An intelligent material identification system and method based on multi-source heterogeneous data fusion

Abstract

The present invention discloses an intelligent material identification system and method based on multi-source heterogeneous data fusion, and relates to the field of intelligent material identification management technology. The intelligent material identification system based on multi-source heterogeneous data fusion includes: an identification-related data acquisition module, an acquisition quality assessment and judgment module, a data processing assessment and judgment module, and an environmental optimization comprehensive judgment module. The present invention collects identification-related data, then evaluates the acquisition quality of the identification-related data to determine whether to perform identification-related data processing, then obtains the identification processing compliance evaluation index after performing identification-related data processing and determines whether to perform material identification-related environmental optimization, and finally makes a comprehensive judgment on whether to perform material identification, thereby improving the data quality in the intelligent material identification process of multi-source heterogeneous data fusion, and solving the problem of low data quality in the intelligent material identification process of multi-source heterogeneous data fusion in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent material identification management, and in particular to an intelligent material identification system and method based on multi-source heterogeneous data fusion. Background Art

[0002] With the development of smart manufacturing, the Internet of Things, and big data technologies, the manufacturing industry is constantly transforming towards digitalization and intelligence. As a basic link in production and logistics management, material identification is constantly expanding its application scenarios. Traditional material identification usually relies on manual operation and a single data source (such as barcodes), which is susceptible to factors such as human errors, signal interference, and data loss. The intelligent material identification system with multi-source heterogeneous data fusion integrates multiple data sources such as images, sensors, and location information, greatly improving the accuracy and efficiency of identification, laying a solid technical foundation for the digitalization and intelligent transformation of the industrial and logistics fields.

[0003] Existing intelligent material identification systems achieve efficient material identification and management by integrating multiple technical means and data sources. Material information is usually collected through visual recognition, radio frequency identification, barcodes, QR codes, sensor data and other methods, and processed through data fusion and artificial intelligence algorithms to achieve automatic identification and management of materials. The synergy of multiple technologies enables the existing material identification system to meet the needs of modern manufacturing and supply chain management for efficient, accurate and automated management.

[0004] For example, a patent application with publication number: CN115310774A discloses a method and system for perceiving environmental resources at a work site, including: receiving work environment data collected by at least one wearable smart device of a worker that is communicatively connected to an edge computing device; performing multi-source heterogeneous data analysis based on the multi-parameter data of the work environment, and organically integrating different types of data from multiple dimensions; transmitting the integrated multi-parameter data of the work environment to an edge intelligent server; receiving the work environment hazard identification and prediction results fed back by the edge intelligent server and transmitting them to the wearable smart device of the worker within a preset range of the hazard's location.

[0005] For example, the invention patent with announcement number CN112053085B announces an airport surface operation management system and method based on digital twins, including: an airport operation scene module, a digital twin airport module, a business system service module and a data fusion module. The airport operation scene module is used to splice and fuse the real-time video information of multiple cameras to obtain panoramic fusion video information, and generate structured sign information, and coordinately respond to the structured sign information to the digital twin airport module. The digital twin airport module is used to build an airport surface operation digital environment that is exactly the same as the airport operation situation according to the fusion results of the data fusion module and the structured sign information, and generate target information and superimpose the target information, air traffic control information, and airport information into the virtual scene, and coordinately respond to the information superimposed into the virtual scene through the digital twin scene and the panoramic video. The data fusion module is used to match and fuse the target information, air traffic control information, and airport information on the virtual scene using a multi-source data fusion method, and analyze the matching and fusion results to realize active decision-making for airport management and complete the airport surface operation management based on digital twins.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] In intelligent material identification, for example, the process of quality inspection and identification of raw tobacco leaves involves many different types of data with dimensional differences. In addition, if there are abnormalities in data collection, processing and transmission, it may cause the detection results to lag and affect the overall identification effect.

[0008] Another thing that needs to be considered is that changes in environmental conditions during the intelligent material identification process can also lead to abnormal material conditions (such as mold), and there is a problem of low data quality during the intelligent material identification process that integrates multi-source heterogeneous data. Summary of the invention

[0009] The embodiments of the present application provide an intelligent material identification system and method based on multi-source heterogeneous data fusion, thereby solving the problem of low data quality in the intelligent material identification process of multi-source heterogeneous data fusion in the prior art, and achieving improved data quality in the intelligent material identification process of multi-source heterogeneous data fusion.

[0010] The embodiment of the present application provides an intelligent material identification system based on multi-source heterogeneous data fusion, including: an identification-related data acquisition module, an acquisition quality assessment and judgment module, a data processing assessment and judgment module, and an environmental optimization comprehensive judgment module; wherein the identification-related data acquisition module is used to collect identification-related data of preset materials within a preset time interval, and the identification-related data includes material identification image data, material identification spectral data, and material identification environmental data; the acquisition quality assessment and judgment module is used to evaluate the acquisition quality of the identification-related data to obtain an identification acquisition quality assessment index, and judge whether to execute identification-related data processing based on the identification acquisition quality assessment index and a preset acquisition quality assessment threshold range, and the identification acquisition quality assessment module is used to evaluate the acquisition quality of the identification-related data to obtain an identification acquisition quality assessment index, and judge whether to execute identification-related data processing based on the identification acquisition quality assessment index and a preset acquisition quality assessment threshold range. The quantity evaluation index is used to quantify the collection quality of identification-related data; the data processing evaluation judgment module is used to obtain the identification processing compliance evaluation index after executing identification-related data processing, and judge whether to execute material identification-related environmental optimization based on the identification processing compliance evaluation index and the preset processing compliance threshold range, and the identification processing compliance evaluation index is used to comprehensively quantify the compliance degree of identification-related data processing; the environmental optimization comprehensive judgment module is used to obtain the identification environment optimization evaluation index after executing material identification-related environmental optimization, and comprehensively judge whether to execute material identification based on the identification collection quality evaluation index and the identification processing compliance evaluation index, and the identification environment optimization evaluation index is used to quantitatively evaluate the effect of material identification-related environmental optimization.

[0011] Furthermore, the specific steps of evaluating the collection quality of the identification related data to obtain the identification collection quality evaluation index are as follows: obtaining image quality evaluation related data, spectral quality evaluation related data and environmental quality evaluation related data of a preset material within a preset time interval, the image quality evaluation related data including edge gradient value, material image mean, material image standard deviation and color saturation mean, the spectral quality evaluation related data including spectral intensity maximum value and spectral intensity minimum value, the environmental quality evaluation related data including ambient temperature measurement value and ambient humidity measurement value; combining the image quality evaluation related data, spectral quality evaluation related data and environmental quality evaluation related data with quality evaluation reference data obtained from a preset database to obtain a quality related evaluation value, the quality related The evaluation values ​​include image quality evaluation values, spectral quality evaluation values ​​and environmental quality evaluation values. The quality evaluation reference data include quality evaluation reference values ​​and quality evaluation influencing factors. The quality evaluation reference values ​​include image clarity reference values, color saturation maximum reference values, spectral intensity maximum reference deviations, ambient temperature reference values ​​and ambient humidity reference values. The quality evaluation influencing factors include image clarity influencing factors, image noise influencing factors, image color influencing factors, spectral intensity influencing factors, ambient temperature influencing factors and ambient humidity influencing factors. The quality-related evaluation values ​​and the acquisition quality evaluation weights obtained from a preset database are combined to obtain an identification acquisition quality evaluation index. The acquisition quality evaluation weights include image quality evaluation weights, spectral quality evaluation weights and environmental quality evaluation weights.

