Cotton quality sorting control method based on multi-technology cooperation

Abstract

The invention provides a cotton quality sorting control method based on multi-technology collaboration, and relates to the technical field of detection control, and the method comprises the steps: obtaining a detection equipment set through cotton quality evaluation index set mapping scheduling, and carrying out the limited collection of the minimum sample number; controlling the standardized sampling tool to randomly sample the cotton sorting pile; performing deviation sample identification and elimination based on visual analysis; carrying out sample pooling treatment to obtain an integrated test sample; generating a weight distribution melon score integrated test sample, and synchronizing the test sample to a detection equipment set for multi-technology collaborative cotton quality sorting; a comprehensive evaluation function is introduced to quantify a cotton quality evaluation result set, and a cotton quality coefficient is matched with a cotton grade number table to obtain a cotton grading result, so that the technical problem of poor sorting effect caused by low sorting efficiency and insufficient cotton quality control accuracy and comprehensiveness in a cotton quality sorting process is solved; the technical effect of improving the accuracy, efficiency and comprehensiveness of cotton sorting is achieved.

Description

Technical Field

[0001] This invention relates to the field of detection and control technology, specifically to a cotton quality sorting and control method based on multi-technology collaboration. Background Technology

[0002] In recent years, the cotton industry has maintained steady development, and the overall quality of cotton has remained stable. However, some regions and farmers still have some quality problems in the cotton production process, such as the use of banned pesticides and high impurity content. These problems directly affect the overall quality and market competitiveness of cotton.

[0003] Currently, cotton quality inspection mainly employs a variety of techniques, including instrumental testing, visual inspection, and laboratory testing. Instrumental testing, such as high-performance liquid chromatography (HPLC) and gas chromatography (GC), is used to detect pesticide residues, heavy metals, and other harmful substances in cotton. Visual inspection uses high-definition cameras and image recognition technology to assess the appearance quality of cotton, such as color and impurities. Laboratory testing uses chemical analysis methods to detect physical indicators such as the length and strength of cotton fibers. While existing testing technologies can ensure cotton quality to a certain extent, their efficiency and accuracy still need improvement, especially in the large-scale, fast-moving cotton market, where more efficient and accurate testing technologies are needed to ensure the quality of each batch of cotton.

[0004] In summary, existing technologies suffer from low sorting efficiency and insufficient precision and comprehensiveness in cotton quality control during the cotton quality sorting process, resulting in poor sorting performance and impacting subsequent processing, production, and sales. Summary of the Invention

[0005] This application provides a cotton quality sorting and control method based on multi-technology collaboration, which is used to address the technical problems of low sorting efficiency, insufficient accuracy and comprehensiveness in cotton quality control in the existing cotton quality sorting process, resulting in poor sorting effect and affecting subsequent processing and sales.

[0006] In view of the above problems, this application provides a cotton quality sorting and control method based on multi-technology collaboration.

[0007] The cotton quality sorting and control method based on multi-technology collaboration provided in this application includes: A set of cotton quality evaluation indicators is preset, and a standardized testing equipment set is obtained by mapping and scheduling according to the cotton quality evaluation indicator set; the standardized testing equipment set is subjected to minimum sample quantity limit collection to obtain a sample limit set; a standardized sampling tool is controlled to randomly sample the cotton sorting pile to obtain a sample cotton set; based on visual analysis, the sample cotton set is used to identify and remove biased samples to obtain a reliable sample set; the reliable sample set is subjected to sample pooling processing to obtain integrated test samples; the integrated test samples are distributed according to the sample limit set with weights, and the distribution results are synchronized to the standardized testing equipment set; the standardized testing equipment set is controlled to perform multi-technology collaborative cotton quality sorting to obtain a cotton quality evaluation result set; a comprehensive evaluation function is introduced to quantify the cotton quality evaluation result set to obtain a cotton quality coefficient, and the cotton quality coefficient is used to match a cotton grade table to obtain the cotton grading result of the cotton sorting pile.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: The cotton quality sorting and control method based on multi-technology collaboration provided in this application involves: pre-setting a set of cotton quality evaluation indicators; obtaining a standardized testing equipment set by mapping and scheduling the cotton quality evaluation indicator set; collecting samples from the standardized testing equipment set with a minimum sample size limit to obtain a sample limit set; controlling a standardized sampling tool to randomly sample from the cotton sorting pile to obtain a sample cotton set; identifying and removing biased samples from the sample cotton set based on visual analysis to obtain a reliable sample set; performing sample pooling processing on the reliable sample set to obtain integrated test samples; and generating a weight allocation to distribute the integrated test samples according to the sample limit set, and then sharing the distribution results with... The process proceeds to the standardized testing equipment set, which is then controlled to perform multi-technology collaborative cotton quality sorting to obtain a cotton quality evaluation result set. A comprehensive evaluation function is introduced to quantify the cotton quality evaluation result set, obtaining a cotton quality coefficient. This coefficient is then matched with a cotton grade table to obtain the cotton grading results of the sorted cotton pile. This addresses the technical problems of low sorting efficiency, insufficient accuracy and comprehensiveness in cotton quality control during the cotton quality sorting process, leading to poor sorting results and affecting subsequent processing and sales. The process achieves the technical effect of improving the accuracy, efficiency, and comprehensiveness of cotton sorting, thereby ensuring the reliability of subsequent processing and sales. Attached Figure Description

[0009] Figure 1 This application provides a flowchart of a cotton quality sorting and control method based on multi-technology collaboration.

