Plastic recovery tracking method and system, program product and storage medium

Abstract

The invention provides a plastic recycling and tracking method and system, a program product and a storage medium, and relates to the technical field of intelligent environment-friendly recycling. The target plastic group is divided into different categories to realize basic classification, so that mixing of plastics of different materials is avoided; the plastic source is traceable by collecting geographic position information and generating a digital map; and finally, the plastics at different geographic positions are distributed to the corresponding packaging mechanisms according to a preset classification rule, and accurate classification treatment is achieved. According to the progressive treatment mode, the plastics of the same kind but different producing areas can be accurately distinguished and classified, mixing treatment of the plastics of different technological levels and different modification formulas is effectively avoided, and therefore the classification management precision is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent environmentally friendly recycling technology, and in particular to a plastic recycling tracking method, system, program product and storage medium. Background Art

[0002] As global awareness of environmental protection increases, plastic recycling has become one of the key links in reducing environmental pollution.

[0003] At present, the plastic recycling process mainly relies on an automatic classification system, which uses optical recognition equipment to identify the material type of the plastic. The system automatically matches the collected information with the preset classification database, and then the sorting equipment transports different types of plastics to the corresponding processing areas.

[0004] However, the relevant technology will generate some new problems in the actual use process. For example, the actual performance and recycling value of the same type of plastics produced by different manufacturers vary due to differences in process level and raw material selection. Secondly, some manufacturers use special modified formulas. If these plastics are mixed with ordinary plastics for recycling, they may cause adverse reactions in the recycling process. In summary, the accuracy of the classification management of the relevant technology is too low. Summary of the invention

[0005] The present application provides a plastic recycling tracking method, system, program product and storage medium for improving the accuracy of classification management.

[0006] In a first aspect, the present application provides a plastic recycling tracking method, including: dividing a target plastic group into different categories to obtain a number of sub-target plastic groups; collecting geographic location information of the sub-target plastic groups; generating a digital map of the plastic source based on the geographic location information; and allocating plastics in different geographic locations in the sub-target plastic group to corresponding packaging organizations for processing based on the digital map based on preset classification rules.

[0007] By adopting the above technical solution, the target plastic group is divided into different categories to achieve basic classification, avoiding the mixing of plastics of different materials; then by collecting geographical location information and generating digital maps, the source of plastics can be traced; finally, plastics in different geographical locations are assigned to corresponding packaging agencies according to preset classification rules, achieving accurate classification and processing. This step-by-step processing method enables plastics of the same type but different origins to be accurately distinguished and classified, effectively avoiding the mixed processing of plastics with different process levels and different modified formulas, thereby improving the accuracy of classification management.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of dividing the target plastic group into different categories to obtain several sub-target plastic groups specifically includes: obtaining surface image data of the target plastic group; performing image enhancement processing on the surface image data to obtain enhanced image data; extracting surface features from the enhanced image data; detecting the contamination area in the surface features to obtain contamination degree data; determining whether the contamination degree data is greater than a preset contamination threshold; if it is greater than the preset contamination threshold, collecting material feature data and spectral feature data of the target plastic group; inputting the material feature data and spectral feature data into a deep learning model; generating a plastic category prediction result through the deep learning model; and classifying the target plastic group based on the plastic category prediction result to obtain several sub-target plastic groups.

[0009] By adopting the above technical solutions, the accuracy of feature extraction is improved through image enhancement processing, making the surface features clearer and more identifiable; the reliability of the classification process is ensured through threshold judgment of the contamination degree data; the introduction of deep learning models can learn and adapt to the characteristic manifestations of different types of plastics and can handle plastics with severe surface contamination, continuously improving the recognition accuracy, laying the foundation for subsequent origin traceability.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the step of collecting the geographic location information of the sub-target plastic group specifically includes: obtaining text image data on the surface of the sub-target plastic group; determining whether the text image data is complete; if the text image data is complete, preprocessing the text image data to obtain preprocessed image data; performing optical character recognition processing on the preprocessed image data to obtain text recognition data; and extracting the geographic location information from the text recognition data.

[0011] By adopting the above technical solution, the system judges the integrity of the text image data to ensure the quality of the data for subsequent processing; then, the clarity and recognizability of the text image are enhanced through the pre-processing step; and then the optical character recognition technology is used to convert the text information in the image into processable digital text. This improves the processing efficiency and maintains stable recognition accuracy.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the step of extracting geographic location information from text recognition data specifically includes: performing keyword matching on the text recognition data to obtain geographic location keyword matching results; extracting text data that conforms to the address format from the matching results; converting the text data into geographic coordinate data, and determining the geographic coordinate data as geographic location information.

[0013] By adopting the above technical solution, the text recognition data is initially screened through keyword matching technology to extract content that may contain geographic location information; then, the address format verification mechanism is used to ensure that the extracted information meets the standard address format requirements; finally, the verified text data is converted into standard geographic coordinate data. This multiple verification and conversion mechanism not only improves the accuracy of geographic location information, but also realizes the standardized expression of location information. Standardized geographic coordinate data facilitates spatial analysis and classification management, and also provides a reliable data basis for subsequent data statistics and analysis.