[0012] Furthermore, the method for obtaining the recognition acquisition quality evaluation index is as follows:

[0013] ;

[0014] ;

[0015] ;

[0016] ;

[0017] In the formula, represents the recognition and collection quality evaluation index of the jth preset material, Indicates the number of the preset material. , Indicates the total quantity of the preset material. represents the image quality assessment weight, represents the spectral quality assessment weight, represents the environmental quality assessment weight, represents the image quality evaluation value of the j-th preset material, represents the spectral quality evaluation value of the j-th preset material, represents the environmental quality assessment value of the jth preset material, represents the image clarity impact factor, represents the image noise impact factor, represents the image color impact factor, Represents the number of horizontal columns of image pixels, , Represents the total number of horizontal columns of image pixels, Represents the number of horizontal rows of image pixels, , Indicates the total number of horizontal rows of image pixels, Indicates the edge gradient value of the image pixel of the j-th preset material located at the y-th row and x-th column, Indicates the image clarity reference value, represents the mean value of the material image of the j-th preset material, represents the material image standard deviation of the j-th preset material, represents the mean color saturation of the jth preset material, Indicates the maximum reference value of color saturation. represents the spectral intensity influencing factor, Indicates the maximum spectral intensity of the jth preset material, Indicates the minimum spectral intensity of the j-th preset material, Indicates the maximum reference deviation of the spectral intensity, represents the ambient temperature measurement value of the jth preset material, Indicates the ambient temperature reference value, Represents the measured value of the ambient humidity of the jth preset material, Indicates the reference value of ambient humidity. represents the environmental temperature influence factor, Represents the environmental humidity impact factor.

[0018] Furthermore, the specific steps of determining whether to execute identification-related data processing based on the identification and acquisition quality assessment index and the preset acquisition quality assessment threshold range are as follows: Compare the identification and acquisition quality assessment index with the preset acquisition quality assessment threshold range: If the identification and acquisition quality assessment index is within the preset acquisition quality assessment threshold range, identification-related data processing is not executed, and at the same time, the identification-related data and the identification and acquisition quality assessment index are backed up in real time and uploaded to cloud storage; If the identification and acquisition quality assessment index exceeds the preset acquisition quality assessment threshold range, identification-related data processing is executed.

[0019] Furthermore, the specific steps of the identification-related data processing are as follows: A1, after performing image data processing on the material identification image data of the preset material to be processed, it is determined whether the monitored identification acquisition quality assessment index is within the preset acquisition quality assessment threshold range. If so, the identification-related data processing is stopped, otherwise A2 is executed, and the image data processing includes image data preprocessing, image data feature extraction and image data enhancement; A2, after performing spectral data processing on the material identification spectral data of the preset material to be processed, it is determined whether the monitored identification acquisition quality assessment index is within the preset acquisition quality assessment threshold range. If so, the identification-related data processing is stopped, otherwise A3 is executed, and the spectral data processing includes spectral data preprocessing and spectral data feature extraction; A3, performing environmental data processing on the material identification environmental data of the preset material to be processed, and at the same time obtaining the identification processing compliance assessment index after performing the identification-related data processing, and the environmental data processing includes environmental data preprocessing and environmental data feature extraction.

[0020] Furthermore, the specific steps of obtaining the identification processing compliance evaluation index after executing the identification related data processing are as follows: obtaining the processing related evaluation standard deviation of the preset material to be processed after executing the identification related data processing, the processing related evaluation standard deviation includes the processing pixel intensity standard deviation, the processing spectral intensity standard deviation, the processing environment temperature standard deviation and the processing environment humidity standard deviation; performing calculations on the processing environment temperature standard deviation and the processing environment humidity standard deviation with the corresponding environment temperature standard deviation reference value and environment humidity standard deviation reference value obtained from the preset database, and then combining them with the environment processing compliance weight to obtain the environment processing coefficient; performing calculations on the processing pixel intensity standard deviation and the processing spectral intensity standard deviation with the corresponding pixel intensity standard deviation reference value and spectral intensity standard deviation reference value obtained from the preset database, and then combining them with the image processing compliance weight and the spectral processing compliance weight to obtain the image processing coefficient and the spectral processing coefficient; performing addition operation on the environment processing coefficient, the image processing coefficient and the spectral processing coefficient to obtain the identification processing compliance evaluation index.

[0021] Furthermore, the specific steps of judging whether to execute material identification-related environmental optimization based on the identification processing compliance evaluation index and the preset processing compliance threshold range are as follows: compare the identification processing compliance evaluation index with the preset processing compliance threshold range: if the identification processing compliance evaluation index is within the preset processing compliance threshold range, material identification-related environmental optimization is not executed, and at the same time, the identification-related data, the identification acquisition quality evaluation index and the identification processing compliance evaluation index are backed up in real time and transmitted to the preset material identification database; if the identification processing compliance evaluation index exceeds the preset processing compliance threshold range, material identification-related environmental optimization is executed, and at the same time, the identification environment optimization evaluation index after executing the material identification-related environmental optimization is obtained.

[0022] Furthermore, the specific steps of obtaining the identification environment optimization evaluation index after executing the material identification related environment optimization are as follows: obtaining the initial environment evaluation related data of the preset material to be optimized within the initial preset time interval before executing the material identification related environment optimization, the initial environment evaluation related data including the initial identification processing compliance evaluation index, the initial environment light intensity, the initial environment temperature and the initial environment humidity; obtaining the optimized environment evaluation related data of the preset material to be optimized within the optimized preset time interval after executing the material identification related environment optimization, the optimized environment evaluation related data including the optimized identification processing compliance evaluation index, the optimized environment light intensity, the optimized environment temperature and the optimized environment humidity; processing the initial environment evaluation related data and the optimized environment evaluation related data to obtain the optimized environment related change amount, the optimized environment related change amount including the optimized light intensity change amount, the optimized temperature change amount and the optimized humidity change amount; performing operation processing on the difference between the optimized identification processing compliance evaluation index and the initial identification processing compliance evaluation index and the complement of the initial identification processing compliance evaluation index, and combining the optimized environment related change amount and the optimized evaluation weight to obtain the identification environment optimization evaluation index, the optimized evaluation weight including the ambient light intensity optimization weight, the ambient temperature optimization weight and the ambient humidity optimization weight.

[0023] The embodiment of the present application provides an intelligent material identification method based on multi-source heterogeneous data fusion, comprising the following steps: S1, collecting identification-related data of a preset material within a preset time interval, wherein the identification-related data includes material identification image data, material identification spectral data, and material identification environment data; S2, evaluating the collection quality of the identification-related data to obtain an identification collection quality evaluation index, and judging whether to perform identification-related data processing based on the identification collection quality evaluation index and a preset collection quality evaluation threshold range, wherein the identification collection quality evaluation index is used to quantify the collection quality of the identification-related data; S3, obtaining an identification processing compliance evaluation index after performing identification-related data processing, and judging whether to perform material identification-related environment optimization based on the identification processing compliance evaluation index and a preset processing compliance threshold range, wherein the identification processing compliance evaluation index is used to comprehensively quantify the degree of compliance of identification-related data processing; S4, obtaining an identification environment optimization evaluation index after performing material identification-related environment optimization, and comprehensively judging whether to perform material identification based on the identification collection quality evaluation index and the identification processing compliance evaluation index, wherein the identification environment optimization evaluation index is used to quantitatively evaluate the effect of material identification-related environment optimization.

[0024] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0025] 1. By evaluating the collection quality of identification-related data to determine whether to execute identification-related data processing, then obtaining the identification processing compliance evaluation index after executing the identification-related data processing and determining whether to execute material identification-related environment optimization, and finally making a comprehensive judgment on whether to execute material identification, the evaluation and processing of the collection quality of identification-related data are realized, thereby achieving the improvement of data quality in the intelligent material identification process of multi-source heterogeneous data fusion, and effectively solving the problem of low data quality in the intelligent material identification process of multi-source heterogeneous data fusion in the prior art.

[0026] 2. By acquiring the preset material image quality assessment related data, spectral quality assessment related data and environmental quality assessment related data within a preset time interval and combining them with the quality assessment reference data obtained from the preset database, a quality-related assessment value is obtained, and the quality-related assessment value and the acquisition quality assessment weight obtained from the preset database are combined to obtain the recognition acquisition quality assessment index, thereby realizing the numerical evaluation of the recognition-related data acquisition quality, and further realizing a more accurate evaluation of the recognition-related data acquisition quality.

