Recycling detection method and device for plastic product recycling and computer equipment

Abstract

The invention relates to a recovery detection method and device for plastic product recovery and computer equipment. The method comprises the following steps: acquiring a current shot image and a current spectrogram on a recovery belt, so as to generate an image feature matrix and a spectral feature vector of the recovery belt; then, through a feature fusion strategy, generating fusion feature data of the recovery belt, and based on the fusion feature data, identifying recovery material distribution information and recovery value distribution information through a recovery analysis positioning model; and then according to a plastic product recovery strategy, target recovery material distribution information is screened, and a current recovery detection report of the recovery belt is generated based on the target recovery material distribution information and the recovery value distribution information. By adopting the method, the precision of recycling and detecting the universal plastic products can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data and inventory optimization, in particular to a recycling detection method and device for plastic product recycling and a computer equipment. BACKGROUND

[0002] With the rapid development of the plastic industry, plastic products are widely used in production and life, and at the same time, a large amount of plastic waste has been generated, causing serious environmental problems. Plastic product recycling is a key way to alleviate plastic pollution and realize resource recycling, and recycling detection, as a preceding link of plastic product recycling, directly determines the purity and reuse value of recycled plastics. Therefore, how to improve the efficient and high-precision detection of plastic products is the current research focus.

[0003] The existing plastic product recycling detection methods mainly include manual sorting detection and automatic detection based on single technology. Manual sorting detection relies on the experience of operators, which is not only low in efficiency, high in cost, unstable in detection accuracy, and susceptible to subjective factors, etc., and is difficult to meet the needs of large-scale industrialized recycling. The automatic detection method based on single technology cannot effectively complete the detection of plastic products without identification, contaminated plastic, blurred identification or identification blocked, and the scope of application is limited, resulting in low precision of universal plastic product recycling detection. SUMMARY

[0004] Therefore, it is necessary to provide a recycling detection method, device and computer equipment for plastic product recycling in view of the above technical problems.

[0005] In a first aspect, the present application provides a recycling detection method for plastic product recycling, comprising: obtaining a current shooting image on a recycling belt and a current spectrum image on the recycling belt, and generating an image feature matrix of the recycling belt and a spectrum feature vector of the recycling belt based on the current shooting image and the current spectrum image; based on the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt, generating fusion feature data of the recycling belt through a feature fusion strategy, and identifying recycling material distribution information and recycling value distribution information through a recycling analysis positioning model based on the fusion feature data; based on the recycling material distribution information and the recycling value distribution information, screening target recycling material distribution information according to a plastic product recycling strategy, and generating a current recycling detection report of the recycling belt based on the target recycling material distribution information and the recycling value distribution information.

[0006] Optionally, the generating, based on the current captured image and the current spectral image, an image feature matrix of the recycling belt and a spectral feature vector of the recycling belt comprises: graying the current captured image to obtain a gray image of the current captured image, and identifying structural features of each structural edge range in the gray image based on the gray image through an edge identification strategy; constructing an image feature matrix based on the structural features of each structural edge range; preprocessing the current spectral image to obtain a standard spectral image of the current spectral image, and extracting a feature spectral vector of the standard spectral image through a feature extraction strategy.

[0007] Optionally, the generating, based on the image feature matrix of the recycling belt and the spectral feature vector of the recycling belt, a fusion feature data of the recycling belt through a feature fusion strategy comprises: obtaining current pollution degree information of the recycling belt, and performing feature dimension alignment processing on the spectral feature vector and the image feature matrix to obtain a new spectral feature vector and a new image feature matrix; generating a first weight value of the new spectral feature vector and a second weight value of the new image feature matrix through a weight self-adaptive adaptation strategy based on the current pollution degree information of the recycling belt; performing weighted fusion processing on the new spectral feature vector and the new image feature matrix through the first weight value and the second weight value according to a data fusion strategy to obtain the fusion feature data.

[0008] Optionally, the identifying, based on the fusion feature data, recycling material distribution information and recycling value distribution information through a recycling analysis positioning model comprises: identifying sub-fusion features of each plastic material type through a feature classification sub-model based on the fusion feature data, and identifying a recycling belt position range corresponding to each sub-fusion feature through a position positioning sub-model based on the sub-fusion features of each plastic material type; identifying a recyclability judgment result corresponding to each sub-fusion feature and a recycling type corresponding to each sub-fusion feature through a recycling fusion judgment strategy, and generating recycling material distribution information based on the recycling type corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature; generating recycling value distribution information based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature.

[0009] Optionally, the target recycling material distribution information is screened according to a plastic product recycling strategy based on the recycling material distribution information and the recycling value distribution information, and the target recycling material distribution information includes: The recycling material distribution information is split into sub-recycling material distribution information of each recycling type through a plastic product recycling strategy; Each sub-recycling material distribution information is subjected to material value labeling processing based on the recycling value distribution information, to obtain each sub-target recycling material distribution information; Each sub-target recycling material distribution information of each recycling type is taken as the target recycling material distribution information.

[0010] Optionally, the current recycling detection report of the recycling belt is generated based on the target recycling material distribution information and the recycling value distribution information, and the current recycling detection report of the recycling belt includes: Each sub-target recycling material distribution information of each recycling type is taken as the target recycling material distribution information. Each sub-target recycling material distribution information of each recycling type is taken as the target recycling material distribution information. Each sub-target recycling material distribution information of each recycling type is taken as the target recycling material distribution information.