[0014] In combination with some embodiments of the first aspect, in some embodiments, the step of converting text data into geographic coordinate data specifically includes: matching the text data with data in a geographic location information mapping database; the geographic location information mapping database contains standard address text format data, geographic location keyword data, and address hierarchical structure data; generating standardized address text based on the matching results; and converting the standardized address text into geographic coordinate data.

[0015] By adopting the above technical solution, a mapping database containing standard address text format data, geographic location keyword data and address hierarchical structure data is used to provide a comprehensive reference for coordinate conversion; then, through the data matching process, the standardization and accuracy of address information are ensured; finally, accurate geographic coordinate data is obtained through the conversion of standardized address text. This conversion mechanism based on a complete mapping database not only improves the accuracy of coordinate conversion, but also ensures the consistency of address information from different sources.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining whether the text image data is complete, the method also includes: if the text image data is incomplete, adjusting the focal length parameters and the exposure parameters, and reacquiring new text image data on the surface of the sub-target plastic group; determining whether the new text image data is complete; if the new text image data is complete, treating the new text image data as text image data, and performing preprocessing on the text image data to obtain the preprocessed image data; if the new text image data is incomplete, constructing an address hierarchy dependency graph, establishing the relationship between address fields, including a unique correspondence between a lower-level address field and an upper-level address field; a spatial adjacent relationship between address fields of the same level; and an association relationship between multiple fields of the same address; executing an address information completion strategy according to the address hierarchy dependency graph to obtain the completed text image data, wherein the address information completion strategy includes: completing the upper-level address field based on a unique correspondence; completing the same-level address field based on a spatial adjacent relationship; and completing multiple fields of the same address based on an association relationship; treating the completed text image data as text image data, and performing preprocessing on the text image data to obtain the preprocessed image data.

[0017] By adopting the above technical solution, the system determines the data quality through the integrity judgment of text images; when the data is found to be incomplete, it first re-collects by adjusting the focal length and exposure parameters to improve the quality of the original data; if it is still incomplete after optimization, the information completion mechanism based on the address hierarchy dependency graph is activated. By establishing the unique correspondence, spatial adjacent relationship and association relationship between address fields, the system can intelligently infer and complete the missing address information. This maximizes the integrity of the geographic location information.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after the step of extracting geographic location information from text recognition data, the method also includes: collecting color features, texture features, and shape features of historical sub-target plastic groups to obtain feature data; constructing a contaminated sample database, which stores feature data of sub-target plastic groups with different historical geographic location information; calculating the similarity between the feature data of the new sub-target plastic group and the feature data in the contaminated sample database to obtain a feature similarity score; and determining the historical geographic location information with the highest feature similarity score as the geographic location information of the new sub-target plastic group.

[0019] By adopting the above technical solution, the system has established a comprehensive feature database by collecting the color, texture and shape features of historical samples; then, by building a database of contaminated samples, a mapping relationship is established between the feature data and the geographic location information; when processing new samples, by calculating the feature similarity, its possible geographic location can be intelligently inferred. In the case of plastics without text information, this method can quickly identify and supplement the missing geographic location information, improving processing efficiency. In addition, as the number of samples in the database increases, the inference accuracy will continue to improve, realizing the self-optimization and evolution of the classification system.

[0020] In a second aspect, the present application provides a plastic recycling tracking system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the plastic recycling tracking system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, the present application provides a computer program product comprising instructions, which, when executed on a plastic recycling tracking system, enables the plastic recycling tracking system to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer-readable storage medium comprising instructions, which, when executed on a plastic recycling tracking system, causes the plastic recycling tracking system to execute the method described in the first aspect and any possible implementation of the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Divide the target plastic group into different categories to achieve basic classification, avoiding the mixing of plastics of different materials; then collect geographic location information and generate digital maps to make the source of plastics traceable; finally, according to the preset classification rules, plastics in different geographical locations are assigned to corresponding packaging agencies to achieve accurate classification and processing. This progressive processing method enables plastics of the same type but different origins to be accurately distinguished and classified, effectively avoiding the mixed processing of plastics with different process levels and different modified formulas, thereby improving the accuracy of classification management.

[0024] 2. The system determines the data quality by judging the integrity of text and images. When the data is found to be incomplete, it first re-collects by adjusting the focal length and exposure parameters to improve the quality of the original data. If it is still incomplete after optimization, the information completion mechanism based on the address hierarchy dependency graph is activated. By establishing the unique correspondence, spatial adjacent relationship and association relationship between address fields, the system can intelligently infer and complete the missing address information. This maximizes the integrity of the geographic location information.