[0010] Figure 2This application provides a flowchart illustrating the random sampling process of cotton sorting piles in a cotton quality sorting and control method based on multi-technology collaboration. Detailed Implementation

[0011] 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 merely illustrative and not intended to limit the scope of this application.

[0012] Examples, such as Figure 1 As shown, this application provides a cotton quality sorting and control method based on multi-technology collaboration, the method comprising: Step S100: Preset a set of cotton quality evaluation indicators, and obtain a set of standardized testing equipment by mapping and scheduling according to the cotton quality evaluation indicator set.

[0013] Furthermore, the cotton quality evaluation index set includes fiber length, fiber strength, micronaire value, color, and impurity content.

[0014] The cotton quality sorting and control method based on multi-technology synergy proposed in this application aims to achieve rapid and accurate detection and control of cotton quality by comprehensively utilizing various advanced technologies. Sorting refers to the process of conducting quality testing and evaluation on the cotton to be tested, and then classifying and grading the cotton according to certain standards and rules based on the test results. This process involves not only the testing of physical properties of the cotton, such as fiber length and strength, but may also include the evaluation of chemical properties, such as impurity content and pigment content. Multi-technology synergy refers to using multiple testing technologies or methods for cotton sampling and testing to improve the accuracy, efficiency, and comprehensiveness of the testing.

[0015] Specifically, in the cotton quality evaluation process, it is first necessary to determine a series of key quality indicators. These indicators can comprehensively reflect the physical and chemical properties of cotton, thereby ensuring the accuracy and reliability of the evaluation results. All determined quality indicators together constitute the cotton quality evaluation indicator set. In this embodiment, the cotton quality evaluation indicator set includes fiber length, fiber strength, micronaire value, color, and impurity content. Among these parameters, fiber length is a crucial parameter for measuring cotton fiber quality, directly impacting spinning performance and yarn quality. Longer fibers are generally more suitable for spinning high-count yarns, imparting better luster and hand feel to textiles. Fiber strength reflects the cotton fiber's resistance to breakage and is an important indicator for assessing yarn durability and the strength of the final product. High-strength fibers can produce stronger and more durable textiles. Micronaire value is a comprehensive indicator of cotton fiber fineness and maturity, significantly affecting yarn uniformity and the hand feel of the final product. An appropriate micronaire value ensures yarn uniformity and fabric comfort. Color is an important aspect of evaluating cotton's appearance quality, including color uniformity and whiteness. High-quality cotton should have uniform color and good whiteness to meet the color requirements of different textiles. Impurity content refers to the amount of non-fibrous substances in cotton, including dust, sand, and short fibers. A low impurity content can improve spinning efficiency, reduce yarn breakage during spinning, and improve the appearance and hand feel of the final product.

[0016] After determining the set of cotton quality evaluation indicators, a corresponding set of standardized testing equipment is mapped and scheduled according to these indicators. This equipment should possess high precision, high stability, and high reliability to ensure the accuracy and consistency of the test results. The mapping process involves associating each quality evaluation indicator with a standardized testing device capable of measuring that indicator. For example, fiber length corresponds to a fiber length tester, fiber strength to a tensile strength meter, micronaire value to a micronaire value tester, color to a colorimeter, and impurity content to an impurity analyzer. Scheduling involves rationally arranging the order and schedule of use of each testing device based on factors such as the testing process, equipment performance, and testing time. This ensures the efficiency and orderliness of the testing process, avoiding conflicts and waiting times between devices. Through the mapping and scheduling process, a standardized set of testing equipment containing all necessary testing equipment is ultimately obtained. The equipment in this set should conform to international or industry standards and undergo regular calibration and maintenance to ensure the accuracy and reliability of its measurement results.

[0017] Step S200: Collect samples from the standardized testing equipment set with a minimum sample quantity limit to obtain a sample limit set.

[0018] Optionally, firstly, it's necessary to clarify the purpose of the test and the required level of accuracy. Different tests and purposes may have different sample size requirements. For example, tests requiring high-precision results may require more samples. Refer to relevant statistical literature, industry standards, or equipment manufacturer recommendations to understand commonly used methods and formulas for determining the minimum sample size. These methods may include confidence interval estimation, sample size calculation formulas, etc. The minimum sample size refers to the minimum number of samples required for effective testing. Simultaneously, the equipment's detection accuracy also affects the determination of the minimum sample size. High-precision equipment may obtain reliable results with a smaller sample size, while low-precision equipment may require more samples to ensure accuracy. Therefore, based on the testing purpose, accuracy requirements, and equipment accuracy, use appropriate statistical methods or formulas to calculate the minimum sample size for each test. This number should ensure that, at a given confidence level, the error range of the test results is within an acceptable range. Furthermore, organize the calculated minimum sample sizes for each test into a table or list to form a sample limit set. This set will serve as a guide for subsequent sample collection and testing. In addition, it should be noted that in actual operation, factors such as sample availability, testing time, and cost need to be considered. Sometimes, it may be necessary to make appropriate adjustments to the minimum sample size while ensuring testing accuracy.