[0011] In a second aspect, the present application further provides a recycling detection device for plastic product recycling, which includes: An acquisition module is configured to acquire a current shooting image on a recycling belt and a current spectrum image on the recycling belt, and generate an image feature matrix of the recycling belt and a spectrum feature vector of the recycling belt based on the current shooting image and the current spectrum image. An identification module is configured to generate fusion feature data of the recycling belt through a feature fusion strategy based on the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt, and identify recycling material distribution information and recycling value distribution information through a recycling analysis positioning model based on the fusion feature data. A generation module is configured to screen target recycling material distribution information according to a plastic product recycling strategy based on the recycling material distribution information and the recycling value distribution information, and generate a current recycling detection report of the recycling belt based on the target recycling material distribution information and the recycling value distribution information.

[0012] Optionally, the acquisition module is specifically configured to: graying the current photographed image to obtain a gray image of the current photographed image, and identifying structural features of each structural edge range in the gray image based on the gray image through an edge identification strategy; constructing an image feature matrix based on the structural features of each structural edge range; preprocessing the current spectral image to obtain a standard spectral image of the current spectral image, and extracting a feature spectral vector of the standard spectral image through a feature extraction strategy.

[0013] Optionally, the identification module is specifically configured to: obtain current pollution degree information of the recycling belt, and perform feature dimension alignment processing on the spectral feature vector and the image feature matrix to obtain a new spectral feature vector and a new image feature matrix; generate a first weight value of the new spectral feature vector and a second weight value of the new image feature matrix through a weight self-adaptive adaptation strategy based on the current pollution degree information of the recycling belt; perform weighted fusion processing on the new spectral feature vector and the new image feature matrix through the first weight value and the second weight value according to a data fusion strategy to obtain fusion feature data.

[0014] Optionally, the identification module is specifically configured to: identify sub-fusion features of each plastic material type through a feature classification sub-model based on the fusion feature data, and identify a recycling belt position range corresponding to each sub-fusion feature through a position positioning sub-model based on the sub-fusion features of each plastic material type; identify a recyclability judgment result corresponding to each sub-fusion feature and a recycling type corresponding to each sub-fusion feature through a recycling fusion judgment strategy, and generate recycling material distribution information based on the recycling type corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature; generate recycling value distribution information based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature.

[0015] Optionally, the generation module is specifically configured to: split the recycling material distribution information into sub-recycling material distribution information of each recycling type through a plastic product recycling strategy; perform material value labeling processing on each sub-recycling material distribution information based on the recycling value distribution information to obtain each sub-target recycling material distribution information; use each sub-target recycling material distribution information of each recycling type as target recycling material distribution information.

[0016] Optionally, the generating module is specifically configured to: generate, based on the sub-target recycling material distribution information of each recycling type, the current recycling sorting instruction of each recycling type through the recycling control strategy of each recycling type; identify the recycling value judgment information of each recycling type based on the sub-target recycling material distribution information of each recycling type, and generate, based on the sub-target recycling material distribution information of each recycling type and the recycling value judgment information of each recycling type, the sub-recycling detection report of each recycling type according to a recycling detection report template; use the sub-recycling detection report of each recycling type as the current recycling detection report of the recycling belt.

[0017] In a third aspect, a computer device is provided. The computer device includes a memory and a processor. The memory stores a computer program. The processor implements the steps of the method in any one of the first aspect when executing the computer program.

[0018] In a fourth aspect, a computer readable storage medium is provided. The computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the steps of the method in any one of the first aspect.

[0019] In a fifth aspect, a computer program product is provided. The computer program product includes a computer program. The computer program is executed by a processor to implement the steps of the method in any one of the first aspect.

[0020] The recycling detection method, device and computer equipment for plastic product recycling described above, by acquiring a current shooting image on a recycling belt and a current spectrum image on the recycling belt, and based on the current shooting image and the current spectrum image, generating an image feature matrix of the recycling belt and a spectrum feature vector of the recycling belt; based on the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt, generating fusion feature data of the recycling belt through a feature fusion strategy, and based on the fusion feature data, identifying recycling material distribution information and recycling value distribution information through a recycling analysis positioning model; based on the recycling material distribution information and the recycling value distribution information, screening target recycling material distribution information according to a plastic product recycling strategy, and based on the target recycling material distribution information and the recycling value distribution information, generating a current recycling detection report of the recycling belt. Through the three-dimensional fusion detection mode of the shooting image, the spectrum image and the recycling analysis positioning model, the scheme can comprehensively analyze the plastic product with more pollutants, blurred image / identification and poor distinguishability, and can comprehensively judge the recycling material of the plastic product and its positioning distribution information from the angles of the image feature, the spectrum feature and the material property of the plastic product, thereby effectively improving the identification accuracy and positioning accuracy of the plastic product with more pollutants, blurred image / identification and poor distinguishability. Secondly, the traditional technical solution focuses on the technical solution of sorting efficiency and false alarm calibration, and the scheme can effectively improve the high-precision identification of the plastic product with more pollutants through the correlation feature enhancement technology from the angles of the material property and the positioning identification, thereby avoiding the problem of poor identification accuracy of the traditional edge detection in the blurred identification scene. Finally, the scheme does not need to improve the structure of the traditional recycling belt, but can directly add a shooting device and a spectrum scanning device to the traditional recycling belt, and the specific algorithm execution process only needs to be optimized and improved from the software program angle, thereby being applicable to the recycling scene of a large recycling station and the recycling scene of a small community recycling station, effectively improving the coverage of the recycling scene, and thereby improving the accuracy of the universal plastic product recycling detection. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the description of the embodiments of the present application or the related art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other related drawings according to these drawings without creative labor.