[0025] 3. The system has established a comprehensive feature database by collecting the color, texture and shape features of historical samples; then, by building a database of contaminated samples, it has established a mapping relationship between feature data and geographic location information; when processing new samples, it can intelligently infer their possible geographic locations by calculating feature similarities. In the case of plastics without text information, this method can quickly identify and supplement missing geographic location information, improving processing efficiency. In addition, as the number of samples in the database increases, the inference accuracy will continue to improve, realizing the self-optimization and evolution of the classification system. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of a plastic recycling tracking method in an embodiment of the present application; Figure 2 This is a specific flow chart of step S101 in the embodiment of the present application; Figure 3 is another flow chart of the plastic recycling tracking method in an embodiment of the present application; Figure 4 It is a schematic diagram of an exemplary hardware structure of the plastic recycling tracking system in the embodiment of the present application. DETAILED DESCRIPTION

[0027] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items.

[0028] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0029] See also Figure 1 , Figure 1 It is a flow chart of a plastic recycling tracking method in an embodiment of the present application; S101, dividing the target plastic group into different categories to obtain a plurality of sub-target plastic groups; Among them, the target plastic group represents the overall plastics to be classified and processed; the sub-target plastic group refers to the subset of plastics with similar characteristics formed after classification.

[0030] In some embodiments, the collected plastics are pre-processed to remove obvious impurities and contaminants; characteristic information of the plastics is obtained through visual inspection, material analysis, etc.; finally, the plastics are divided into different categories based on this characteristic information to form sub-target plastic groups that are convenient for subsequent processing.

[0031] In some specific embodiments, the plastic surface is scanned by an optical recognition device to extract color and shape features, and automatic classification is performed in combination with a machine learning algorithm. Classification methods can also be used to classify plastics, which is not limited here.

[0032] In some specific embodiments, the material composition of the plastic is detected by a near-infrared spectrometer, a material database is established for comparative analysis, and the plastics are grouped according to material similarity. A classification method can also be used to classify the plastics, which is not limited here.

[0033] S102, collecting geographical location information of the sub-target plastic group; Among them, geographic location information represents the source location of the plastic and related spatial data.

[0034] After completing the plastic classification, the specific source location of each group of plastics needs to be determined. In some embodiments, the plastic surface is first checked for origin identification or address information; then a clear image of this information is obtained through an image acquisition device; then the image is converted into text data using character recognition technology; and finally, the text address is converted into standard geographic coordinate data through geocoding.

[0035] In some specific embodiments, the text information on the plastic surface is photographed by a smart camera, the address text is recognized using OCR technology, and the geographic information system is used for verification and standardization to finally generate standardized geographic location data. It is understandable that other methods can also be used to obtain geographic location information, which is not limited here.

[0036] S103, generating a digital map of the plastic source based on the geographic location information; Among them, digital map refers to the visual expression of geographic spatial information displayed in electronic form.

[0037] In some embodiments, the collected geographic location data are first formatted in a unified manner; then the data are imported into a geographic information system; different legend styles are then set according to different categories of plastics; and finally, all location points are visualized on a base map to form an intuitive digital map.

[0038] Optionally, use the WebGIS platform to import geographic coordinate data into the system, set map symbols and styles, add legends and annotations, and finally generate an interactive Web map. Other map production and visualization methods can also be used to generate digital maps, which are not limited here.

[0039] S104, based on the preset classification rules and according to the digital map, the plastics in different geographical locations in the sub-target plastic group are allocated to corresponding packaging organizations for processing.

[0040] In some embodiments, the geographical location distribution of all plastics is confirmed based on a digital map; then the plastics in different geographical locations are centrally allocated to the same packaging organization; then it is ensured that the packaging organization has sufficient processing capacity; and finally, the plastics are transported in sequence for packaging processing.

[0041] It can be seen that the target plastic group is divided into different categories to achieve basic classification, avoiding the mixing of plastics of different materials; then by collecting geographic location information and generating digital maps, the source of plastics can be traced; finally, plastics in different geographic locations are assigned to corresponding packaging agencies according to preset classification rules, achieving accurate classification and processing. This progressive processing method enables plastics of the same type but different origins to be accurately distinguished and classified, effectively avoiding the mixed processing of plastics with different process levels and different modified formulas, thereby improving the accuracy of classification management.

[0042] See also Figure 2 , Figure 2 This is a specific flow chart of step S101 in the embodiment of the present application; In some specific embodiments, step S101 specifically includes: S1011, acquiring surface image data of a target plastic group; Among them, the surface image data represents the digital visual information of the plastic appearance;.

[0043] Before starting plastic classification, it is necessary to first obtain visual information of the plastic. In some embodiments, the plastic on the conveyor belt is continuously photographed by an industrial camera to collect images from multiple angles.

[0044] S1012, performing image enhancement processing on the surface image data to obtain enhanced image data; In some embodiments, noise is first removed from the image; then the brightness, contrast, and sharpness of the image are adjusted; then color correction and geometric correction are performed; and finally an enhanced image with improved quality is output.