[0019] By obtaining a set of sample limit values, unnecessary sample waste and testing time can be avoided while ensuring the accuracy of test results. This helps to improve testing efficiency, shorten the testing cycle, and thus provide stronger support for cotton quality evaluation more quickly.

[0020] Step S300: Control the standardized sampling tool to randomly sample the cotton sorting pile to obtain a sample cotton set.

[0021] For example, using standardized sampling tools can significantly reduce human bias and improve the accuracy and representativeness of sampling. Standardized sampling tools can be sampling probes or sampling boxes. Sampling probes are suitable for scenarios where samples need to be obtained from deep within cotton piles. The probes should be designed to be long enough to penetrate cotton layers at different depths, and their tips should be smooth to avoid damaging the cotton fibers during sampling. Sampling boxes, on the other hand, are suitable for scenarios where random sampling is required from the surface or sides of cotton piles. Sampling boxes should have sufficient volume to hold a sufficient number of cotton samples, and their openings should be designed for easy operation and to minimize the introduction of external impurities.

[0022] Specifically, during sampling, ensure that sampling tools are clean, undamaged, and meet pre-defined standardization requirements. Mark the tools to track their usage history and maintenance status. Prepare sufficient clean containers or bags to store samples taken from the cotton sorting pile. Based on the size, shape, and distribution of the cotton sorting pile, use random number tables, grid division, or other randomization methods to determine the sampling points, ensuring they are evenly distributed throughout the pile to reflect overall quality. Next, following predetermined operating procedures, use standardized sampling tools to take samples at each sampling point. For sampling probes, insert them into the cotton pile to a predetermined depth, rotate, and pull them out to obtain representative samples. For sampling boxes, place them above the sampling points, open the opening, and gently shake or tap the cotton pile to allow samples to fall into the box. Finally, carefully place the samples taken from each sampling point into clean containers or bags and seal them immediately to prevent contamination or moisture loss. In addition, each sample needs to be uniquely identified and its origin location (such as the sampling point number or coordinates) needs to be recorded. A label should be affixed to the sample container or packaging bag, indicating the sample number, sampling date, sampling personnel, and other information. This completes the entire sampling process.

[0023] All collected samples are merged into a single cotton sample set, ready for subsequent biased sample identification, pooling processing, and multi-technology collaborative detection. By using standardized sampling tools and implementing rigorous sampling procedures, the impact of human bias on cotton sorting results can be minimized, improving the accuracy and reliability of sorting.

[0024] Step S400: Based on visual analysis, identify and remove biased samples from the sample cotton set to obtain a reliable sample set.

[0025] Furthermore, high-resolution cameras or image acquisition devices are used to photograph the sample cotton collection to obtain clear image data. The acquired images are preprocessed, including noise reduction, contrast enhancement, and brightness adjustment, to improve image quality and facilitate subsequent analysis. The color distribution and uniformity of the sample cotton are analyzed to identify samples with color abnormalities (such as excessive bleaching or uneven dyeing). Simultaneously, texture analysis techniques are used to extract the texture features of the sample cotton to identify samples with abnormal fiber structures (such as impurities or fiber breakage). The shape and size of the sample cotton are also analyzed to identify samples with morphological abnormalities (such as irregularly shaped cotton or short-staple cotton).

[0026] Next, a model for identifying deviation samples is built based on the extracted visual features. This model can be built using machine learning algorithms (such as classifiers and clustering algorithms), which require training with a large amount of labeled sample data. The preprocessed sample cotton images are input into the identification model, which automatically analyzes the features in the image and compares them with preset deviation standards to identify deviation samples. For more accurate judgment, manual verification can be performed on top of automated identification; this is not limited here. Then, based on the identification results, deviation samples are physically removed from the sample cotton set to ensure that subsequent detected samples are reliable. The time, sample number, and identification result of each identification operation are recorded for subsequent tracking and analysis. Simultaneously, the identification model and preprocessing process can be continuously optimized based on the identification results and actual conditions to improve the accuracy and efficiency of identification.

[0027] By using visual analysis technology to identify and remove biased samples from the cotton sample set, we can ensure that the obtained reliable sample set has a high degree of representativeness and consistency, providing a solid foundation for subsequent cotton quality sorting.

[0028] Step S500: Perform sample pooling processing on the reliable sample set to obtain integrated test samples.

[0029] Pooling a reliable sample set to obtain an integrated test sample is a process of merging multiple individual samples into a unified sample set. This process aims to eliminate or reduce differences between individual samples, improving the stability and representativeness of subsequent testing. Specifically, ensure that all samples in the reliable sample set have undergone rigorous screening and validation, and do not contain any biased samples. Prepare sufficient containers or equipment for mixing and storing the integrated test sample. Mix all samples in the reliable sample set uniformly together; mixing can be achieved through manual stirring, mechanical mixing, or automated equipment. During mixing, care should be taken to avoid over-stirring to prevent damage to the cotton fibers or the introduction of new impurities.