[0022] Figure 1 A flowchart of a recycling detection method for plastic product recycling in an embodiment; Figure 2 a flowchart of a recycling detection example for plastic product recycling in an embodiment; Figure 3 a structural block diagram of a recycling detection device for plastic product recycling in an embodiment; Figure 4 an internal structural diagram of a computer device in an embodiment. DETAILED DESCRIPTION

[0023] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not intended to limit the present application.

[0024] It should be noted that the terms "first", "second" and the like used in the present application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "include" and "have" and any variations thereof used in the present application are intended to cover non-exclusive inclusion. The term "multiple" used in the present application refers to two and more than two. The term "and / or" used in the present application refers to one of the options or any combination of multiple options.

[0025] The recycling detection method for plastic product recycling provided by the embodiments of the present application can be applied to a system for recycling detection of plastic product recycling. The system can be applied to a terminal, which can be, but is not limited to, various personal computers, notebook computers, medium-sized computers, etc. Among them, the terminal can comprehensively analyze the plastic product with more pollutants, blurred image / identification, and poor distinguishability through the three-dimensional fusion detection mode of the photographed image, the spectrum diagram, and the recycling analysis positioning model, and can comprehensively judge the recycling material of the plastic product and the positioning distribution information from the angles of the image features, the spectrum features, and the material properties of the plastic product, so as to effectively improve the identification accuracy and positioning accuracy of the plastic product with more pollutants, blurred image / identification, and poor distinguishability. Secondly, the traditional technical solution focuses on the technical solution of sorting efficiency and false alarm calibration, while the present solution can effectively improve the high-precision identification of the plastic product with more pollutants through the correlation feature enhancement technology from the angles of material properties and positioning identification, avoiding the problem of poor identification accuracy of the traditional edge detection in the identification of the blurred identification scene. Finally, the present solution does not need to improve the structure of the traditional recycling belt, but can directly increase the shooting device and the spectrum scanning device on the traditional recycling belt, and the specific algorithm execution process only needs to be optimized and improved from the software program angle, so that it can be applied to the recycling scene of large recycling stations and the recycling scene of small community recycling stations, effectively improving the coverage of the recycling scene, and thus improving the accuracy of the universal plastic product recycling detection.

[0026] In one exemplary embodiment, as shown in Figure 1 A recycling detection method for plastic product recycling is provided. Taking the terminal as an example, the method comprises the following steps S101 to S103. Among them: Step S101, obtaining the current photographed image on the recycling belt and the current spectrum diagram on the recycling belt, and generating the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt based on the current photographed image and the current spectrum diagram.

[0027] In this embodiment, before image acquisition, the terminal will now all the items to be recycled, transmission value recycling belt (rotating speed 0.5-2m / s) on the recycling belt will all the items to be recycled to the preset detection station, the surface of the conveyor belt positioning groove to the product for preliminary limiting, prevent transport deviation, then, the terminal trigger high pressure air blowing dust mechanism, air compressor output 0.6-1.0MPa high pressure gas, through 8 evenly distributed (detection area above 10cm) air nozzle, from multiple angles to spray the product surface dust, loose impurities and residual attachments, dust removal time control in 0.3-0.5s. Among them, the high pressure air blowing dust mechanism can also be through manual blowing way of blowing dust. When image acquisition is performed, the terminal starts the near infrared spectrum detection unit, the attenuated total reflectance spectrometer scans in the wavelength range of 400-4000 nm, the 805nm infrared laser emitter with a power of 50mW is synchronously turned on, and the signal strength of weak spectrum response materials such as PVC is enhanced; the condenser (focal length 20mm) focuses the reflected spectrum to the multi-channel optical sensor (resolution 0.1 nm), acquires multi-channel spectrum data through 5 special wavelength filters (1600nm, 1700nm, 1800nm, 2000nm, 3000nm), the sampling frequency is 10Hz, the acquisition time is 0.5s, and the original spectrum curve (i.e. the current spectrum graph) is generated. Then, the terminal synchronously starts the machine vision detection unit, the IMX219 sensor industrial camera (resolution 1920x1080) shoots at an angle of 30°, and the 12W ring-shaped LED fill light automatically adjusts the brightness according to the ambient light intensity (light range 500-2000lux), focusing on capturing the product surface recycling mark area, color distribution and contour features, continuously collecting 3 frames of images, and selecting the one with the highest clarity as the original image data (i.e. the current shooting image). Then, the terminal generates the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt based on the current shooting image and the current spectrum graph. The specific generation process will be described in detail later.

[0028] Step S102, based on the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt, the fusion feature data of the recycling belt is generated through the feature fusion strategy, and based on the fusion feature data, the recycling material distribution information and the recycling value distribution information are identified through the recycling analysis positioning model.