[0045] In some specific embodiments, histogram equalization is used to enhance contrast, Gaussian filtering is used to eliminate noise, a sharpening algorithm is applied to enhance edge features, and finally a color space conversion is performed. Other image processing algorithms may also be used to achieve image enhancement, which is not limited here.

[0046] S1013, extracting surface features from the enhanced image data; Among them, surface features refer to various visual characteristics of the plastic appearance.

[0047] In some specific embodiments, an image processing algorithm is used to calculate color moment features, extract grayscale co-occurrence matrix features, detect edge contour features, and finally form a feature vector. Other feature extraction methods can also be used to obtain surface features, which are not limited here.

[0048] In some specific embodiments, the gray level co-occurrence matrix (GLCM) is first used to extract texture features, including energy, contrast, correlation and entropy; then the color features are extracted through the color histogram, including the statistical distribution of RGB or HSV space; then the edge detection operator is used to extract structural features, including gradient direction and amplitude; finally, all features are combined to form surface features. Other feature extraction methods can also be used to obtain surface features, which are not limited here.

[0049] S1014, detecting the contaminated area in the surface features and obtaining contamination degree data; First, the abnormal regions in the image are identified; then the area ratio of the abnormal regions is calculated.

[0050] In some specific embodiments, an image segmentation algorithm is used to divide normal areas and damaged areas, calculate the area ratio of each type of damage, evaluate the severity of the damage, and finally generate a damage report. Other detection methods can also be used to evaluate the damage situation, which is not limited here.

[0051] S1015, determining whether the contamination degree data is greater than a preset contamination threshold; If it is not greater than the preset contamination threshold, the plastic category is determined based on whether the surface characteristics are within the preset threshold range, and plastics with similar characteristics are grouped together to achieve the division of the target plastic group into different categories.

[0052] In some specific embodiments, the judgment is made based on a preset color threshold range, texture threshold, and transparency threshold; finally, the plastic category is determined based on whether the characteristic value is within the preset threshold range, and plastics with similar characteristics are grouped together.

[0053] In some specific embodiments, based on the color feature: RGB three-channel values ​​are calculated, and the color category is determined according to the proportional relationship between different color components.

[0054] Based on texture features: calculate the grayscale distribution of the image and distinguish the surface texture categories through the grayscale standard deviation.

[0055] Based on transparency features: Analyze the visibility of the image and the background to determine the degree of transparency.

[0056] Feature combination judgment: Combine multiple feature thresholds to improve classification accuracy.

[0057] If it is greater than the preset contamination threshold, then S1016, collecting material characteristic data and spectral characteristic data of the target plastic group; When the contamination level exceeds a threshold, more feature information is needed to assist in classification. In some embodiments, near infrared spectroscopy is performed, followed by measuring other physical properties, and finally all feature data are integrated.

[0058] In some specific embodiments, hardness and density are measured, thermal properties are analyzed, and finally a feature data set is generated; other detection methods may also be used to obtain material and spectral features, which are not limited here.

[0059] In some specific embodiments, a multi-sensor system is used to simultaneously collect Raman spectra and X-ray fluorescence data, measure the conductive properties, and form multi-dimensional feature data; other detection methods may also be used to obtain material and spectral features, which are not limited here.

[0060] S1017, inputting material feature data and spectral feature data into a deep learning model; In some embodiments, the feature data is first preprocessed and standardized; then the data is converted into a model input format; then the data is fed into a trained model; and finally, the model processing is waited for to be completed.

[0061] In some specific embodiments, a training data set is first constructed, including material characteristics and spectral characteristics data of a large number of annotated plastic samples, as well as corresponding label information; then a suitable deep neural network architecture is selected, such as a convolutional neural network (CNN) or a multi-layer perceptron (MLP); the training data is then used to train and verify the model; finally, the trained model is deployed and used to receive new feature data input.

[0062] S1018. Generate plastic category prediction results through deep learning model; S1019. Classify the target plastic group based on the plastic category prediction result to obtain several sub-target plastic groups.

[0063] It can be seen that the accuracy of feature extraction is improved through image enhancement processing, making the surface features clearer and more identifiable; the reliability of the classification process is ensured through threshold judgment of the contamination degree data; the introduction of the deep learning model can learn and adapt to the characteristic manifestations of different types of plastics and can handle plastics with severe surface contamination, continuously improving the recognition accuracy, and laying the foundation for subsequent origin traceability.

[0064] See also Figure 3 , Figure 3 is another flow chart of the plastic recycling tracking method in an embodiment of the present application; In some specific embodiments, step S102 specifically includes: S201, obtaining text image data on the surface of the sub-target plastic group; It should be noted that the principle and process of this step are similar to those of step S1011. The relevant principles and processes can be referred to step S1011 and will not be repeated here.