[0027] 3. By processing the initial environmental assessment related data and the optimized environmental assessment related data, the optimized environment related changes are obtained. At the same time, the difference between the optimized identification processing compliance evaluation index and the initial identification processing compliance evaluation index and the complement of the initial identification processing compliance evaluation index are processed and combined with the optimized environment related changes and the optimized evaluation weight to obtain the identification environment optimization evaluation index, thereby realizing the numerical evaluation of the material identification related environmental optimization effect, and then realizing a more accurate evaluation of the material identification related environmental optimization effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A schematic diagram of the structure of an intelligent material identification system based on multi-source heterogeneous data fusion provided in an embodiment of the present application;

[0029] Figure 2 A schematic diagram of changes in the identification processing compliance evaluation index provided in an embodiment of the present application, wherein (a) is a schematic diagram of changes in the identification processing compliance evaluation index with the standard deviation of the processed pixel intensity, and (b) is a schematic diagram of changes in the identification processing compliance evaluation index with the standard deviation of the processed spectral intensity;

[0030] Figure 3 A flowchart of an intelligent material identification method based on multi-source heterogeneous data fusion provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] The embodiment of the present application solves the problem of low data quality in the intelligent material identification process of multi-source heterogeneous data fusion in the prior art by providing an intelligent material identification system and method based on multi-source heterogeneous data fusion. The system and method collect identification-related data of a preset material within a preset time interval, and then evaluate the collection quality of the identification-related data to obtain an identification collection quality evaluation index and determine whether to perform identification-related data processing. Then, the identification processing compliance evaluation index after performing identification-related data processing is obtained. Based on the identification processing compliance evaluation index and a preset processing compliance threshold range, it is determined whether to perform material identification-related environmental optimization. Finally, the identification environment optimization evaluation index after performing material identification-related environmental optimization is obtained. A comprehensive judgment is made on whether to perform material identification based on the identification collection quality evaluation index and the identification processing compliance evaluation index, thereby achieving improved data quality in the intelligent material identification process of multi-source heterogeneous data fusion.

[0032] The technical solution in the embodiment of the present application is to solve the problem of low data quality in the intelligent material recognition process of multi-source heterogeneous data fusion. The overall idea is as follows:

[0033] By evaluating the collection quality of identification-related data to determine whether to execute identification-related data processing, then obtaining the identification processing compliance evaluation index after executing identification-related data processing and determining whether to execute material identification-related environmental optimization, and finally making a comprehensive judgment on whether to execute material identification, the effect of improving data quality in the intelligent material identification process of multi-source heterogeneous data fusion is achieved.

[0034] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0035] like Figure 1As shown, it is a structural schematic diagram of an intelligent material identification system based on multi-source heterogeneous data fusion provided by an embodiment of the present application. An intelligent material identification system based on multi-source heterogeneous data fusion provided by an embodiment of the present application includes: an identification-related data acquisition module, an acquisition quality assessment and judgment module, a data processing assessment and judgment module, and an environment optimization comprehensive judgment module; wherein the identification-related data acquisition module is used to collect identification-related data of preset materials within a preset time interval, and the identification-related data includes material identification image data, material identification spectral data, and material identification environmental data; the acquisition quality assessment and judgment module is used to evaluate the acquisition quality of the identification-related data to obtain an identification acquisition quality assessment index, and based on the identification acquisition quality assessment index and a preset acquisition quality assessment threshold range, ... acquisition module is used to collect identification-related data of preset materials within a preset time interval, and the identification-related data acquisition module is used to collect identification-related data of preset materials within a preset time interval, and the identification-related The identification collection quality evaluation index is used to quantify the collection quality of identification-related data; the data processing evaluation judgment module is used to obtain the identification processing compliance evaluation index after the identification-related data processing is executed, and based on the identification processing compliance evaluation index and the preset processing compliance threshold range, it is judged whether to execute the material identification-related environmental optimization, and the identification processing compliance evaluation index is used to comprehensively quantify the compliance degree of the identification-related data processing; the environmental optimization comprehensive judgment module is used to obtain the identification environment optimization evaluation index after the material identification-related environmental optimization is executed, and comprehensively judge whether to execute material identification based on the identification collection quality evaluation index and the identification processing compliance evaluation index, and the identification environment optimization evaluation index is used to quantitatively evaluate the effect of the material identification-related environmental optimization.

[0036] In this embodiment, the identification-related data includes material identification image data, material identification spectral data and material identification environmental data; wherein, the material identification image data is used to monitor the appearance characteristics of the preset material, specifically including the material image, the material image pixel brightness value and the material color saturation; the material identification spectral data is used to monitor the chemical composition (moisture) content of the preset material, specifically referring to the material spectral intensity; the material identification environmental data is used to monitor the external environmental state of the preset material, specifically including identifying the ambient temperature and identifying the ambient humidity.

[0037] Specifically, the material image is obtained through a high-definition industrial camera, the pixel brightness value of the material image is obtained through image processing software or programming library (such as OpenCV, MATLAB), and the color saturation of the material is obtained through color space conversion.

[0038] The spectral intensity of the material is obtained by performing spectral scanning on the preset material (the preset materials involved in this example refer to tobacco leaves) through a near-infrared spectrometer.

[0039] The ambient temperature and humidity are obtained by deploying temperature sensors and humidity sensors in the preset material storage area.

[0040] By acquiring identification-related data to conduct comprehensive monitoring of preset materials, the reliability of preset material identification-related data is improved, thereby achieving an improvement in the accuracy of preset material identification.

[0041] Furthermore, the specific steps of evaluating the collection quality of the recognition-related data to obtain the recognition collection quality evaluation index are as follows: obtaining image quality evaluation-related data, spectral quality evaluation-related data and environmental quality evaluation-related data of a preset material within a preset time interval, the image quality evaluation-related data including edge gradient value, material image mean value, material image standard deviation and color saturation mean value, the spectral quality evaluation-related data including spectral intensity maximum value and spectral intensity minimum value, and the environmental quality evaluation-related data including ambient temperature measurement value and ambient humidity measurement value; combining the image quality evaluation-related data, spectral quality evaluation-related data and environmental quality evaluation-related data with the quality evaluation reference data obtained from a preset database to obtain quality-related evaluation values, the quality-related evaluation values ​​including the image quality evaluation value (i.e., the value in the recognition collection quality evaluation index) ), spectral quality assessment value (i.e., identifying the acquisition quality assessment index ) and environmental quality assessment values ​​(i.e., identifying the ), the quality assessment reference data includes quality assessment reference values ​​and quality assessment influencing factors, the quality assessment reference values ​​include image clarity reference values, color saturation maximum reference values, spectral intensity maximum reference deviations, ambient temperature reference values, and ambient humidity reference values, and the quality assessment influencing factors include image clarity influencing factors, image noise influencing factors, image color influencing factors, spectral intensity influencing factors, ambient temperature influencing factors, and ambient humidity influencing factors; the quality-related assessment values ​​and the acquisition quality assessment weights obtained from the preset database are combined to obtain the recognition acquisition quality assessment index (i.e. ), the acquisition quality assessment weight includes image quality assessment weight, spectral quality assessment weight and environmental quality assessment weight.

[0042] The method for obtaining the recognition acquisition quality assessment index is as follows:

[0043] ;

[0044] ;

[0045] ;

[0046] ;

[0047] In the formula, represents the recognition and collection quality evaluation index of the jth preset material, Indicates the number of the preset material. , Indicates the total quantity of the preset material. represents the image quality assessment weight, represents the spectral quality assessment weight, represents the environmental quality assessment weight, represents the image quality evaluation value of the j-th preset material, represents the spectral quality evaluation value of the j-th preset material, represents the environmental quality assessment value of the jth preset material, represents the image clarity impact factor, represents the image noise impact factor, represents the image color impact factor, Represents the number of horizontal columns of image pixels, , Represents the total number of horizontal columns of image pixels, Represents the number of horizontal rows of image pixels, , Indicates the total number of horizontal rows of image pixels, Indicates the edge gradient value of the image pixel of the j-th preset material located at the y-th row and x-th column, Indicates the image clarity reference value, represents the mean value of the material image of the j-th preset material, represents the material image standard deviation of the j-th preset material, represents the mean color saturation of the jth preset material, Indicates the maximum reference value of color saturation. represents the spectral intensity influencing factor, Indicates the maximum spectral intensity of the jth preset material, Indicates the minimum spectral intensity of the j-th preset material, Indicates the maximum reference deviation of the spectral intensity, represents the ambient temperature measurement value of the jth preset material, Indicates the ambient temperature reference value, Represents the measured value of the ambient humidity of the jth preset material, Indicates the reference value of ambient humidity. represents the environmental temperature influence factor, Represents the environmental humidity impact factor.

[0048] In this embodiment, the edge gradient value is used to describe the clarity of a preset material image, and the edge gradient value is obtained through an image processing algorithm (such as Sobel and Canny edge detection).