[0029] In this embodiment, the terminal generates fusion feature data of the recycling belt based on the image feature matrix of the recycling belt and the spectral feature vector of the recycling belt through a feature fusion strategy, and identifies recycling material distribution information and recycling value distribution information based on the fusion feature data through a recycling analysis positioning model. The feature fusion strategy is to perform data optimization and feature extraction on the image feature matrix and the spectral feature vector respectively, and then perform multi-dimensional feature fusion between one-dimensional vectors and two-dimensional feature matrices to obtain fusion feature data. The multi-dimensional feature fusion method is a feature fusion method that automatically assigns weights according to the detection scene after dynamic weighted average fusion strategy. The specific fusion process will be described in detail later, and the recycling analysis positioning model is to identify the sub-fusion feature data of each plastic material type based on the fusion feature data through an improved RCF convolutional neural network. The improved RCF convolutional neural network strengthens the association between material features and identification features through a cross-layer feature fusion module, focuses on core features through an attention mechanism, and matches each plastic material type one by one with the built-in mainstream plastic material feature library (including PE, Polyethylene, PP, Polypropylene, PS, Polystyrene, PVC, Polyvinyl Chloride, PET, Polyethylene Terephthalate, etc.) to identify the sub-fusion feature data of each plastic material type. Then, through the GPS module, the position range corresponding to each sub-fusion feature data in the latitude and longitude coordinates of the recycling belt is identified. Finally, through the cloud database, the recycling grade, recyclability, sorting requirements, and prohibited recycling situations of each sub-fusion feature data are identified, and the recyclability and specific recycling categories (such as general recyclable, special recyclable, and non-recyclable) of plastic products are determined based on the parameters, so as to obtain recycling material distribution information and recycling value distribution information. The recycling material distribution information is the position distribution information of each recycling category except the non-recyclable category in the recycling belt. The recycling value distribution information is the distribution information of the material value of the sub-fusion feature data corresponding to each position range. The specific identification process will be described in detail later.

[0030] In step S103, based on the recycling material distribution information and the recycling value distribution information, the target recycling material distribution information is screened according to the plastic product recycling strategy, and the current recycling detection report of the recycling belt is generated based on the target recycling material distribution information and the recycling value distribution information.

[0031] In this embodiment, the terminal screens target recycling material distribution information according to the plastic product recycling strategy based on the recycling material distribution information and the recycling value distribution information. The target recycling material distribution information includes sub-target recycling material distribution information of each recycling type, and the recycling type is a recycling material corresponding to a plastic material type, i.e., a recycling type preset by the worker on the terminal, and each recycling type corresponds to one or more plastic material types.

[0032] Then, the terminal generates a current recycling detection report of the recycling belt based on the target recycling material distribution information and the recycling value distribution information. The current recycling detection report includes sub-recycling detection reports of each recycling type, and each sub-recycling detection report includes sub-target recycling material distribution information of each recycling type and recycling value judgment information of each recycling type. The recycling value judgment information is distribution information of recycling values of all recycling items on the recycling belt. The sub-target recycling material distribution information includes position distribution information of all recycling items of the recycling type on the recycling belt and recycling reason information thereof. The recycling reason of each recycling item includes the recycling grade, the plastic material type, and the judgment result of whether it is recyclable.

[0033] Based on the above scheme, by using the three-dimensional fusion detection mode of the shooting image, the spectrum image, and the recycling analysis positioning model, the comprehensive analysis of the plastic products with more pollutants, blurred images / identifiers, and poor recognizability can be performed, the recycling material of the plastic product and the positioning distribution information thereof can be comprehensively judged from the angles of the plastic product image features, the spectrum features, and the material properties, and thus the identification accuracy and the positioning accuracy of the plastic products with more pollutants, blurred images / identifiers, and poor recognizability can be effectively improved. Secondly, the traditional technical scheme focuses on the technical scheme of sorting efficiency and false alarm calibration, and the present scheme can effectively improve the high-precision identification of the plastic products with more pollutants by using the correlation feature enhancement technology from the angles of the material properties and the positioning identifier, thereby avoiding the problem of poor identification accuracy of the traditional edge detection in the identifier blurred scene. Finally, the present scheme does not need to improve the structure of the traditional recycling belt, but directly adds a shooting device and a spectrum scanning device to the traditional recycling belt, and the specific algorithm execution process only needs to be optimized and improved from the software program angle, so that it can be applied to both large recycling stations and small community recycling stations, effectively improving the coverage of the recycling scene, and thus improving the accuracy of the universal plastic product recycling detection.

[0034] Optionally, based on the current captured image and the current spectral image, an image feature matrix of the recycling band and a spectral feature vector of the recycling band are generated, including: performing grayscale processing on the current captured image to obtain a grayscale image of the current captured image, and identifying structural features of each structural edge range in the grayscale image based on the grayscale image through an edge identification strategy; constructing the image feature matrix based on the structural features of each structural edge range; preprocessing the current spectral image to obtain a standard spectral image of the current spectral image, and extracting a feature spectral vector of the standard spectral image through a feature extraction strategy.

[0035] In this embodiment, the terminal performs grayscale processing on the current captured image to obtain a grayscale image of the current captured image, and identifies structural features of each structural edge range in the grayscale image based on the grayscale image through an edge identification strategy. Then, the terminal constructs the image feature matrix based on the structural features of each structural edge range. Specifically, the terminal uses a weighted average method with weight proportions R:G:B=0.299:0.587:0.114 to convert the captured image into a grayscale image. The edge identification recognition strategy is to perform edge detection (threshold range 50-150) through a Canny operator to segment out the product contour; and to use a template matching algorithm to identify the recycling mark and extract a 64-dimensional image feature matrix (including 32-dimensional mark features, 16-dimensional color features, and 16-dimensional shape features). The image feature matrix includes sub-image feature matrices of different position ranges of the recycling band. The position range is the latitude and longitude range corresponding to each grid after grid processing of the recycling band. The size of the grid is preset in the terminal. For example, 0.5 cm, 1 cm, 2 cm, 3 cm, etc.