[0030] Furthermore, to ensure the homogeneity of the integrated test samples, additional measures may be needed to further mix the samples. For example, sieves or airflow devices can be used to homogenize the mixed samples to eliminate any potential clumping or uneven distribution. The total amount of integrated test samples should be determined based on the needs of subsequent testing and equipment limitations, ensuring a sufficiently large sample volume to reflect the overall characteristics of the original sample set while avoiding waste and unnecessary complexity. The mixed and homogenized integrated test samples should be stored in appropriate, clean, uncontaminated, and well-sealed containers to prevent contamination or moisture loss during storage. Finally, the integrated test samples should be labeled with their origin, mixing time, mixing method, etc., and key steps and parameters throughout the pooling process should be recorded for subsequent tracking and analysis. After the integrated test samples are prepared, preliminary quality control checks should be performed to ensure that the homogeneity, representativeness, and stability of the samples meet the requirements of subsequent testing.

[0031] The above steps transform a reliable sample set into an integrated test sample set, which can more accurately reflect the overall quality status of the original cotton sorting pile in subsequent multi-technology collaborative testing, thus ensuring the accuracy of sorting.

[0032] Step S600: Generate a weight allocation to divide the integrated test samples according to the sample limit set, and synchronize the division results to the standardized testing equipment set, control the standardized testing equipment set to perform multi-technology collaborative cotton quality sorting, and obtain a cotton quality evaluation result set.

[0033] For example, each quality indicator is assigned a corresponding weight based on the sample limit set (i.e., the upper and lower limits of different quality indicators such as color, fiber length, and strength). The weight allocation should be based on the importance of each indicator to the overall quality of the cotton. For example, color can be set to 30%, fiber length to 25%, strength to 20%, and other indicators such as uniformity and impurity content to a combined 25% weight. Next, using methods such as stratified sampling or proportional sampling, the integrated test sample is divided into multiple sub-sample sets according to the weight allocation. Each sub-sample set should maintain the representativeness of the original sample set as much as possible and meet the requirements of specific quality indicators. For example, a sub-sample set with a higher weight for the color indicator can include more cotton samples with uniform color distribution; a sub-sample set with a higher weight for the fiber length indicator can include more samples with lengths close to the standard value. The information of the divided sub-sample sets (including sample number, quality indicator requirements, weights, etc.) is synchronized to the standardized testing equipment set, ensuring that each testing device can accurately receive and identify its corresponding sub-sample set information. Based on the distribution results and the quality index requirements of the sub-sample sets, the standardized testing equipment is configured accordingly, including adjusting the testing parameters, calibrating the equipment accuracy, and setting the testing process.

[0034] In the actual sorting process, a set of standardized testing equipment is activated simultaneously or sequentially to perform multi-technology collaborative cotton quality sorting on the sub-sample sets under their respective responsibilities. This multi-technology collaboration may include various techniques such as optical detection (for color and impurity identification), mechanical detection (for fiber length and strength measurement), and chemical detection (for specific component analysis). The test results data from each testing device are collected in real time, summarized, and analyzed. Data processing software or algorithms are used to process the test data to assess the overall quality of the cotton. Furthermore, a cotton quality evaluation result set is generated based on the analysis results and weight allocation. This result set includes the specific values ​​of each quality indicator, grade ratings, and comprehensive quality evaluation information. The evaluation result set can be presented in the form of charts, reports, or electronic documents for subsequent quality traceability and decision analysis.

[0035] Step S700: Introduce a comprehensive evaluation function to quantify the cotton quality evaluation result set, obtain the cotton quality coefficient, and use the cotton quality coefficient to match the cotton grade number table to obtain the cotton grading result of the cotton sorting pile.

[0036] Furthermore, the comprehensive evaluation function is as follows: ; in, This is the cotton quality coefficient. To determine fiber length, To determine fiber strength, To determine the micronaire value, To determine the color, To determine the impurity content.

[0037] Specifically, a cotton grading table is developed based on industry standards and market demand. This table should clearly list the quality coefficient ranges and corresponding grade designations for different grades of cotton. Furthermore, a comprehensive evaluation function is introduced to quantify the cotton quality evaluation result set, wherein the comprehensive evaluation function is... ;in, This is the cotton quality coefficient. To determine fiber length, To determine fiber strength, To determine the micronaire value, To determine the color, To determine the impurity content, the calculated cotton quality coefficient is matched against a grading table to find the corresponding grade identifier. If the cotton quality coefficient falls within a certain grade range, that grade is used as the grading result for the cotton sorting pile. The cotton grading results are output in the form of reports, labels, or electronic documents for subsequent quality traceability, sales, and decision analysis. Simultaneously, the grading results are fed back to relevant departments or personnel to collect their feedback. Based on the feedback and actual conditions, necessary optimizations and improvements are made to the comprehensive evaluation function, grading table, or the entire sorting process to improve the accuracy and reliability of the grading results.

[0038] By constructing a comprehensive evaluation function to quantify the cotton quality evaluation result set and matching the cotton grades, cotton grading results are obtained, which improves sorting efficiency and accuracy.

[0039] Furthermore, such as Figure 2 As shown, random sampling is performed on the cotton sorting pile to obtain a sample cotton set. Step S300 of this application further includes: Step S310: Transfer the cotton sorting pile to a sampling container to obtain a cotton stack.

[0040] Step S320: Measure and obtain the stacking height parameter of the cotton stack.

[0041] Step S330: Synchronize the stacking height parameters to the pre-built sampling analysis network for sampling analysis and generate a sampling strategy.

[0042] Step S340: Use the sampling strategy described above to sample the cotton stack and obtain the sample cotton set.