[0049] The material image mean is used to describe the brightness of the preset material image; the material image standard deviation is used to describe the contrast of the preset material image; the color saturation mean is used to describe the color of the preset material image; the material image mean, material image standard deviation and color saturation mean are obtained through image processing software (such as OpenCV).

[0050] The maximum spectral intensity is the maximum value of the reflected light intensity in the preset material (tobacco leaf) spectrum, and the minimum spectral intensity is the minimum value of the reflected light intensity in the preset material (tobacco leaf) spectrum; the maximum spectral intensity and the minimum spectral intensity are obtained by a near-infrared spectrometer.

[0051] The ambient temperature measurement value and the ambient humidity measurement value are the ambient temperature value and the ambient humidity value at any time within a preset time interval. The ambient temperature measurement value and the ambient humidity measurement value are obtained by deploying temperature sensors and humidity sensors in a preset material storage area.

[0052] It should be understood that the image clarity impact factor is used to describe the degree of influence of image clarity on the image quality evaluation value. Specifically, the image clarity impact factor is an impact factor corresponding to the image clarity preset in a preset database, and represents a numerical value of the degree of influence of the image clarity on the image quality evaluation value. When used, the preset image clarity impact factor can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship. For example, the image clarity and the impact factor corresponding to the image clarity preset in the preset database form a mapping set, and the real-time image clarity is input into the mapping set to obtain the corresponding impact factor. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0053] The image noise impact factor is used to describe the degree of influence of image noise on the image quality assessment value. Specifically, the image noise impact factor is the impact factor corresponding to the image noise preset in the preset database, and represents the numerical value of the influence of the image noise on the image quality assessment value. When used, the preset image noise impact factor can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship. For example, the image noise and the impact factor corresponding to the image noise preset in the preset database form a mapping set, and the real-time image noise is input into the mapping set to obtain the corresponding impact factor. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0054] In this example, the sum of the image clarity impact factor, the image noise impact factor, and the image color impact factor is 1. The image color impact factor is used to describe the degree of influence of image color on the image quality assessment value.

[0055] The spectral intensity influence factor is used to describe the influence of the spectral intensity on the spectral quality assessment value. Specifically, the spectral intensity influence factor is the influence factor corresponding to the spectral intensity preset in the preset database, which represents the numerical value of the influence of the spectral intensity on the spectral quality assessment value. When used, the preset spectral intensity influence factor can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship. For example, the spectral intensity and the influence factor corresponding to the spectral intensity preset in the preset database form a mapping set, and the real-time spectral intensity is input into the mapping set to obtain the corresponding influence factor. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0056] The ambient temperature impact factor is used to describe the degree of influence of the ambient temperature on the environmental quality assessment value. Specifically, the ambient temperature impact factor is the impact factor corresponding to the ambient temperature preset in the preset database, which represents the numerical value of the influence of the ambient temperature on the environmental quality assessment value. When used, the preset ambient temperature impact factor can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship. For example, the ambient temperature and the impact factor corresponding to the ambient temperature preset in the preset database form a mapping set, and the real-time ambient temperature is input into the mapping set to obtain the corresponding impact factor. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0057] In this example, the sum of the ambient temperature impact factor and the ambient humidity impact factor is 1. The ambient humidity impact factor is used to describe the degree of influence of ambient humidity on the environmental quality assessment value.

[0058] Among them, the image quality assessment weight is used to describe the degree of influence of the compliance of the material identification image data collection of the preset material on the identification and collection quality assessment index. Specifically, the image quality assessment weight is the weight corresponding to the image quality assessment preset in the preset database, which represents the numerical value of the degree of influence of the compliance of the material identification image data collection of the preset material on the identification and collection quality assessment index. When used, the weight corresponding to the preset image quality assessment can be directly obtained from the preset database, and the corresponding relationship can be a pre-set mapping relationship. For example, the compliance of the material identification image data collection of the preset material and the weight corresponding to the preset image quality assessment in the preset database form a mapping set, and the compliance of the material identification image data collection of the real-time preset material is input into the mapping set to obtain the corresponding weight. The mapping relationship can be a one-to-one correspondence or a many-to-one relationship. In this example, the value range is [0, 1].

[0059] The spectral quality assessment weight is used to describe the degree of influence of the compliance of the material identification spectral data collection of the preset material on the identification and collection quality assessment index. Specifically, the spectral quality assessment weight is the weight corresponding to the spectral quality assessment preset in the preset database, which represents the numerical value of the influence of the compliance of the material identification spectral data collection of the preset material on the identification and collection quality assessment index. When used, the weight corresponding to the preset spectral quality assessment can be directly obtained from the preset database, and the corresponding relationship can be a pre-set mapping relationship. For example, the compliance of the material identification spectral data collection of the preset material and the weight corresponding to the spectral quality assessment preset in the preset database form a mapping set, and the compliance of the material identification spectral data collection of the real-time preset material is input into the mapping set to obtain the corresponding weight. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0060] In this example, the sum of the image quality assessment weight, the spectral quality assessment weight, and the environmental quality assessment weight is 1. The environmental quality assessment weight is used to describe the degree of influence of the compliance of the material recognition environment data collection of the preset material on the recognition collection quality assessment index.

[0061] The recognition acquisition quality assessment index is used to quantify the acquisition quality of recognition-related data; among them, the recognition acquisition quality assessment index is affected by the calculation of the image quality assessment value, the spectral quality assessment value and the environmental quality assessment value; specifically, the recognition acquisition quality assessment index is obtained by comprehensive analysis of the correlation and mutual influence between multiple parameters. For example, the edge gradient value is used to measure the clarity of the material image, while the material image mean and the material image standard deviation are used to reflect the brightness and contrast of the image. For preset materials such as tobacco leaves, an increase in image contrast usually enhances the edge recognition ability, that is, the edge gradient value increases; similarly, as the mean value of the material image decreases, the image may become too dark and the edge blurring may increase, that is, the edge gradient value decreases; in addition, changes in the mean value of the material image often lead to changes in the material The distortion of image color affects the color saturation mean. It also needs to be considered that the difference between the maximum and minimum spectral intensity also affects the color saturation mean. As the dryness of the tobacco leaves increases, the difference in spectral intensity increases, and the color saturation mean will also change. At the same time, the ambient temperature measurement and the ambient humidity measurement also affect the imaging of tobacco leaves. For example, an increase in the ambient humidity measurement will lead to an increase in light absorption on the tobacco leaf surface, thereby affecting the brightness of the image. Therefore, a numerical evaluation of the quality of identification-related data acquisition is achieved in a quantitative manner. The quality of identification-related data acquisition is judged through numerical evaluation, thereby improving the reliability of the judgment of the quality of identification-related data acquisition, and thereby improving the accuracy of the judgment of abnormal identification-related data acquisition quality.

[0062] Furthermore, the specific steps for determining whether to execute identification-related data processing based on the identification and collection quality assessment index and the preset collection quality assessment threshold range are as follows: Compare the identification and collection quality assessment index with the preset collection quality assessment threshold range: If the identification and collection quality assessment index is within the preset collection quality assessment threshold range, identification-related data processing is not executed, and at the same time, the identification-related data and the identification and collection quality assessment index are backed up in real time and uploaded to cloud storage; If the identification and collection quality assessment index exceeds the preset collection quality assessment threshold range, identification-related data processing is executed.

[0063] In this embodiment, the preset acquisition quality assessment threshold range is obtained from a database. Specifically, the preset acquisition quality assessment threshold range is set by professionals according to standards in the field. For example, the image quality assessment value range (generally 0 to 1), the spectral quality assessment value range (generally 0 to 5) and the environmental quality assessment value range (generally 0 to 2) are generally set to determine the preset acquisition quality assessment threshold range; by making judgments in combination with the preset acquisition quality assessment threshold range, a more accurate assessment of the acquisition quality of the identified relevant data is achieved.