[0036] Then, the terminal preprocesses the current spectral image to obtain a standard spectral image of the current spectral image, and extracts a feature spectral vector of the standard spectral image through a feature extraction strategy. The preprocessing method of the current spectral image is to use a Savitzky-Golay algorithm to smooth and denoise the original spectral data (window size 5 points), complete baseline correction through a polynomial fitting method, and eliminate background interference and instrument noise. The feature extraction strategy is to extract a 128-dimensional feature spectral vector based on the wavelength, intensity, and half-peak width of the feature absorption peak to represent the molecular structure features of the plastic material. The feature spectral vector includes sub-feature spectral vectors of different position ranges.

[0037] Based on the above scheme, the purification and feature extraction of the spectral image and the captured image are performed respectively, thereby effectively improving the identification comprehensiveness of each plastic product on the recycling band.

[0038] Optionally, based on the image feature matrix of the recycling belt and the spectral feature vector of the recycling belt, the fusion feature data of the recycling belt is generated through a feature fusion strategy, including: obtaining the current pollution degree information of the recycling belt, and performing feature dimension alignment processing on the spectral feature vector and the image feature matrix to obtain a new spectral feature vector and a new image feature matrix; based on the current pollution degree information of the recycling belt, the first weight value of the new spectral feature vector and the second weight value of the new image feature matrix are generated through a weight self-adaptive adaptation strategy; based on the new spectral feature vector and the new image feature matrix, the first weight value and the second weight value are weighted and fused according to the data fusion strategy to obtain the fusion feature data.

[0039] In this embodiment, the terminal obtains the current pollution degree information of the recycling belt, and performs feature dimension alignment processing on the spectral feature vector and the image feature matrix to obtain a new spectral feature vector and a new image feature matrix. Among them, since the feature spectral vector is 128-dimensional and the image feature matrix is 64-dimensional, for each position range of the recycling belt, the terminal first aligns the to-be-fused dimension interface of the image feature matrix with the to-be-fused dimension interface of the feature spectral vector, thereby forming a corresponding relationship between different dimension interfaces, and then the terminal adds the corresponding relationship to the spectral feature vector and the image feature matrix respectively to obtain the new spectral feature vector and the new image feature matrix. The current pollution degree is the pollution degree value obtained by the staff after judging the pollution degree of each recycling article of the recycling belt (from 1% to 100%).

[0040] Then, the terminal generates the first weight value of the new spectral feature vector and the second weight value of the new image feature matrix based on the current pollution degree information of the recycling belt through a weight self-adaptive adaptation strategy. Among them, the weight range of the new spectral feature vector is 60%-70%, and the weight range of the new image feature matrix is 30%-40%, the terminal performs normalization processing on the pollution degree value to obtain a pollution degree influence weight value, and based on the influence weight value, the weight value is selected in the above weight range to obtain the first weight value of the new spectral feature vector and the second weight value of the new image feature matrix. For example, the pollution degree value is 85%, the normalized influence weight value is 8.5%, and the higher the pollution degree, the higher the weight value of the spectral feature vector, and the lower the pollution degree, the higher the weight value of the image feature matrix, so in the above example, the first weight value of the new spectral feature vector is 68.5%, and the second weight value of the new image feature matrix is 31.5%.

[0041] Then, the terminal performs weighted fusion processing on the new spectral feature vector and the new image feature matrix according to a data fusion strategy through the first weight value and the second weight value to obtain fusion feature data. The fusion feature data is 192-dimensional comprehensive feature data.

[0042] Based on the above scheme, the comprehensive feature data is obtained by performing dimension alignment on vectors and matrices of different dimensions and performing dynamic weighted average according to pollution weights, and then performing dimension fusion on the vectors and matrices, thereby improving the accuracy of the obtained comprehensive feature data.

[0043] Optionally, based on the fusion feature data, the recycling material distribution information and the recycling value distribution information are identified through a recycling analysis positioning model, including: based on the fusion feature data, sub-fusion features of each plastic material type are identified through a feature classification sub-model, and based on the sub-fusion features of each plastic material type, a recycling zone position range corresponding to each sub-fusion feature is identified through a position positioning sub-model; through a recycling fusion judgment strategy, a recyclability judgment result corresponding to each sub-fusion feature and a recycling type corresponding to each sub-fusion feature are identified, and based on the recycling type corresponding to each sub-fusion feature and the recycling zone position range corresponding to each sub-fusion feature, the recycling material distribution information is generated; based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling zone position range corresponding to each sub-fusion feature, the recycling value distribution information is generated.

[0044] In this embodiment, based on the fusion feature data, the terminal identifies sub-fusion features of each plastic material type through a feature classification sub-model, and based on the sub-fusion features of each plastic material type, a recycling zone position range corresponding to each sub-fusion feature is identified through a position positioning sub-model. The sub-fusion feature is a fusion feature data belonging to different plastic material types in fusion feature data of different position ranges. Specifically, the terminal first generates a material type corresponding to each fusion feature data and a confidence degree of the material type corresponding to each fusion feature data, and when the confidence degree is ≥95%, the material type is directly confirmed to be correct; when the confidence degree is between 80% and 95%, the terminal calls a backup feature comparison algorithm to verify the material type again, and returns to the confidence degree judgment process; when the confidence degree is <80%, the terminal marks the fusion feature data as “to be reviewed” and triggers a subsequent artificial intervention prompt.