[0043] Furthermore, the sampling analysis network includes cascaded sampling stratification matching units and sampling analysis execution units.

[0044] Optionally, the cotton to be sampled is first transferred to a specific sampling container and stacked within the container to form a cotton stack. This ensures the cotton remains stable during sampling, facilitating subsequent operations. Next, the stack height parameter is measured using appropriate measuring tools or methods. Stack height is an important physical quantity, reflecting information such as the stack's volume and density, and is crucial for subsequent sampling strategy development. The measured stack height parameter is synchronized to a pre-built sampling analysis network. This network is an intelligent system that, based on the input stack height and other parameters, analyzes the data through internal algorithms and models to produce the optimal sampling strategy.

[0045] The sampling analysis network comprises cascaded sampling stratification matching units and sampling analysis execution units. The sampling stratification matching unit is responsible for stratifying the cotton stack according to input parameters such as stack height and matching corresponding sampling strategies. Stratification ensures that cotton at different levels and locations is sampled uniformly and reasonably, thereby improving sample representativeness. The sampling analysis execution unit executes specific sampling operations after the sampling stratification matching unit determines the sampling strategy, including controlling robotic arms, conveyor belts, and other equipment to extract samples from the cotton stack according to the predetermined strategy. Finally, the actual sampling operation is performed on the cotton stack according to the sampling strategy produced by the sampling analysis network to obtain the required sample cotton set, which will be used for subsequent quality inspection, component analysis, and other work.

[0046] The above steps enable efficient and accurate sampling of cotton sorting piles. By measuring the stack height parameter and inputting it into the sampling analysis network, an optimal sampling strategy can be generated, thereby ensuring the representativeness of the sample and the accuracy of the sampling.

[0047] Furthermore, the stacking height parameter is synchronized to a pre-built sampling analysis network for sampling analysis to generate a sampling strategy. Step S330 of this application also includes: Step S331: Synchronize the stacking height parameter to the sampling stratification matching unit and obtain the stratification coefficient. The sampling stratification matching unit has a built-in stacking stratification number table.

[0048] Step S332: Digitally model the sampling container to obtain a digital container, and divide the digital container with the stratification coefficient as a constraint to obtain a multi-level sampling space.

[0049] Step S333: After locating multiple sampling starting points in the multi-layer sampling space, using 1 / K of the stacking height parameter as the sampling interval constraint, the cross-sampling engine built into the sampling analysis execution unit is used to perform random sampling in the multi-layer sampling space to obtain multiple sets of sampling space points, where K is a positive integer greater than 10.

[0050] Step S334: Aggregate the multiple sets of sampling spatial points to obtain the sampling strategy, and schedule the standardized sampling tool based on the sampling strategy to perform location sampling on the cotton stack to obtain the sample cotton set.

[0051] Furthermore, the stacking height parameter is sent to the sampling stratification matching unit, which stores a stacking stratification number table. This table, constructed based on historical data, physical characteristics, or expert experience, is used to convert the stacking height into a corresponding stratification coefficient. This stratification coefficient determines how many layers the stack will be divided into for sampling. The sampling stratification matching unit then calculates the most suitable stratification coefficient based on the stacking height parameter and the stratification number table. This coefficient guides the subsequent stratification sampling process.

[0052] Next, the sampling container is digitally modeled to form a digital container. This model is virtual but accurately reflects the shape, size, and location information of the sampling container. Specifically, the sampling container is precisely measured using 3D scanning, laser measurement, or other techniques, and this measurement data is input into a computer. Computer graphics or CAD software is used to digitally model the sampling container, generating an accurate digital container model. Then, based on the stratification coefficients obtained earlier through the sampling stratification matching unit, the digital container model is divided into multiple layers vertically (i.e., multi-layer sampling space). The stratification coefficients determine the number of layers and the thickness of each layer, ensuring that sampling covers the entire stack and reflects its internal variations.

[0053] Within each sampling layer, multiple sampling starting points are randomly located or positioned according to certain rules (such as uniform distribution, grid distribution, etc.) as needed by the sampling strategy. These starting points are the initial positions of the sampling process and are used to generate specific sampling paths or point sets later. The sampling interval is constrained by 1 / K of the stack height parameter, where K is a positive integer greater than 10. This interval determines the distance between sampling points within the same layer. A larger K value means denser sampling, which can more accurately reflect subtle changes within the stack; while a smaller K value may result in overly sparse sampling points, failing to capture important variation information.

[0054] The sampling analysis execution unit utilizes its built-in cross-sampling engine to perform random sampling within each sampling stratum, constrained by a set sampling starting point and sampling interval. The cross-sampling engine employs complex algorithms to ensure the randomness and uniformity of sampling points while avoiding excessive overlap or omissions between adjacent strata. This step yields multiple sets of sampling spatial points, each representing a potential sample location within the stack. The sampling spatial points from all strata are then aggregated to form a complete sampling strategy, encompassing all required sampling location information.

[0055] Finally, standardized sampling tools (such as robotic arm sampling probes and automated sampling machines) are deployed according to the aggregated sampling strategy to perform location sampling on the cotton stacks. These tools can accurately reach the designated sampling locations and collect sample cotton in a predetermined manner (such as grabbing and cutting). After location sampling, all collected sample cotton is combined into a sample cotton set, which will be used for subsequent quality testing and composition analysis.

