized in that: The step S3 comprises: Step S3.1: Build a Kafka queue service and configure partitions, consumer groups, and consumer instances that meet the preset requirements to achieve parallel consumption; Step S3.2: Based on the pre-processed original industrial product images in the object storage server, use the task scheduling tool to periodically trigger the crawling operation; Step S3.2: The captured pre-processed original industrial product images are stored in the corresponding partition of the Kafka queue service, and the large model is asynchronously called to perform parallel processing on the images in the Kafka queue service to obtain the processed images.

5. The method for generating industrial product images based on a large model image processing method according to claim 1, characterized in that: The step S4 includes: judging through the large model whether the compression ratio, brightness, contrast and clarity of the processed image meet the preset quality requirements.

6. An industrial product image generation system based on a large model image processing method, characterized in that: include: Module M1: Obtain original industrial product images and store them in the object storage server; Module M2: preprocessing the original industrial product commodity pictures stored in the object storage server to obtain the preprocessed original industrial product commodity pictures, and storing the preprocessed original industrial product commodity pictures in the object storage server; Module M3: Capture the pre-processed original industrial product images in the object storage server, store the captured images in the Kafka queue service, and asynchronously call the large model to process the images in the Kafka queue service in parallel to obtain the processed images; Module M4: Determine whether the processed image meets the preset quality requirements. If it does, the currently processed image is regarded as a usable finished image and stored in the object storage server. If it does not meet the preset requirements, the large model is called for repeated processing until a picture that meets the requirements is obtained.

7. The industrial product commodity image generation system based on the large model image processing method according to claim 6 is characterized in that: The module M1 includes: obtaining original industrial product images by uploading images or capturing product data.

8. The industrial product commodity image generation system based on the large model image processing method according to claim 6 is characterized in that: The module M2 includes: using a large model to identify the quality of the original industrial product commodity image, including: compression ratio, brightness, contrast and clarity; and preprocessing the current original industrial product commodity image according to preset quality parameters to obtain a preprocessed original industrial product commodity image, and storing the preprocessed original industrial product commodity image in an object storage server.

9. The industrial product commodity image generation system based on the large model image processing method according to claim 6 is characterized in that: The module M3 comprises: Module M3.1: Build a Kafka queue service and configure partitions, consumer groups, and consumer instances that meet the preset requirements to achieve parallel consumption; Module M3.2: Based on the pre-processed original industrial product images in the object storage server, use the task scheduling tool to regularly trigger the crawling operation; Module M3.2: Store the captured pre-processed original industrial product images in the corresponding partition of the Kafka queue service, and asynchronously call the large model to perform parallel processing on the images in the Kafka queue service to obtain the processed images.

10. The industrial product commodity image generation system based on the large model image processing method according to claim 6 is characterized in that: The module M4 includes: judging through a large model whether the compression ratio, brightness, contrast and clarity of the processed image meet the preset quality requirements.

Images