[0045] Then, the terminal identifies the recyclability judgment result corresponding to each sub-fusion feature and the recycling type corresponding to each sub-fusion feature through the recycling fusion judgment strategy, and generates recycling material distribution information based on the recycling type corresponding to each sub-fusion feature and the recycling band position range corresponding to each sub-fusion feature. Specifically, the terminal sends a request to the cloud regional recycling standard database by uploading the GPS coordinates of the recycling band, carrying the position information and the confirmed plastic material type. The recycling fusion judgment strategy is to match the corresponding administrative region recycling standard according to the position information through the cloud database, wherein the administrative region recycling standard is the division, classification, sorting requirement, and prohibited recycling situation of the recyclable recycling type in each region, province, or place. Then, the terminal identifies the recyclability judgment result and the recycling type of each sub-fusion feature based on the above-mentioned restriction standards. The recyclability judgment result is the recycling level, recyclability, sorting requirement, and judgment result of whether it belongs to the prohibited recycling situation. The recycling type can be, but is not limited to, general recyclable, special recyclable, and non-recyclable recycling type. Finally, the terminal generates recycling value distribution information based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling band position range corresponding to each sub-fusion feature. Among them, the special recyclable value is the highest, followed by the general recyclable, and the non-recyclable value is the lowest (0).

[0046] Based on the above scheme, the recyclability judgment result of each sub-fusion feature can be obtained in real time through GPS backtracking, which improves the recyclability judgment accuracy of the plastic material type of the recycling band.

[0047] Optionally, based on the recycling material distribution information and the recycling value distribution information, the target recycling material distribution information is screened according to the plastic product recycling strategy, including: splitting the recycling material distribution information into sub-recycling material distribution information of each recycling type through the plastic product recycling strategy; performing material value labeling processing on each sub-recycling material distribution information based on the recycling value distribution information to obtain each sub-target recycling material distribution information; and taking each sub-target recycling material distribution information of each recycling type as the target recycling material distribution information.

[0048] In this embodiment, the terminal divides the recycled material distribution information into sub-recycled material distribution information of each recycling type through the plastic product recycling strategy. Then, the terminal labels the material value of each sub-recycled material distribution information based on the recycling value distribution information to obtain each sub-target recycled material distribution information. Finally, the terminal takes the sub-target recycled material distribution information of each recycling type as the target recycled material distribution information. Among them, the sub-recycled material distribution information of the non-recyclable type is deleted, and only the sub-recycled material distribution information of the general recyclable type and the special recyclable type is retained.

[0049] Based on the above scheme, the sub-target recycled material distribution information of the recyclable recycling type can be quickly screened, and the recycling screening accuracy, efficiency and comprehensiveness of the plastic material product are improved.

[0050] Optionally, based on the target recycled material distribution information and the recycling value distribution information, a current recycling detection report of the recycling belt is generated, including: based on the sub-target recycled material distribution information of each recycling type, generating a current recycling sorting instruction of each recycling type through a recycling control strategy of each recycling type; based on the sub-target recycled material distribution information of each recycling type, identifying recycling value judgment information of each recycling type, and based on the sub-target recycled material distribution information of each recycling type and the recycling value judgment information of each recycling type, generating a sub-recycling detection report of each recycling type according to a recycling detection report template; taking the sub-recycling detection report of each recycling type as the current recycling detection report of the recycling belt.

[0051] In this embodiment, the terminal generates a current recycling sorting instruction of each recycling type based on the sub-target recycled material distribution information of each recycling type through a recycling control strategy of each recycling type. The current recycling sorting instruction is an object selection instruction of different recycling sorting devices (for example, a recycling sorting mechanical arm). The recycling control strategy is a control strategy of different recycling sorting devices.

[0052] Then, the terminal identifies the recycling value judgment information of each recycling type based on the sub-target recycled material distribution information of each recycling type, and generates a sub-recycling detection report of each recycling type according to a recycling detection report template based on the sub-target recycled material distribution information of each recycling type and the recycling value judgment information of each recycling type. Finally, the terminal takes the sub-recycling detection report of each recycling type as the current recycling detection report of the recycling belt. When generating the recycling detection report, the sub-recycling detection report of the non-recyclable type also needs to be generated, and the sub-recycling detection report includes the sub-recycled material distribution information of the recycling type and the recyclability judgment result of each sub-fusion feature in the sub-recycling value distribution information of the recycling type.

[0053] Based on the above scheme, when generating the recycling detection report, the material type, confidence, recyclability, recycling category and judgment basis are marked, and the reason for the recyclable material not meeting the regional recycling standard is additionally marked, thereby improving the comprehensiveness and accuracy of recycling detection of plastic products.

[0054] The application also provides a recycling detection example for recycling of plastic products, as shown in Figure 2 The specific processing process includes the following steps: Step S201, acquiring a current shooting image on the recycling belt and a current spectrum image on the recycling belt.

[0055] Step S202, performing gray-scale processing on the current shooting image to obtain a gray-scale image of the current shooting image, and identifying the structural features of each structural edge range in the gray-scale image based on the gray-scale image through an edge identification strategy.

[0056] Step S203, constructing an image feature matrix based on the structural features of each structural edge range.

[0057] Step S204, preprocessing the current spectrum image to obtain a standard spectrum image of the current spectrum image, and extracting a feature spectrum vector of the standard spectrum image through a feature extraction strategy.

[0058] Step S205, acquiring current pollution degree information of the recycling belt, and performing feature dimension alignment processing on the spectrum feature vector and the image feature matrix to obtain a new spectrum feature vector and a new image feature matrix.

[0059] Step S206, generating a first weight value of the new spectrum feature vector and a second weight value of the new image feature matrix based on the current pollution degree information of the recycling belt through a weight self-adaptive adaptation strategy.

[0060] Step S207, performing weighted fusion processing according to a data fusion strategy based on the new spectrum feature vector and the new image feature matrix through the first weight value and the second weight value to obtain fusion feature data.