[0065] S202, determining whether the text image data is complete; After acquiring the text image, its integrity needs to be evaluated. In some embodiments, the text image is first binarized to obtain a clear text outline; then the connectivity of the text area is calculated to detect whether there is a break; then the integrity of the text strokes is analyzed to determine whether there is a missing stroke; finally, it is determined whether the text image meets the requirements based on a preset integrity threshold.

[0066] In some specific embodiments, the number of connected components in the text area is calculated. If it exceeds the expectation, it means that the text is broken, or the number of pixels of the text strokes is counted. If it is lower than a threshold, it means that the text is missing, or the clarity of the text edge is analyzed. If the blurriness exceeds the threshold, it means that the image quality is unqualified, or the closedness of the text area is detected to ensure the integrity of the character structure. There is no limitation here.

[0067] S203, if the text image data is complete, preprocessing the text image data to obtain preprocessed image data; S204, performing optical character recognition processing on the pre-processed image data to obtain text recognition data; After image preprocessing is completed, the text in the image needs to be converted into text. In some embodiments, the trained OCR model is first loaded; then the image is subjected to text detection and positioning; then the detected text is recognized one by one; and finally the recognition results are integrated to form a complete text.

[0068] In some specific embodiments, a multilingual OCR system is used to automatically identify the language of text, apply corresponding recognition models, and generate multilingual text results. Other recognition methods can also be used to achieve text conversion, which is not limited here.

[0069] S205: Extracting geographic location information from the text recognition data.

[0070] The geographical location information indicates the source or production location of the plastic. After obtaining the text recognition result, it is necessary to locate the information related to the geographical location. In some embodiments, a geographical location keyword library is first established, then the text is segmented and matched, and then the matched geographical information is extracted.

[0071] It can be seen that the system judges the integrity of the text image data to ensure the quality of the data for subsequent processing; then, the clarity and recognizability of the text image are enhanced through the preprocessing step; and then the optical character recognition technology is used to convert the text information in the image into processable digital text. This improves the processing efficiency and maintains stable recognition accuracy.

[0072] In some embodiments, step S205 specifically includes: S2051, performing keyword matching on the text recognition data to obtain a geographic location keyword matching result; Among them, keyword matching refers to the process of comparing text with a preset geographic location vocabulary; keywords include place name information such as country, region, city, street, etc.

[0073] After obtaining the OCR text, it is necessary to identify the geographic location related words therein. In some embodiments, the text is segmented and standardized based on a pre-built multi-level geographic location keyword library; then a matching algorithm is used to find possible geographic location words; and finally the results are sorted and filtered according to the matching degree.

[0074] In some specific embodiments, a geographic location dictionary tree is constructed, text is segmented, a string matching algorithm is executed, a matching score is calculated, and a high-confidence matching result is output. Other matching methods may also be used to identify geographic location keywords, which are not limited here.

[0075] S2052, extracting text data conforming to the address format from the matching results; The address format refers to a method of describing a geographic location that complies with specific specifications.

[0076] After obtaining the keyword matching results, it is necessary to identify the complete address information. In some embodiments, firstly, multiple standard address format templates are defined; then pattern matching is performed on the matching results; then address texts that meet the format requirements are extracted; and finally the extracted addresses are normalized.

[0077] In some specific embodiments, regular expression matching is used to construct address format rules, verify address integrity, standardize address formats, and output standardized address text. Other extraction methods may also be used to obtain standard address text, which is not limited here.

[0078] S2053: Convert the text data into geographic coordinate data, and determine the geographic coordinate data as geographic location information.

[0079] Among them, geographic coordinate data represents the latitude and longitude information of the geographic location; conversion refers to the process of mapping a text address into specific coordinates; and geographic location information is used to represent the finalized location data.

[0080] After obtaining the standard address, it needs to be converted into coordinate data that can be used for positioning. In some embodiments, first connect to the geographic information service system; then send an address resolution request; then receive the returned coordinate data; finally verify the accuracy of the coordinates and save the results.

[0081] In some specific embodiments, the geocoding service API is called, an address query request is sent, the response data is parsed, the coordinate validity is verified, and the final result is stored; other conversion methods may also be used to obtain geographic coordinate data, which are not limited here.

[0082] In some specific embodiments, an offline geographic database is used to construct an address index system, query geographic coordinate information, and output standard coordinate data. It is understandable that other conversion methods can also be used to obtain geographic coordinate data, which is not limited here.

[0083] It can be seen that the text recognition data is initially screened through keyword matching technology to extract content that may contain geographic location information; then, the address format verification mechanism is used to ensure that the extracted information meets the standard address format requirements; finally, the verified text data is converted into standard geographic coordinate data. This multiple verification and conversion mechanism not only improves the accuracy of geographic location information, but also realizes the standardized expression of location information. Standardized geographic coordinate data facilitates spatial analysis and classification management, and also provides a reliable data basis for subsequent data statistics and analysis.