[0056] By following the steps above, we can ensure the accuracy, representativeness, and efficiency of the sampling process, providing reliable data support for subsequent cotton quality evaluation.

[0057] Furthermore, to obtain the sampling strategy by aggregating the multiple sets of sampling spatial points, step S334 of this application also includes: Step S3341: Decouple the multiple sets of sampling spatial points in the digital container to obtain a spatial sampling scatter set.

[0058] Step S3342: Using 1 / K of the stacking height parameter as a constraint for repeated sampling, traverse the spatial sampling scatter set to remove scattered points and obtain the sampling strategy.

[0059] For example, to ensure that the sampling strategy is both representative and avoids unnecessary duplicate sampling, the multiple sets of sampling spatial points are aggregated. Specifically, firstly, multiple sets of sampling spatial points generated from different layers or different sampling strategies are aggregated, merging all potential sampling points into a single set for subsequent processing. Then, this aggregated set of sampling spatial points is decoupled in a digital container, transforming it into a spatial sampling scatter set. The purpose of decoupling is to treat each sampling point as an independent entity, allowing for individual evaluation of its necessity and effectiveness. Next, using 1 / K of the stacking height parameter as a duplicate sampling constraint, the spatial sampling scatter set is traversed to remove scatter points. Here, the duplicate sampling constraint means that when determining whether two sampling points are too close (i.e., whether they constitute duplicate sampling), the reciprocal of the stacking height parameter multiplied by an integer K greater than 1 is used as a distance threshold. If the distance between two sampling points is less than this threshold, they are considered too close, and one of the points needs to be removed to avoid duplicate sampling.

[0060] The specific process of scatter point removal and duplicate sampling can be performed as follows: First, initialization is performed by setting an empty set to store the final sampling strategy point set. Then, for each point in the spatial sampling scatter set, a traversal operation is performed, calculating the distance between the current point and all points already existing in the final sampling strategy point set. If any distance is less than 1 / K of the stacking height parameter, the current point is considered a duplicate and skipped; otherwise, the current point is added to the final point set. This process is repeated until all points in the spatial sampling scatter set have been processed.

[0061] After scatter sampling, the final sampling strategy point set is the optimized sampling strategy. This strategy ensures the representativeness of the sampling (because the point set covers different areas and layers of the stack) while avoiding unnecessary duplicate sampling (through the scatter sampling process). Finally, based on this optimized sampling strategy, a standardized sampling tool is used to perform location sampling on the cotton stack, obtaining a sample cotton set. This set will be used for subsequent quality inspection and composition analysis, further improving sorting efficiency and reducing sorting costs.

[0062] Furthermore, based on visual analysis, biased samples are identified and removed from the sample cotton set to obtain a reliable sample set. Step S400 of this application also includes: Step S410: Call a standard cotton image, wherein the standard cotton image has an image acquisition identifier, wherein the image acquisition identifier includes resolution, illumination and background.

[0063] Step S420: Based on the image acquisition identifier, control the standardized image acquisition device to acquire images of multiple sample cottons in the sample cotton set to obtain a sample cotton image set, wherein the sample cotton image set includes multiple sample cotton images of the multiple sample cottons.

[0064] Step S430: Extract features from the standard cotton image and the sample cotton image set, and perform similarity analysis based on the extraction results to obtain multiple comprehensive similarity scores.

[0065] Step S440: Based on the multiple comprehensive similarity scores, identify and remove biased samples from the sample cotton set to obtain the reliable sample set.

[0066] Next, one or more standard cotton images are used as references. These standard cotton images should have clear image acquisition identifiers, including resolution, lighting conditions, and background, to ensure consistency and comparability in subsequent image acquisition. Using standardized image acquisition equipment, images of multiple sample cotton plants in the sample cotton set are acquired according to the image acquisition identifiers of the standard cotton images, ensuring that the acquired sample cotton images are as close as possible in quality to the standard cotton images for accurate similarity analysis. The acquired images of multiple sample cotton plants are combined into a sample cotton image set for subsequent image processing and analysis.

[0067] Next, feature extraction is performed on each image in both the standard cotton image set and the sample cotton image set to extract useful information (such as color, texture, and shape) for subsequent analysis and comparison. Similarity analysis is then performed on each image in both sets based on the extracted features. Similarity analysis can be implemented using various methods, such as calculating the distance between feature vectors (e.g., Euclidean distance, Manhattan distance) or using machine learning algorithms (e.g., support vector machines, neural networks) for pattern recognition. Finally, a comprehensive similarity score is generated for each sample cotton image, reflecting its degree of similarity to the standard cotton image. Each sample in the sample cotton set is evaluated based on its comprehensive similarity score. If a sample's comprehensive similarity score is below a preset threshold, it is considered a biased sample and removed from the sample set.