[0061] Step S208, identifying sub-fusion features of each plastic material type based on the fusion feature data through a feature classification sub-model, and identifying a recycling belt position range corresponding to each sub-fusion feature based on the sub-fusion features of each plastic material type through a position positioning sub-model.

[0062] Step S209, identifying a recyclability judgment result corresponding to each sub-fusion feature and a recycling type corresponding to each sub-fusion feature through a recycling fusion judgment strategy, and generating recycling material distribution information based on the recycling type corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature.

[0063] Step S210, based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling zone position range corresponding to each sub-fusion feature, generating recycling value distribution information.

[0064] Step S211, splitting the recycling material distribution information into sub-recycling material distribution information of each recycling type through the plastic product recycling strategy.

[0065] Step S212, based on the recycling value distribution information, performing material value labeling processing on each sub-recycling material distribution information to obtain each sub-target recycling material distribution information.

[0066] Step S213, taking the sub-target recycling material distribution information of each recycling type as the target recycling material distribution information.

[0067] Step S214, based on the sub-target recycling material distribution information of each recycling type, generating the current recycling sorting instruction of each recycling type through the recycling control strategy of each recycling type.

[0068] Step S215, based on the sub-target recycling material distribution information of each recycling type, identifying the recycling value judgment information of each recycling type, and based on the sub-target recycling material distribution information of each recycling type and the recycling value judgment information of each recycling type, generating the sub-recycling detection report of each recycling type according to the recycling detection report template.

[0069] Step S216, taking the sub-recycling detection report of each recycling type as the current recycling detection report of the recycling zone.

[0070] It should be understood that although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow indication, these steps are not necessarily executed in sequence according to the arrow indication. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which are not necessarily executed at the same time but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential but can be alternately or alternately executed with at least part of other steps or steps or stages in other steps. It can be understood that the steps in different embodiments can be freely combined as needed, and various non-contradictory schemes formed by the combination are within the scope of protection of the present application.

[0071] Based on the same inventive concept, the application further provides a recycling detection device for plastic product recycling for implementing the recycling detection method for plastic product recycling as described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more recycling detection device embodiments for plastic product recycling provided below can refer to the limitations of the recycling detection method for plastic product recycling described above, which will not be repeated here.

[0072] In one exemplary embodiment, as shown in Figure 3 A recycling detection device for plastic product recycling is provided, comprising: an acquisition module 310, an identification module 320, and a generation module 330, wherein: The acquisition module 310 is configured to acquire a current shooting image on a recycling belt and a current spectrum image on the recycling belt, and generate an image feature matrix of the recycling belt and a spectrum feature vector of the recycling belt based on the current shooting image and the current spectrum image; The identification module 320 is configured to generate fusion feature data of the recycling belt by a feature fusion strategy based on the image feature matrix of the recycling belt and the spectrum feature vector of the recycling belt, and identify recycling material distribution information and recycling value distribution information by a recycling analysis positioning model based on the fusion feature data; The generation module 330 is configured to filter target recycling material distribution information according to a plastic product recycling strategy based on the recycling material distribution information and the recycling value distribution information, and generate a current recycling detection report of the recycling belt based on the target recycling material distribution information and the recycling value distribution information.

[0073] Optionally, the acquisition module 310 is specifically configured to: perform gray scale processing on the current shooting image to obtain a gray scale image of the current shooting image, and identify structural features of each structural edge range in the gray scale image by an edge identification strategy based on the gray scale image; construct an image feature matrix based on the structural features of each structural edge range; perform preprocessing on the current spectrum image to obtain a standard spectrum image of the current spectrum image, and extract a feature spectrum vector of the standard spectrum image by a feature extraction strategy.

[0074] Optionally, the identification module 320 is specifically configured to: acquire current pollution degree information of the recycling belt, and perform feature dimension alignment processing on the spectrum feature vector and the image feature matrix to obtain a new spectrum feature vector and a new image feature matrix; generate the first weight value of the new spectral feature vector and the second weight value of the new image feature matrix based on the current pollution degree information of the recycling belt through a weight self-adaptive adaptation strategy; generate the fusion feature data through weighted fusion processing according to a data fusion strategy based on the new spectral feature vector and the new image feature matrix through the first weight value and the second weight value.

[0075] Optionally, the identification module 320 is specifically configured to: identify the sub-fusion features of each plastic material type based on the fusion feature data through a feature classification sub-model, and identify the recycling belt position range corresponding to each sub-fusion feature based on the sub-fusion features of each plastic material type through a position positioning sub-model; identify the recyclability judgment result corresponding to each sub-fusion feature and the recycling type corresponding to each sub-fusion feature through a recycling fusion judgment strategy, and generate the recycling material distribution information based on the recycling type corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature; generate the recycling value distribution information based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature.

[0076] Optionally, the generation module 330 is specifically configured to: split the recycling material distribution information into sub-recycling material distribution information of each recycling type through a plastic product recycling strategy; perform material value labeling processing on each sub-recycling material distribution information based on the recycling value distribution information to obtain each sub-target recycling material distribution information; use the sub-target recycling material distribution information of each recycling type as the target recycling material distribution information.

[0077] Optionally, the generation module 330 is specifically configured to: generate the current recycling sorting instruction of each recycling type based on the sub-target recycling material distribution information of each recycling type through a recycling control strategy of each recycling type; identify the recycling value judgment information of each recycling type based on the sub-target recycling material distribution information of each recycling type, and generate a sub-recycling detection report of each recycling type according to a recycling detection report template based on the sub-target recycling material distribution information of each recycling type and the recycling value judgment information of each recycling type; use the sub-recycling detection report of each recycling type as the current recycling detection report of the recycling belt.