[0084] In some specific embodiments, step S2053 specifically includes: S20531, matching the text data with data in a geographic location information mapping database; the geographic location information mapping database includes standard address text format data, geographic location keyword data, and address hierarchical structure data; Among them, the geographic location information mapping database represents a structured data set that stores address-related information; the standard address text format data refers to the address representation format that complies with the standard; the geographic location keyword data is used to represent the place name entity vocabulary; the address hierarchical structure data represents the administrative division hierarchy of the geographic location; and the matching represents the correspondence process between the text and the database record.

[0085] In some embodiments, the geographic location information mapping database is first connected; then matching is performed with different types of data, including format matching, keyword matching, and hierarchical matching; then the matching scores are calculated; and finally, a comprehensive evaluation is performed to obtain the best matching result.

[0086] In some specific embodiments, a multidimensional index structure is constructed, a fuzzy query algorithm is executed, similarity is calculated, matching results are merged, and the best matching item is output. Other matching algorithms can also be used to achieve the correspondence between text and database, which is not limited here.

[0087] S20532. Generate a standardized address text according to the matching result; The normalized address text refers to the address string that has been standardized.

[0088] After obtaining the database matching results, it is necessary to construct a standard address text. In some embodiments, firstly, the valid address components are extracted according to the matching results; then the address information is reorganized according to the standard format template; then the necessary missing information is supplemented; and finally, the integrity and standardization of the generated address text are verified.

[0089] In some specific embodiments, a preset address format template library is loaded, including templates for multiple levels of administrative divisions such as countries, provinces, cities, districts, counties, and streets; word segmentation and semantic analysis are performed on the input address to identify the hierarchical relationship and key information in the address; the parsed address elements are matched with the standard template, missing fields are supplemented, and incorrect expressions are corrected; the address integrity is checked to verify the validity and relevance of administrative divisions at all levels; the writing format of the address is unified, including the use of symbols, abbreviation specifications, and sorting methods; a standardized address representation is generated; it is understandable that other methods can also be used to achieve the standardization of address text, which is not limited here.

[0090] In some specific embodiments, an address rewriting rule engine is initialized, including a grammar rule base and a conversion strategy set; natural language processing technology is used to extract components and attribute information in the address; the order of each part is rearranged and the format is converted according to address specification requirements; standardization rules are applied to unify naming standards and expressions; address validity verification is performed to ensure the accuracy of conversion results; and a standardized address representation is generated. It is understandable that other methods can also be used to achieve standardization of address text, which is not limited here.

[0091] S20533. Convert the normalized address text into geographic coordinate data.

[0092] It should be noted that the principle and process of this step are similar to those of step S2053. The relevant principles and processes can be referred to step S2053 and will not be repeated here.

[0093] It can be seen that the use of a mapping database containing standard address text format data, geographic location keyword data, and address hierarchical structure data provides a comprehensive reference for coordinate conversion; then, through the data matching process, the standardization and accuracy of address information are ensured; finally, accurate geographic coordinate data is obtained through the conversion of standardized address text. This conversion mechanism based on a complete mapping database not only improves the accuracy of coordinate conversion, but also ensures the consistency of address information from different sources.

[0094] In some embodiments, S206, adjusting the focus parameter and the exposure parameter, and reacquiring new text image data on the surface of the sub-target plastic group; Among them, the focal length parameter refers to the distance setting from the camera lens to the imaging plane; the exposure parameter refers to the related settings for controlling the amount of light entering.

[0095] In some embodiments, the camera parameters are adjusted; then the shooting area is repositioned; and finally new image data is acquired.

[0096] In some specific embodiments, the image clarity index is calculated, the focal range is automatically optimized, the exposure time is adjusted, and the image is refocused and captured, which is not limited here.

[0097] S207, determining whether the new text image data is complete; It should be noted that the principle and process of this step are similar to those of step S202. The relevant principles and processes can be referred to step S202 and will not be repeated here.

[0098] S208, if the new text image data is complete, the new text image data is regarded as text image data, and a step of preprocessing the text image data is performed to obtain preprocessed image data; S209, if the new text image data is incomplete, construct an address hierarchical dependency graph to establish the relationship between address fields, including the unique correspondence between the lower-level address field and the upper-level address field; the spatial adjacent relationship between the same-level address fields; and the association relationship between multiple fields of the same address; Among them, the address hierarchy dependency graph represents a data structure that describes the relationship between the various components of the address; the unique correspondence relationship refers to the subordinate relationship between the upper and lower level addresses; the spatial adjacent relationship is used to represent the positional association between addresses of the same level; and the association relationship represents the logical connection between different fields of the address.

[0099] In some embodiments, first, complete address hierarchy data is collected; then, dependencies between the hierarchies are analyzed; then, a relationship graph data structure is constructed; and finally, mapping rules for various relationships are established.