[0064] Furthermore, the specific steps of identification-related data processing are as follows: A1, after performing image data processing on the material identification image data of the preset material to be processed, determine whether the monitored identification acquisition quality assessment index is within the preset acquisition quality assessment threshold range. If so, stop the identification-related data processing, otherwise execute A2, the image data processing includes image data preprocessing, image data feature extraction and image data enhancement; A2, after performing spectral data processing on the material identification spectral data of the preset material to be processed, determine whether the monitored identification acquisition quality assessment index is within the preset acquisition quality assessment threshold range. If so, stop the identification-related data processing, otherwise execute A3, the spectral data processing includes spectral data preprocessing and spectral data feature extraction; A3, perform environmental data processing on the material identification environmental data of the preset material to be processed, and at the same time obtain the identification processing compliance assessment index after performing the identification-related data processing, and the environmental data processing includes environmental data preprocessing and environmental data feature extraction.

[0065] In this embodiment, image data preprocessing refers to normalization processing after removing noise through a filter (such as Gaussian filtering); image data feature extraction refers to: extracting edge features of image data by using the Canny algorithm or the Sobel operator to obtain edge strength and edge direction, and extracting image data area features by using the directional gradient histogram to obtain local gradient direction distribution; image data enhancement refers to image data enhancement by histogram equalization and contrast stretching.

[0066] Spectral data preprocessing refers to reducing noise using moving average method or Savitzky-Golay filter; spectral data feature extraction refers to data dimension reduction through principal component analysis (PCA).

[0067] Environmental data preprocessing includes data cleaning and data normalization. Specifically, data cleaning refers to identifying missing values ​​through visualization tools (such as heat maps) and filling missing values ​​through interpolation methods, such as using the mean interpolation method to calculate the mean of the column containing missing values, replacing the missing values ​​in the column with the mean, and performing data normalization through Z-Score standardization; Environmental data feature extraction refers to: extracting environmental data features through statistical analysis methods (such as calculating specific quantiles through Python's pandas library, obtaining the correlation matrix through the corr() function in pandas, and visualizing it as a heat map through the seaborn library) to obtain specific quantiles and correlation matrices; By identifying related data processing, the reliability of identifying related data processing is improved in the case of abnormal quality of related data collection.

[0068] Furthermore, the specific steps for obtaining the identification processing compliance evaluation index after executing the identification related data processing are as follows: obtain the processing related evaluation standard deviation of the preset material to be processed after executing the identification related data processing, the processing related evaluation standard deviation includes the processing pixel intensity standard deviation, the processing spectral intensity standard deviation, the processing ambient temperature standard deviation and the processing ambient humidity standard deviation; perform calculations on the processing ambient temperature standard deviation and the processing ambient humidity standard deviation respectively with the corresponding ambient temperature standard deviation reference value and ambient humidity standard deviation reference value obtained from the preset database, and then combine them with the environmental processing compliance weight to obtain the environmental processing coefficient; perform calculations on the processing pixel intensity standard deviation and the processing spectral intensity standard deviation respectively with the corresponding pixel intensity standard deviation reference value and spectral intensity standard deviation reference value obtained from the preset database, and then combine them with the image processing compliance weight and the spectral processing compliance weight to obtain the image processing coefficient and the spectral processing coefficient; perform addition operation on the environmental processing coefficient, the image processing coefficient and the spectral processing coefficient to obtain the identification processing compliance evaluation index.

[0069] The method to obtain the identification and treatment compliance assessment index is as follows:

[0070] ;

[0071] In the formula, Indicates that the identification and processing of the dth preset material to be processed meets the evaluation index, Indicates the number of the preset material to be processed, , Indicates the total quantity of preset materials to be processed, where: , Indicates that image processing complies with weights, Indicates that the spectral processing complies with the weights, Indicates that the environmental treatment meets the weight, represents the standard deviation of the pixel intensity of the dth preset material to be processed, Indicates the standard deviation reference value of the pixel intensity of the dth preset material to be processed, It represents the standard deviation of the intensity of the processed spectrum of the dth preset material to be processed. It represents the reference value of the standard deviation of the spectral intensity of the dth preset material to be processed. It represents the standard deviation of the processing environment temperature of the dth preset material to be processed. It represents the standard deviation of the processing environment humidity of the dth preset material to be processed. It represents the standard deviation reference value of the ambient temperature of the dth preset material to be processed. Indicates the standard deviation reference value of the ambient humidity of the dth preset material to be processed, Indicates the preset collection quality assessment threshold range. represents the recognition and collection quality evaluation index of the jth preset material, Indicates the number of the preset material. , Indicates the total quantity of the preset material.

[0072] In this embodiment, the pixel intensity standard deviation, the spectral intensity standard deviation, the ambient temperature standard deviation and the ambient humidity standard deviation are obtained through Python's NumPy library or Pandas library; the ambient temperature standard deviation reference value is represented by summing and averaging the collected historical ambient temperature standard deviations; the ambient humidity standard deviation reference value is represented by summing and averaging the collected historical ambient humidity standard deviations; the pixel intensity standard deviation reference value is represented by summing and averaging the collected historical pixel intensity standard deviations; the spectral intensity standard deviation reference value is represented by summing and averaging the collected historical spectral intensity standard deviations.

[0073] It should be understood that the image processing compliance weight is used to describe the degree of influence of the compliance of the material identification image data processing of the preset material on the identification processing compliance evaluation index. Specifically, the image processing compliance weight is the corresponding weight of the image processing compliance preset in the preset database, which represents the numerical value of the influence of the compliance of the material identification image data processing of the preset material on the identification processing compliance evaluation index. When used, the corresponding weight of the preset image processing compliance can be directly obtained from the preset database, and the corresponding relationship can be a pre-set mapping relationship. For example, the compliance of the material identification image data processing of the preset material and the corresponding weight of the image processing compliance preset in the preset database form a mapping set, and the compliance of the material identification image data processing of the real-time preset material is input into the mapping set to obtain the corresponding weight. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0074] The spectral processing compliance weight is used to describe the degree of influence of the compliance of the material identification spectral data processing of the preset material on the identification processing compliance evaluation index. Specifically, the spectral processing compliance weight is the corresponding weight of the spectral processing compliance preset in the preset database, which represents the numerical value of the influence of the compliance of the material identification spectral data processing of the preset material on the identification processing compliance evaluation index. When used, the corresponding weight of the preset spectral processing compliance can be directly obtained from the preset database, and the corresponding relationship can be a pre-set mapping relationship. For example, the compliance of the material identification spectral data processing of the preset material and the corresponding weight of the spectral processing compliance preset in the preset database form a mapping set, and the compliance of the material identification spectral data processing of the real-time preset material is input into the mapping set to obtain the corresponding weight. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0075] In this example, the sum of the image processing compliance weight, the spectral processing compliance weight, and the environmental processing compliance weight is 1. The environmental processing compliance weight is used to describe the degree of influence of the compliance of the material recognition environment data processing of the preset material on the recognition processing compliance evaluation index.

[0076] like Figure 2As shown, it is a change diagram of the return data quality evaluation index provided in an embodiment of the present application, wherein the following settings are made: the image processing compliance weight is 0.4, the spectral processing compliance weight is 0.3, the environmental processing compliance weight is 0.3, the pixel intensity standard deviation reference value is 10 grayscale, the spectral intensity standard deviation reference value is 8, the processing environment temperature standard deviation is 2 degrees Celsius, the processing environment humidity standard deviation is 1.5 relative humidity, the ambient temperature standard deviation reference value is 2 degrees Celsius, and the ambient humidity standard deviation reference value is 1.5 relative humidity; wherein (a) is a schematic diagram of the change of the recognition processing compliance evaluation index with the processing pixel intensity standard deviation. Specifically, the processing spectral intensity standard deviation is set to 8, the pixel intensity standard deviation reference value is 10 grayscale, and when the processing pixel When the intensity standard deviation is equal to the pixel intensity standard deviation reference value, the recognition processing compliance evaluation index reaches the maximum value. As the deviation between the processing pixel intensity standard deviation and the pixel intensity standard deviation reference value gradually increases, the recognition processing compliance evaluation index decreases accordingly. (b) is a schematic diagram of the change of the recognition processing compliance evaluation index with the processing spectral intensity standard deviation. Specifically, the processing pixel intensity standard deviation is set to 10 grayscales and the spectral intensity standard deviation reference value is 8. When the processing spectral intensity standard deviation is equal to the spectral intensity standard deviation reference value, the recognition processing compliance evaluation index reaches the maximum value. As the deviation between the processing spectral intensity standard deviation and the spectral intensity standard deviation reference value gradually increases, the recognition processing compliance evaluation index also decreases accordingly.