[0068] The identification and removal of deviation samples is achieved using discrete value removal. The identification and removal methods can include the MAD method or the 3σ method. Taking the MAD method as an example, firstly, all samples are sorted according to a certain feature (such as impurity content), the median value is calculated, and the absolute deviation of each sample from the median value is calculated. Then, the median value of all absolute deviations is taken as the MAD. A threshold (e.g., n times the MAD) is set based on experience or actual needs, and all samples exceeding this threshold are considered deviation samples and removed. Finally, the dataset after removing deviation samples is verified to ensure that the quality of the remaining samples meets the standards. If the verification results show that a large number of deviation samples have not been removed or normal samples have been mistakenly removed, the threshold needs to be adjusted or other methods need to be used for further processing. Ultimately, the above steps are integrated into an automated sorting system to achieve rapid and accurate sorting of cotton samples. By continuously monitoring the sorting results and adjusting the sorting strategy based on feedback, the stability and reliability of the cotton sorting process are ensured.

[0069] This process yields a reliable sample set consisting of samples that meet the standards. Finally, this reliable sample set is used for subsequent quality testing and component analysis, ensuring that the results of these analyses are more accurate and reliable.

[0070] Furthermore, feature extraction is performed on the standard cotton image and the sample cotton image set, and similarity analysis is conducted based on the extraction results to obtain multiple comprehensive similarity scores. Step S430 of this application also includes: Step S431: Pre-construct an image comparison model, wherein the image comparison model includes an image feature extraction layer, an image feature comparison layer, and a comprehensive similarity calculation layer. The image feature comparison layer includes parallel texture feature analysis branches, color feature analysis branches, and fiber arrangement analysis branches. The comprehensive similarity calculation layer pre-stores texture weight assignment, color weight assignment, and arrangement weight assignment.

[0071] Step S432: Synchronize the first sample cotton image and the standard cotton image to the image feature extraction layer of the image comparison model for feature extraction to obtain the first sample feature set and the first standard feature set.

[0072] Step S433: Perform parallel feature similarity analysis on the first sample feature set and the first standard feature set through the texture feature analysis branch, color feature analysis branch, and fiber arrangement analysis branch of the image feature comparison layer to obtain the first sample similarity set.

[0073] Step S434: The first sample similarity set is weighted and integrated using the texture weight assignment, color weight assignment, and arrangement weight assignment in the comprehensive similarity calculation layer to obtain the first comprehensive similarity score.

[0074] Step S435: Perform feature extraction and similarity analysis based on the extraction results to obtain multiple comprehensive similarity scores.

[0075] Step S436: Similarly, calculate the multiple comprehensive similarity scores of the standard cotton image and the sample cotton image set.

[0076] Specifically, a pre-constructed image comparison model is used, comprising an image feature extraction layer, an image feature comparison layer, and a comprehensive similarity calculation layer. The image feature extraction layer extracts key features from the input image; these features can be pixel-level, texture-based, color-based, or other characteristics that help distinguish different images. The image feature comparison layer contains three parallel branches, used to analyze texture features, color features, and fiber arrangement features, respectively. These three branches work independently, evaluating image similarity from different perspectives. The texture feature analysis branch analyzes texture characteristics such as roughness and smoothness; the color feature analysis branch analyzes color attributes such as color distribution, hue, and saturation; and the fiber arrangement analysis branch (specific to cotton images) analyzes the arrangement and density of cotton fibers. The comprehensive similarity calculation layer then integrates the similarity analysis results from each branch according to preset weights (texture weight, color weight, and arrangement weight) to obtain a comprehensive similarity score.

[0077] Furthermore, the first sample cotton image and the standard cotton image are input into the image feature extraction layer to extract the first sample feature set and the standard feature set, respectively. These two feature sets are then input into the three branches of the image feature comparison layer for parallel analysis, resulting in the first sample similarity set (including texture similarity, color similarity, and fiber arrangement similarity) for each branch. The weights (texture weight assignment, color weight assignment, and arrangement weight assignment) pre-stored in the comprehensive similarity calculation layer are used to perform a weighted summation of the similarities in the first sample similarity set, yielding the first comprehensive similarity score. For each sample image in the sample cotton image set, the above process of feature extraction, parallel feature comparison, and weighted comprehensive similarity calculation is repeated to obtain their respective comprehensive similarity scores with the standard cotton image. Finally, multiple comprehensive similarity scores are obtained, reflecting the degree of similarity between each sample in the sample cotton image set and the standard cotton image. Based on these scores, further evaluation, classification, or selection operations can be performed. The above steps allow for a comprehensive consideration of features across multiple dimensions (texture, color, fiber arrangement), and a weighted approach to balance the importance of different features, thereby improving the accuracy and reliability of similarity assessment.

[0078] Through the technical solutions of the above embodiments, the cotton quality sorting and control method based on multi-technology collaboration provided in this application solves the technical problems of low sorting efficiency, insufficient accuracy and comprehensiveness of cotton quality control in the cotton quality sorting process, which leads to poor sorting results and affects subsequent processing and sales. It achieves the technical effect of improving the accuracy, efficiency and comprehensiveness of cotton sorting, thereby ensuring the reliability of subsequent processing and sales.

[0079] 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 specification.

[0080] The embodiments described above are merely examples of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. 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.