[0078] The various modules in the aforementioned recycling and testing device for plastic product recycling can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0079] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a recycling detection method for plastic product recycling. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0080] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0081] In one exemplary embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of a beer warehouse inventory optimization method.

[0082] In one embodiment, a computer readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, implements the steps of the inventory optimization method for a beer brewery.

[0083] In one embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the inventory optimization method for a beer brewery.

[0084] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0085] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. The volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., without being limited thereto.

[0086] The technical features of the above embodiments can be combined in any manner. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, but as long as the combinations of the technical features do not exist, they should be considered as the scope of the present application.

[0087] The above-described embodiments are merely illustrative of several embodiments of the present application, which are described in more detail and in a specific manner, but should not be construed as limiting the scope of the patent of the present application. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A recycling detection method for plastic product recycling, characterized by, The method comprises: acquiring a current photographed image on a recycling belt and a current spectral image on the recycling belt, and generating an image feature matrix of the recycling belt and a spectral feature vector of the recycling belt based on the current photographed image and the current spectral image; based on the image feature matrix of the recycling belt and the spectral feature vector of the recycling belt, generating fusion feature data of the recycling belt through a feature fusion strategy, and identifying recycling material distribution information and recycling value distribution information through a recycling analysis positioning model based on the fusion feature data; based on the recycling material distribution information and the recycling value distribution information, screening target recycling material distribution information according to a plastic product recycling strategy, and generating a current recycling detection report of the recycling belt based on the target recycling material distribution information and the recycling value distribution information.

2. The method of claim 1, wherein, The method comprises: graying the current photographed image to obtain a gray image of the current photographed image, and identifying structural features of each structural edge range in the gray image through an edge identification strategy based on the gray image; constructing an image feature matrix based on the structural features of each structural edge range; preprocessing the current spectral image to obtain a standard spectral image of the current spectral image, and extracting a feature spectral vector of the standard spectral image through a feature extraction strategy.

3. The method of claim 2, wherein, The method comprises: acquiring current pollution degree information of the recycling belt, and performing feature dimension alignment processing on the spectral feature vector and the image feature matrix to obtain a new spectral feature vector and a new image feature matrix; based on the current pollution degree information of the recycling belt, generating a first weight value of the new spectral feature vector and a second weight value of the new image feature matrix through a weight self-adaptive adaptation strategy; based on the new spectral feature vector and the new image feature matrix, performing weighted fusion processing according to a data fusion strategy through the first weight value and the second weight value to obtain fusion feature data.

4. The method of claim 3, wherein, The method comprises: based on the fusion feature data, identifying sub-fusion features of each plastic material type through a feature classification sub-model, and based on the sub-fusion features of each plastic material type, identifying a recycling belt position range corresponding to each sub-fusion feature through a position positioning sub-model; identifying a recyclability judgment result corresponding to each sub-fusion feature and a recycling type corresponding to each sub-fusion feature through a recycling fusion determination strategy, and generating recycling material distribution information based on the recycling type corresponding to each sub-fusion feature and the recycling belt position range corresponding to each sub-fusion feature; The recycling value distribution information is generated based on the recyclability judgment result corresponding to each sub-fusion feature and the recycling zone position range corresponding to each sub-fusion feature.

5. The method of claim 4, wherein, The target recycling material distribution information is screened according to a plastic product recycling strategy based on the recycling material distribution information and the recycling value distribution information, and the target recycling material distribution information includes: The recycling material distribution information is split into sub-recycling material distribution information of each recycling type through a plastic product recycling strategy; Material value labeling processing is performed on each sub-recycling material distribution information based on the recycling value distribution information to obtain sub-target recycling material distribution information of each recycling type; Each sub-target recycling material distribution information of each recycling type is used as target recycling material distribution information.

6. The method of claim 5, wherein, The current recycling detection report of the recycling zone is generated based on the target recycling material distribution information and the recycling value distribution information, and the current recycling detection report includes: Current recycling sorting instructions of each recycling type are generated through a recycling control strategy of each recycling type based on the sub-target recycling material distribution information of each recycling type; Recycling value judgment information of each recycling type is identified based on the sub-target recycling material distribution information of each recycling type, and a sub-recycling detection report of each recycling type is generated according to a recycling detection report template based on the sub-target recycling material distribution information of each recycling type and the recycling value judgment information of each recycling type; Each sub-recycling detection report of each recycling type is used as the current recycling detection report of the recycling zone.

7. A recycling detection device for recycling of plastic articles, characterized in that The device includes: An acquisition module is configured to acquire a current image on a recycling zone and a current spectrum image on the recycling zone, and generate an image feature matrix of the recycling zone and a spectrum feature vector of the recycling zone based on the current image and the current spectrum image; An identification module is configured to generate fusion feature data of the recycling zone through a feature fusion strategy based on the image feature matrix of the recycling zone and the spectrum feature vector of the recycling zone, and identify recycling material distribution information and recycling value distribution information through a recycling analysis positioning model based on the fusion feature data; A generation module is configured to screen target recycling material distribution information according to a plastic product recycling strategy based on the recycling material distribution information and the recycling value distribution information, and generate a current recycling detection report of the recycling zone based on the target recycling material distribution information and the recycling value distribution information.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to implement the steps of the method of any one of claims 1 to 6.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 6.

Images