[0077] The identification and processing conformity assessment index is used to comprehensively quantify the conformity of the identification-related data processing. Specifically, the identification and processing conformity assessment index includes parameters from multiple aspects, and the various parameters are related to each other and do not exist independently. For example, as the moisture content of a preset material such as tobacco leaves decreases, the spectral reflectance of the moisture may change, thereby causing the standard deviation of the processed pixel intensity and the standard deviation of the processed spectral intensity to change accordingly; in addition, changes in temperature and humidity also affect the moisture content of the tobacco leaves, thereby causing the standard deviation of the processed spectral intensity to change accordingly. For example, as temperature fluctuations increase, tobacco leaves are prone to drying, thereby affecting the color of the tobacco leaves, that is, the standard deviation of the processed spectral intensity changes. Therefore, through a quantitative approach, the correlation and mutual influence between various factors are taken into account, and a comprehensive analysis is performed to obtain the identification and processing conformity assessment index; a numerical evaluation of the conformity of the identification-related data processing is achieved, and the conformity of the identification-related data processing is judged through a numerical evaluation, thereby achieving an improvement in the accuracy of the evaluation of the conformity of the identification-related data processing.

[0078] Furthermore, the specific steps for determining whether to execute material identification-related environmental optimization based on the identification processing compliance evaluation index and the preset processing compliance threshold range are as follows: Compare the identification processing compliance evaluation index with the preset processing compliance threshold range: If the identification processing compliance evaluation index is within the preset processing compliance threshold range, material identification-related environmental optimization will not be executed, and at the same time, the identification-related data, the identification acquisition quality evaluation index and the identification processing compliance evaluation index will be backed up in real time and transmitted to the preset material identification database; If the identification processing compliance evaluation index exceeds the preset processing compliance threshold range, material identification-related environmental optimization will be executed, and at the same time, the identification environment optimization evaluation index after executing the material identification-related environmental optimization will be obtained.

[0079] In this embodiment, the preset processing compliance threshold range is obtained from a database. Specifically, the preset processing compliance threshold range is set by professionals according to standards in the field. For example, the preset processing compliance threshold range is generally set to 0.7 to 1. By combining the preset processing compliance threshold range for judgment, a more accurate judgment of the degree of compliance of related data processing is achieved.

[0080] Furthermore, the specific steps of obtaining the identification environment optimization evaluation index after executing the material identification related environment optimization are as follows: obtaining the initial environment evaluation related data of the preset material to be optimized within the initial preset time interval before executing the material identification related environment optimization, the initial environment evaluation related data including the initial identification processing compliance evaluation index, the initial environment light intensity, the initial environment temperature and the initial environment humidity; obtaining the optimized environment evaluation related data of the preset material to be optimized within the optimized preset time interval after executing the material identification related environment optimization, the optimized environment evaluation related data including the optimized identification processing compliance evaluation index, the optimized environment light intensity, the optimized environment temperature and the optimized environment humidity; processing the initial environment evaluation related data and the optimized environment evaluation related data to obtain the optimized environment related change amount, the optimized environment related change amount including the optimized light intensity change amount (i.e. ), optimize the temperature change (i.e. ) and optimizing the humidity change (i.e. ); the difference between the optimized recognition processing compliance evaluation index and the initial recognition processing compliance evaluation index and the complement of the initial recognition processing compliance evaluation index are processed and then combined with the optimization environment related changes and the optimization evaluation weight to obtain the recognition environment optimization evaluation index.

[0081] The method for obtaining the identification environment optimization evaluation index is as follows:

[0082] ;

[0083] In the formula, It represents the recognition environment optimization evaluation index of the qth preset material to be optimized. Indicates the number of the preset material to be optimized. , Indicates the total quantity of preset materials to be optimized, where: , Indicates that the optimization identification processing of the qth preset material to be optimized meets the evaluation index, Indicates that the initial identification processing of the qth preset material to be optimized meets the evaluation index, represents the ambient light intensity optimization weight, represents the ambient temperature optimization weight, represents the environmental humidity optimization weight, Indicates the optimized light intensity change of the qth preset material to be optimized, Indicates the optimized temperature change of the qth preset material to be optimized, Indicates the optimized humidity change of the qth preset material to be optimized, Indicates that the preset processing meets the threshold range, Indicates that the identification and processing of the dth preset material to be processed meets the evaluation index, Indicates the number of the preset material to be processed, , Indicates the total quantity of preset materials to be processed.

[0084] It should be understood that the optimized identification processing compliance evaluation index is used to describe the identification processing compliance evaluation index after the preset material to be optimized performs the material identification related environment optimization; the initial identification processing compliance evaluation index is used to describe the identification processing compliance evaluation index before the preset material to be optimized performs the material identification related environment optimization. Specifically, the material identification related environment optimization can be performed by adjusting the environmental influencing factors, for example, based on the initial ambient temperature being reduced by 2 degrees Celsius in sequence, and the adjustment is achieved by setting the preset material constant temperature box temperature. It should be understood that as the ambient temperature decreases, it helps to maintain the chemical composition of the preset material (tobacco leaves) stable and reduce the loss of volatile substances, thereby improving the accuracy of material identification; wherein, the initial ambient light intensity, initial ambient temperature, initial ambient humidity, optimized ambient light intensity, optimized ambient temperature and optimized ambient humidity are obtained by deploying a light intensity sensor, a temperature sensor and a humidity sensor in the preset material storage area; the optimized environment related change amount is obtained by subtracting the initial ambient light intensity, the initial ambient temperature and the initial ambient humidity from the corresponding optimized ambient light intensity, the optimized ambient temperature and the optimized ambient humidity and then performing normalization.

[0085] Among them, the optimization evaluation weights include the ambient light intensity optimization weight, the ambient temperature optimization weight and the ambient humidity optimization weight. The ambient light intensity optimization weight is used to describe the influence of the ambient light intensity optimization on the identification environment optimization evaluation index. Specifically, the ambient light intensity optimization weight is the weight corresponding to the ambient light intensity optimization preset in the preset database, which represents the numerical value of the influence of the ambient light intensity optimization on the identification environment optimization evaluation index. When used, the weight corresponding to the preset ambient light intensity optimization can be directly obtained from the preset database, and the corresponding relationship can be a pre-set mapping relationship. For example, the ambient light intensity optimization and the weight corresponding to the ambient light intensity optimization preset in the preset database form a mapping set, and the real-time ambient light intensity optimization is input into the mapping set to obtain the corresponding weight. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0086] The ambient temperature optimization weight is used to describe the influence of ambient temperature optimization on the identification environment optimization evaluation index. Specifically, the ambient temperature optimization weight is the weight corresponding to the ambient temperature optimization preset in the preset database, which represents the numerical value of the influence of ambient temperature optimization on the identification environment optimization evaluation index. When used, the weight corresponding to the preset ambient temperature optimization can be directly obtained from the preset database, and the corresponding relationship can be a pre-set mapping relationship. For example, the ambient temperature optimization and the weight corresponding to the ambient temperature optimization preset in the preset database form a mapping set, and the real-time ambient temperature optimization is input into the mapping set to obtain the corresponding weight. The mapping relationship can be one-to-one or many-to-one. In this example, the value range is [0, 1].

[0087] In this example, the sum of the ambient light intensity optimization weight, the ambient temperature optimization weight, and the ambient humidity optimization weight is 1. The ambient humidity optimization weight is used to describe the influence of ambient humidity optimization on the identification environment optimization evaluation index.

[0088] Specific assumptions: the ambient light intensity optimization weight is 0.3, the ambient temperature optimization weight is 0.3, and the ambient humidity optimization weight is 0.4; the above method can be used to calculate the recognition environment optimization evaluation index, and the change statistics of the recognition environment optimization evaluation index are shown in Table 1:

[0089] Table 1 Statistics of changes in identification environment optimization evaluation index

[0090]

[0091] It can be seen from the first and second groups of data in Table 1 that when the initial recognition processing compliance evaluation index is a fixed value (i.e., 0.20), as the optimized light intensity change increases, the optimized recognition processing compliance evaluation index increases accordingly. As the optimized recognition processing compliance evaluation index increases, it shows that environmental optimization increases the degree of compliance with the recognition-related data processing, thereby increasing the recognition environment optimization evaluation index; changes in the optimized light intensity change, the optimized humidity change, and the optimized humidity change directly lead to changes in the optimized recognition processing compliance evaluation index, thereby affecting the recognition environment optimization evaluation index.