[0100] In some specific embodiments, administrative division data, place name and address data, point of interest data and related attribute data are collected from the target area, and data cleaning and standardization are completed to form an original data set; the original data set is analyzed to extract the hierarchical structure, subordinate relationship, spatial location and related features, and generate a relationship feature set; based on the relationship feature set, a node set and edge attributes are created, the topological structure of the graph is constructed, the weights of the nodes and edges are assigned, and the initial graph structure is output; node connection rules, inheritance mapping rules and association verification rules are configured for the initial graph structure to form a rule constraint set; based on the rule constraint set, the connectivity of the graph structure is tested, the node coverage is checked, the rationality of the relationship is verified, and a complete relationship graph is output; other methods can also be used to construct an address relationship model, which is not limited here.

[0101] S2010, executing the address information completion strategy according to the address hierarchy dependency graph to obtain completed text image data, wherein the address information completion strategy includes: completing the upper-level address field based on the unique corresponding relationship; completing the same-level address field based on the spatial adjacent relationship; and completing multiple fields of the same address based on the association relationship; After the dependency graph is established, the missing address information needs to be supplemented accordingly. In some embodiments, the missing address fields are first identified; then a completion strategy is selected according to different types of relationships; then information inference and verification are performed; and finally all supplemented information is integrated.

[0102] In some specific embodiments, the known information of the current address node is first checked to determine its hierarchical type and attribute value. Then, the possible parent node is searched upward in the dependency graph. When the only matching upper node is found, the standard address information of the node is extracted to complete the missing fields of the current address. If there are multiple possible parent nodes, the completion operation is not performed. First, the spatial location information of the target address is obtained, and a reasonable search range (such as 500 meters) is set. All nodes of the same level address are searched within this range, and the spatial distance and address attribute similarity of each node to the target address are calculated. When the node with the closest distance and similarity exceeding the set threshold is found, the relevant field value of the node is used to complete the missing information of the target address. First, the known attribute set of the current address node is obtained, and matched against the preset attribute association rule library. When the precondition of a rule is met, the new attribute value is calculated according to the derivation logic defined by the rule, and used to complete the relevant fields. Finally, the rationality of the completion result is verified to ensure the accuracy of the derivation process.

[0103] Other methods may also be used to complete the address information, which are not limited here.

[0104] S2011, treating the completed text image data as text image data, and performing preprocessing on the text image data to obtain preprocessed image data.

[0105] It can be seen that the system determines the data quality by judging the integrity of text and images; when the data is found to be incomplete, it first re-collects by adjusting the focal length and exposure parameters to improve the quality of the original data; if it is still incomplete after optimization, the information completion mechanism based on the address hierarchy dependency graph is activated. By establishing the unique correspondence, spatial adjacent relationship and association relationship between address fields, the system can intelligently infer and complete the missing address information. This maximizes the integrity of the geographic location information.

[0106] In actual use, the text image data may not be collected due to various reasons. Therefore, in some specific embodiments, the above embodiments further include: S2012, collecting color features, texture features and shape features of the historical sub-target plastic group to obtain feature data; Among them, color features represent the color distribution and properties of the plastic group; texture features refer to the subtle structure and pattern of the surface; shape features are used to represent the appearance contour and geometric characteristics; feature data represents the numerical representation of the above features; historical sub-target plastic group refers to the recorded plastic samples.

[0107] S2013, constructing a contaminated sample database, wherein the contaminated sample database stores characteristic data of sub-target plastic groups with different historical geographic location information; After completing the feature collection, a systematic data storage structure needs to be established. In some embodiments, the database architecture is first designed; then the mapping relationship between the feature data and the geographic location is established; then the collected feature data is imported; and finally the index and query mechanism is established.

[0108] In some specific embodiments, a distributed storage system is used to construct a feature index and establish a position map. It is understandable that other methods can also be used to construct a sample database, which is not limited here.

[0109] S2014, calculating the similarity between the feature data of the new sub-target plastic group and the feature data in the contaminated sample database to obtain a feature similarity score; In some embodiments, feature data of the new sample is first extracted; then a suitable similarity calculation method is selected; then the new sample is compared with the samples in the database one by one; finally, the similarity scores are calculated and sorted.

[0110] In some specific embodiments, feature vectors are extracted, distance metrics are selected, feature distances are calculated, scores are normalized, and similarity results are output; other methods may also be used to calculate feature similarity, which are not limited here.

[0111] S2015. Determine the historical geographic location information with the highest feature similarity score as the geographic location information of the new sub-target plastic group.

[0112] In some embodiments, the similarity scores are first sorted; then it is verified whether the highest score exceeds a threshold; then the corresponding geographic location information is extracted and the final result is output.

[0113] It can be seen that the system has established a comprehensive feature database by collecting the color, texture and shape features of historical samples; then, by building a database of contaminated samples, a mapping relationship is established between the feature data and the geographic location information; when processing new samples, by calculating the feature similarity, its possible geographic location can be intelligently inferred. In the case of plastics without text information, this method can quickly identify and supplement the missing geographic location information, improving processing efficiency. In addition, as the number of samples in the database increases, the inference accuracy will continue to improve, realizing the self-optimization and evolution of the classification system.