Claims

1. A cotton quality sorting and control method based on multi-technology collaboration, characterized in that, The method includes: A set of cotton quality evaluation indicators is preset, and a set of standardized testing equipment is obtained by mapping and scheduling according to the set of cotton quality evaluation indicators; The standardized testing equipment set is subjected to a minimum sample quantity limit collection to obtain a sample limit set; The standardized sampling tool is used to randomly sample cotton from the sorted piles to obtain a sample cotton set. Based on visual analysis, biased samples are identified and removed from the sample cotton set to obtain a reliable sample set. The reliable sample set is subjected to sample pooling to obtain integrated test samples; The integrated test samples are distributed according to the sample limit set, and the distribution results are synchronized to the standardized testing equipment set. The standardized testing equipment set is then controlled to perform multi-technology collaborative cotton quality sorting to obtain a cotton quality evaluation result set. A comprehensive evaluation function is introduced to quantify the cotton quality evaluation result set, and a cotton quality coefficient is obtained. The cotton quality coefficient is then matched with a cotton grade number table to obtain the cotton grading result of the cotton sorting pile.

2. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 1, characterized in that, The method further includes randomly sampling cotton from sorted piles to obtain a sample cotton set, wherein the sampling method further includes: The cotton sorting pile is transferred to a sampling container and stacked to obtain a cotton stack. The stacking height parameter of the cotton stack was measured and obtained; The stacking height parameters are synchronized to a pre-built sampling analysis network for sampling analysis to generate a sampling strategy. The sampling strategy described above is used to sample the cotton stack to obtain the sample cotton set.

3. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 2, characterized in that, The sampling analysis network includes cascaded sampling stratification matching units and sampling analysis execution units.

4. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 3, characterized in that, The method further includes synchronizing the stacking height parameters to a pre-built sampling analysis network for sampling analysis to generate a sampling strategy. The stacking height parameter is synchronized to the sampling stratification matching unit to obtain the stratification coefficient. The sampling stratification matching unit has a built-in stacking stratification number table. The sampling container is digitally modeled to obtain a digital container, and the digital container is divided with the stratification coefficient as a constraint to obtain a multi-level sampling space; After locating multiple sampling starting points in the multi-layer sampling space, 1 / K of the stacking height parameter is used as the sampling interval constraint. The cross-sampling engine built into the sampling analysis execution unit is used to perform random sampling in the multi-layer sampling space to obtain multiple sets of sampling space points, where K is a positive integer greater than 10. By aggregating the multiple sets of sampling spatial points, the sampling strategy is obtained, and the standardized sampling tool is scheduled to perform location sampling on the cotton stack based on the sampling strategy to obtain the sample cotton set.

5. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 4, characterized in that, The sampling strategy is obtained by aggregating the multiple sets of sampling spatial points, and the method further includes: Decouple the multiple sets of sampling spatial points in the digital container to obtain a spatial sampling scatter set; Using 1 / K of the stacking height parameter as a constraint for repeated sampling, the spatial sampling scatter point set is traversed to remove scatter points, thereby obtaining the sampling strategy.

6. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 1, characterized in that, Based on visual analysis, biased samples are identified and removed from the sample cotton set to obtain a reliable sample set. The method further includes: A standard cotton image is retrieved, wherein the standard cotton image has image acquisition identifiers, wherein the image acquisition identifiers include resolution, illumination, and background; Based on the image acquisition identifier, the standardized image acquisition device is controlled to acquire images of multiple sample cottons in the sample cotton set to obtain a sample cotton image set, wherein the sample cotton image set includes multiple sample cotton images of the multiple sample cottons; Feature extraction is performed on the standard cotton image and the sample cotton image set, and similarity analysis is performed based on the extraction results to obtain multiple comprehensive similarity scores; Based on the multiple comprehensive similarity scores, the sample cotton set is used to identify and remove biased samples to obtain the reliable sample set.

7. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 6, characterized in that, Feature extraction is performed on the standard cotton image and the sample cotton image set, and similarity analysis is conducted based on the extraction results to obtain multiple comprehensive similarity scores. The method further includes: A pre-constructed image comparison model is provided, wherein the image comparison model includes an image feature extraction layer, an image feature comparison layer, and a comprehensive similarity calculation layer. The image feature comparison layer includes parallel texture feature analysis branches, color feature analysis branches, and fiber arrangement analysis branches. The comprehensive similarity calculation layer pre-stores texture weight assignment, color weight assignment, and arrangement weight assignment. The first sample cotton image and the standard cotton image are synchronized to the image feature extraction layer of the image comparison model for feature extraction to obtain the first sample feature set and the first standard feature set. The first sample feature set and the first standard feature set are subjected to parallel feature similarity analysis through the texture feature analysis branch, color feature analysis branch, and fiber arrangement analysis branch of the image feature comparison layer to obtain the first sample similarity set; The first sample similarity set is weighted and integrated using the texture weight assignment, color weight assignment, and arrangement weight assignment in the comprehensive similarity calculation layer to obtain the first comprehensive similarity score; Feature extraction is performed, and similarity analysis is conducted based on the extraction results to obtain multiple comprehensive similarity scores; Similarly, the multiple comprehensive similarity scores of the standard cotton image and the sample cotton image set are calculated.

8. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 1, characterized in that, The cotton quality evaluation index set includes fiber length, fiber strength, micronaire value, color, and impurity content.

9. The cotton quality sorting and control method based on multi-technology collaboration as described in claim 8, characterized in that, The comprehensive evaluation function is as follows: ； in, This is the cotton quality coefficient. To determine fiber length, To determine fiber strength, To determine the micronaire value, To determine the color, To determine the impurity content.

Images