[0092] In addition, the identification environment optimization evaluation index includes parameters from multiple aspects, and each parameter is related to each other and does not exist independently. For example, as the temperature of the optimization environment changes, the optimization identification processing compliance evaluation index changes accordingly. It is also necessary to consider that as the temperature of the optimization environment changes, it will also affect the humidity of the optimization environment. Therefore, through quantitative methods, the correlation and mutual influence between various factors are considered, and the identification environment optimization evaluation index is obtained through comprehensive analysis, which realizes the numerical evaluation of the environmental optimization effect related to material identification. The environmental optimization effect related to material identification is judged through numerical evaluation, thereby achieving a more accurate evaluation of the environmental optimization effect related to material identification.

[0093] Furthermore, the specific steps for comprehensively judging whether to execute material identification based on the identification and collection quality evaluation index and the identification and processing compliance evaluation index are as follows: judge whether the obtained identification and collection quality evaluation index is within the preset collection quality evaluation threshold range; if the identification and collection quality evaluation index is within the preset collection quality evaluation threshold range, execute material identification, otherwise, judge whether the obtained identification and processing compliance evaluation index is within the preset processing compliance threshold range; if the identification and processing compliance evaluation index is within the preset processing compliance threshold range, execute material identification, otherwise, judge whether the obtained identification environment optimization evaluation index is within the preset environment optimization effect threshold range; if the identification environment optimization evaluation index is within the preset environment optimization effect threshold range, execute material identification, otherwise, do not execute material identification, and remind the preset personnel to perform abnormal identification detection measures, and the abnormal identification detection measures include equipment abnormal identification detection and re-acquisition of identification-related data.

[0094] In this embodiment, the preset environmental optimization effect threshold range is obtained from a database. Specifically, the preset environmental optimization effect threshold range is set by professionals according to standards in the field. For example, the optimized light intensity change range (generally 0.5 to 0.7), the optimized temperature change range (generally 0.3 to 0.6), the optimized humidity change range (generally 0.5 to 0.8), the initial recognition processing compliance evaluation index range (generally 0.1 to 0.7) and the optimized recognition processing compliance evaluation index range (generally 0.7 to 1) are set to determine the preset environmental optimization effect threshold range.

[0095] It should be understood that material identification is achieved through machine learning algorithms, such as the K-nearest neighbor algorithm and support vector machine; among them, equipment abnormality identification detection refers to the operation status detection of relevant monitoring equipment during the material identification process, such as sensor aging detection, camera performance detection, etc.; the environmental optimization effect related to material identification is judged by pre-setting the environmental optimization effect threshold range, thereby achieving a more accurate judgment of the environmental optimization effect related to material identification.

[0096] like Figure 3 As shown, it is a flow chart of an intelligent material identification method based on multi-source heterogeneous data fusion provided by an embodiment of the present application, and the method includes the following steps: S1, collecting identification-related data of a preset material within a preset time interval, and the identification-related data includes material identification image data, material identification spectral data and material identification environment data; S2, evaluating the collection quality of the identification-related data to obtain an identification collection quality evaluation index, judging whether to perform identification-related data processing based on the identification collection quality evaluation index and a preset collection quality evaluation threshold range, and the identification collection quality evaluation index is used to quantify the collection quality of the identification-related data; S3, obtaining the identification processing compliance evaluation index after performing identification-related data processing, judging whether to perform material identification-related environment optimization based on the identification processing compliance evaluation index and a preset processing compliance threshold range, and the identification processing compliance evaluation index is used to comprehensively quantify the degree of compliance of identification-related data processing; S4, obtaining the identification environment optimization evaluation index after performing material identification-related environment optimization, comprehensively judging whether to perform material identification based on the identification collection quality evaluation index and the identification processing compliance evaluation index, and the identification environment optimization evaluation index is used to quantitatively evaluate the effect of material identification-related environment optimization.

[0097] In this embodiment, by collecting identification-related data during the material identification process and combining multi-source data (material identification image data, material identification spectral data, and material identification environmental data), a comprehensive judgment is made as to whether to perform material identification. This enables the evaluation of multi-source heterogeneous data during the material identification process, as well as the processing and optimization of abnormal collection quality, thereby improving the reliability of data quality during the material identification process.

[0098] To summarize, the embodiment of the present application evaluates the collection quality of identification-related data to determine whether to execute identification-related data processing, then obtains the identification processing compliance evaluation index after executing the identification-related data processing and determines whether to execute material identification-related environment optimization, and finally makes a comprehensive judgment on whether to execute material identification, thereby realizing the evaluation and processing of the collection quality of identification-related data, and further realizing the improvement of data quality in the intelligent material identification process of multi-source heterogeneous data fusion, effectively solving the problem of low data quality in the intelligent material identification process of multi-source heterogeneous data fusion in the prior art.

[0099] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0100] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0101] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0102] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0103] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0104] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent material identification system based on multi-source heterogeneous data fusion, characterized in that: It includes identification related data collection module, collection quality assessment and judgment module, data processing assessment and judgment module and environment optimization comprehensive judgment module; Wherein, the identification-related data acquisition module is used to acquire identification-related data of a preset material within a preset time interval, and the identification-related data includes material identification image data, material identification spectrum data, and material identification environment data; The collection quality evaluation and judgment module is used to evaluate the collection quality of the identification related data to obtain an identification collection quality evaluation index, and judge whether to perform identification related data processing based on the identification collection quality evaluation index and a preset collection quality evaluation threshold range, wherein the identification collection quality evaluation index is used to quantify the collection quality of the identification related data; The data processing evaluation and judgment module is used to obtain the identification processing compliance evaluation index after executing the identification-related data processing, and judge whether to execute the material identification-related environment optimization based on the identification processing compliance evaluation index and the preset processing compliance threshold range. The identification processing compliance evaluation index is used to comprehensively quantify the compliance degree of the identification-related data processing; The environmental optimization comprehensive judgment module is used to obtain the recognition environment optimization evaluation index after executing the material identification related environmental optimization, and comprehensively judge whether to execute material identification according to the recognition collection quality evaluation index and the recognition processing compliance evaluation index. The recognition environment optimization evaluation index is used to quantitatively evaluate the effect of the material identification related environmental optimization; The specific steps of obtaining the recognition processing compliance evaluation index after the recognition-related data processing is performed are as follows: Obtaining a processing-related evaluation standard deviation of a preset material to be processed after performing identification-related data processing, wherein the processing-related evaluation standard deviation includes a processing pixel intensity standard deviation, a processing spectral intensity standard deviation, a processing environment temperature standard deviation, and a processing environment humidity standard deviation; The processing environment temperature standard deviation and the processing environment humidity standard deviation are respectively calculated and processed with the corresponding environment temperature standard deviation reference value and environment humidity standard deviation reference value obtained from the preset database, and then combined with the environment treatment compliance weight to obtain the environment treatment coefficient; The processed pixel intensity standard deviation and the processed spectral intensity standard deviation are respectively processed with the corresponding pixel intensity standard deviation reference value and spectral intensity standard deviation reference value obtained from the preset database, and then combined with the image processing compliance weight and the spectral processing compliance weight to obtain the image processing coefficient and the spectral processing coefficient; Adding the environmental processing coefficient, the image processing coefficient and the spectrum processing coefficient, the identification processing compliance evaluation index is obtained; The specific steps of obtaining the identification environment optimization evaluation index after the material identification related environment optimization is performed are as follows: Acquire initial environmental assessment related data of a preset material to be optimized within an initial preset time interval before executing material identification related environmental optimization, wherein the initial environmental assessment related data includes an initial identification processing compliance assessment index, initial environmental light intensity, initial environmental temperature, and initial environmental humidity; Obtaining optimization environment assessment related data of a preset material to be optimized within a preset time interval after performing material identification related environment optimization, wherein the optimization environment assessment related data includes an optimization identification processing compliance assessment index, an optimization environment light intensity, an optimization environment temperature, and an optimization environment humidity; Processing the initial environment assessment related data and the optimized environment assessment related data to obtain optimized environment related changes, wherein the optimized environment related changes include optimized light intensity changes, optimized temperature changes, and optimized humidity changes; The difference between the optimized recognition processing compliance evaluation index and the initial recognition processing compliance evaluation index is processed by operation with the complement of the initial recognition processing compliance evaluation index, and then combined with the optimized environment-related changes and the optimized evaluation weights to obtain the recognition environment optimization evaluation index, wherein the optimized evaluation weights include the ambient light intensity optimization weight, the ambient temperature optimization weight, and the ambient humidity optimization weight.