[0114] The following introduces an exemplary plastic recycling tracking system 400 provided in an embodiment of the present application. Figure 4 Schematic diagram of an exemplary hardware structure of a plastic recycling tracking system 400 provided in an embodiment of the present application.

[0115] In some embodiments, the plastic recycling tracking system 400 is a computer device or the plastic recycling tracking system 400 includes a computer device. The computer device includes a processor, a memory, and a network interface connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with other external terminals or servers through a network connection. In some embodiments, the network interface can be a wired network interface, and in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, the method in the embodiment of the present application is implemented.

[0116] Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0117] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0118] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0119] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk), etc.

[0120] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A plastic recycling tracking method, characterized in that: include: Divide the target plastic group into different categories to obtain several sub-target plastic groups; Collecting geographical location information of the sub-target plastic group; generating a digital map of the plastic source based on the geographic location information; Based on the preset classification rules and according to the digital map, the plastics in different geographical locations in the sub-target plastic group are allocated to corresponding packaging agencies for processing.

2. The method according to claim 1, characterized in that The step of dividing the target plastic group into different categories to obtain a plurality of sub-target plastic groups specifically includes: Acquiring surface image data of the target plastic group; Performing image enhancement processing on the surface image data to obtain enhanced image data; extracting surface features from the enhanced image data; Detecting a contaminated area in the surface feature to obtain contamination degree data; Determining whether the contamination degree data is greater than a preset contamination threshold; If it is greater than the preset contamination threshold, collecting material characteristic data and spectral characteristic data of the target plastic group; Inputting the material feature data and the spectral feature data into a deep learning model; Generate plastic category prediction results through the deep learning model; The target plastic group is classified based on the plastic category prediction result to obtain the plurality of sub-target plastic groups.

3. The method according to claim 1, characterized in that The step of collecting the geographical location information of the sub-target plastic group specifically includes: Acquire text image data on the surface of the sub-target plastic group; Determining whether the text image data is complete; If the text image data is complete, preprocessing the text image data to obtain preprocessed image data; Performing optical character recognition processing on the pre-processed image data to obtain text recognition data; The geographical location information is extracted from the text recognition data.

4. The method according to claim 3, characterized in that The step of extracting the geographical location information from the text recognition data specifically includes: Perform keyword matching on the text recognition data to obtain a geographic location keyword matching result; Extracting text data conforming to the address format from the matching results; The text data is converted into geographic coordinate data, and the geographic coordinate data is determined as the geographic location information.

5. The method according to claim 4, characterized in that The step of converting the text data into geographic coordinate data specifically includes: Matching the text data with data in a geographic location information mapping database; the geographic location information mapping database includes standard address text format data, geographic location keyword data, and address hierarchical structure data; Generate a standardized address text according to the matching result; The normalized address text is converted into the geographic coordinate data.

6. The method according to claim 3, characterized in that After the step of determining whether the text image data is complete, the method further includes: If the text image data is incomplete, adjusting the focus parameter and the exposure parameter to reacquire new text image data of the surface of the sub-target plastic group; Determining whether the new text image data is complete; If the new text image data is complete, the new text image data is regarded as the text image data, and the step of preprocessing the text image data to obtain preprocessed image data is performed; If the new text image data is incomplete, an address hierarchical dependency graph is constructed to establish the relationship between address fields, including the unique correspondence between lower-level address fields and upper-level address fields; the spatial adjacency between address fields at the same level; and the association between multiple fields of the same address; Executing the address information completion strategy according to the address hierarchy dependency graph to obtain completed text image data, wherein the address information completion strategy includes: completing the upper-level address field based on the unique corresponding relationship; completing the same-level address field based on the spatial adjacent relationship; and completing multiple fields of the same address based on the association relationship; The completed text image data is regarded as the text image data, and the step of preprocessing the text image data to obtain preprocessed image data is performed.

7. The method according to claim 3, characterized in that After the step of extracting the geographical location information from the text recognition data, the method further includes: Collect the color features, texture features and shape features of the historical sub-target plastic group to obtain feature data; Constructing a contaminated sample database, wherein the contaminated sample database stores the characteristic data of the sub-target plastic groups with different historical geographical location information; Calculate the similarity between the feature data of the new sub-target plastic group and the feature data in the contaminated sample database to obtain a feature similarity score; The historical geographic location information with the highest feature similarity score is determined as the geographic location information of the new sub-target plastic group.

8. A plastic recycling tracking system, characterized in that: The plastic recycling tracking system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the plastic recycling tracking system to execute the method described in any one of claims 1-7.

9. A computer program product comprising instructions, characterized in that When the computer program product runs on a plastic recycling tracking system, the plastic recycling tracking system is caused to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a plastic recycling tracking system, the plastic recycling tracking system is caused to perform the method as described in any one of claims 1-7.

Images