2. The intelligent material identification system based on multi-source heterogeneous data fusion as claimed in claim 1, characterized in that: The specific steps of evaluating the collection quality of the recognition-related data to obtain the recognition collection quality evaluation index are as follows: Acquire image quality assessment related data, spectral quality assessment related data and environmental quality assessment related data of a preset material within a preset time interval, wherein the image quality assessment related data includes edge gradient value, material image mean value, material image standard deviation and color saturation mean value, the spectral quality assessment related data includes spectral intensity maximum value and spectral intensity minimum value, and the environmental quality assessment related data includes ambient temperature measurement value and ambient humidity measurement value; A quality-related evaluation value is obtained by combining the image quality evaluation-related data, the spectral quality evaluation-related data and the environmental quality evaluation-related data with the quality evaluation reference data obtained from a preset database, wherein the quality-related evaluation value includes an image quality evaluation value, a spectral quality evaluation value and an environmental quality evaluation value, the quality evaluation reference data includes a quality evaluation reference value and a quality evaluation influencing factor, the quality evaluation reference value includes an image clarity reference value, a color saturation maximum reference value, a spectral intensity maximum reference deviation, an ambient temperature reference value and an ambient humidity reference value, and the quality evaluation influencing factor includes an image clarity influencing factor, an image noise influencing factor, an image color influencing factor, a spectral intensity influencing factor, an ambient temperature influencing factor and an ambient humidity influencing factor; The quality-related evaluation value and the acquisition quality evaluation weight obtained from a preset database are combined to obtain an identification acquisition quality evaluation index, wherein the acquisition quality evaluation weight includes an image quality evaluation weight, a spectral quality evaluation weight, and an environmental quality evaluation weight.

3. The intelligent material identification system based on multi-source heterogeneous data fusion as claimed in claim 2, characterized in that: The method for obtaining the recognition acquisition quality evaluation index is as follows: ； ； ； ； In the formula, represents the recognition and collection quality evaluation index of the jth preset material, Indicates the number of the preset material. , Indicates the total quantity of the preset material. represents the image quality assessment weight, represents the spectral quality assessment weight, represents the environmental quality assessment weight, represents the image quality evaluation value of the j-th preset material, represents the spectral quality evaluation value of the j-th preset material, represents the environmental quality assessment value of the jth preset material, represents the image clarity impact factor, represents the image noise impact factor, represents the image color impact factor, Represents the number of horizontal columns of image pixels, , Represents the total number of horizontal columns of image pixels, Represents the number of horizontal rows of image pixels, , Indicates the total number of horizontal rows of image pixels, Indicates the edge gradient value of the image pixel of the j-th preset material located at the y-th row and x-th column, Indicates the image clarity reference value, represents the mean value of the material image of the j-th preset material, represents the material image standard deviation of the j-th preset material, represents the mean color saturation of the jth preset material, Indicates the maximum reference value of color saturation. represents the spectral intensity influencing factor, Indicates the maximum spectral intensity of the jth preset material, Indicates the minimum spectral intensity of the j-th preset material, Indicates the maximum reference deviation of the spectral intensity, represents the ambient temperature measurement value of the jth preset material, Indicates the ambient temperature reference value, Represents the measured value of the ambient humidity of the jth preset material, Indicates the reference value of ambient humidity. represents the environmental temperature influence factor, Represents the environmental humidity impact factor.

4. The intelligent material identification system based on multi-source heterogeneous data fusion as claimed in claim 1, characterized in that: The specific steps of determining whether to perform identification-related data processing based on the identification collection quality evaluation index and the preset collection quality evaluation threshold range are as follows: Compare the identification acquisition quality assessment index with the preset acquisition quality assessment threshold range: If the recognition and collection quality assessment index is within the preset collection quality assessment threshold, the recognition-related data processing will not be performed, and the recognition-related data and the recognition and collection quality assessment index will be backed up in real time and uploaded to the cloud storage; If the recognition acquisition quality assessment index exceeds the preset acquisition quality assessment threshold range, recognition-related data processing is performed.

5. The intelligent material identification system based on multi-source heterogeneous data fusion as claimed in claim 4, characterized in that: The specific steps of identifying relevant data processing are as follows: A1, after performing image data processing on the material recognition image data of the preset material to be processed, determine whether the monitored recognition acquisition quality assessment index is within the preset acquisition quality assessment threshold range, if so, stop the recognition related data processing, otherwise execute A2, the image data processing includes image data preprocessing, image data feature extraction and image data enhancement; A2, after performing spectral data processing on the material identification spectral data of the preset material to be processed, determine whether the monitored identification acquisition quality assessment index is within the preset acquisition quality assessment threshold range, if so, stop the identification related data processing, otherwise execute A3, the spectral data processing includes spectral data preprocessing and spectral data feature extraction; A3, performing environmental data processing on the material identification environmental data of the preset material to be processed, and obtaining the identification processing compliance evaluation index after the identification-related data processing is executed, wherein the environmental data processing includes environmental data preprocessing and environmental data feature extraction.

6. The intelligent material identification system based on multi-source heterogeneous data fusion as claimed in claim 1, characterized in that: The specific steps of determining whether to perform material identification-related environmental optimization based on the identification processing compliance evaluation index and the preset processing compliance threshold range are as follows: Compare the identification process compliance assessment index with the preset process compliance threshold range: If the identification processing compliance evaluation index is within the preset processing compliance threshold range, the material identification related environment optimization will not be performed, and the identification related data, the identification collection quality evaluation index and the identification processing compliance evaluation index will be backed up in real time and transmitted to the preset material identification database; If the recognition processing compliance evaluation index exceeds the preset processing compliance threshold range, the material recognition related environment optimization is performed, and the recognition environment optimization evaluation index after performing the material recognition related environment optimization is obtained.

7. The intelligent material identification system based on multi-source heterogeneous data fusion as claimed in claim 1, characterized in that: The specific steps of comprehensively judging whether to perform material identification based on the identification and collection quality evaluation index and the identification and processing compliance evaluation index are as follows: Determine whether the acquired recognition acquisition quality assessment index is within a preset acquisition quality assessment threshold range; If the recognition and collection quality evaluation index is within the preset collection quality evaluation threshold range, material recognition is performed; otherwise, it is determined whether the obtained recognition processing compliance evaluation index is within the preset processing compliance threshold range; If the identification processing compliance evaluation index is within the preset processing compliance threshold range, material identification is performed; otherwise, it is determined whether the obtained identification environment optimization evaluation index is within the preset environment optimization effect threshold range; If the identification environment optimization evaluation index is within the preset environment optimization effect threshold range, material identification is performed; otherwise, material identification is not performed, and the preset personnel are reminded to perform abnormal identification detection measures, which include equipment abnormality identification detection and re-acquisition of identification-related data.

8. A method applied to an intelligent material identification system based on multi-source heterogeneous data fusion as claimed in any one of claims 1 to 7, characterized in that: The following steps are involved: S1, collecting identification-related data of a preset material within a preset time interval, wherein the identification-related data includes material identification image data, material identification spectrum data, and material identification environment data; S2, evaluating the collection quality of the identification related data to obtain an identification collection quality evaluation index, and judging whether to execute the identification related data processing based on the identification collection quality evaluation index and a preset collection quality evaluation threshold range, wherein the identification collection quality evaluation index is used to quantify the collection quality of the identification related data; S3, obtaining an identification processing compliance evaluation index after executing identification-related data processing, and judging whether to execute material identification-related environment optimization based on the identification processing compliance evaluation index and a preset processing compliance threshold range, wherein the identification processing compliance evaluation index is used to comprehensively quantify the compliance degree of identification-related data processing; S4, obtaining the recognition environment optimization evaluation index after executing the material recognition related environment optimization, and making a comprehensive judgment on whether to execute the material recognition according to the recognition acquisition quality evaluation index and the recognition processing compliance evaluation index. The recognition environment optimization evaluation index is used to quantitatively evaluate the effect of the material recognition related environment optimization.

Images