An intelligent identification system for medicinal materials used in pharmaceutical factories
Through the feature order formulation and identification and determination module of the intelligent identification system of medicinal materials, the appearance characteristics of medicinal materials are gradually identified, and the problem of low recognition efficiency of traditional Chinese medicinal materials is solved, and efficient and accurate identification of medicinal materials is achieved.
Patent Information
- Application Number
- CN202411223107.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-03
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-09-03
AI Technical Summary
The existing medicinal material identification system needs to synchronize all appearance characteristics of medicinal materials, resulting in high analysis pressure and low efficiency in identifying medicinal materials with unique characteristics.
An intelligent identification system for medicinal materials was designed, including a medicinal material feature order formulation module, a feature recognition module and an identification and determination module. By formulating the medicinal material feature recognition order, the appearance characteristics of the medicinal material are gradually identified, and the characteristic recognition value and threshold value are used to make judgments to improve identification efficiency. and precision.
It reduces the pressure and difficulty of the system to synchronize all appearance characteristics of medicinal materials, improves the efficiency of identification of medicinal materials, and reasonably controls the number of appearance characteristics analyzed on the basis of ensuring high accuracy.
Smart Images

Figure CN118736334B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of medicinal material identification, and more specifically, to an intelligent medicinal material identification system applied to a pharmaceutical factory. Background Art
[0002] Before producing and processing medicinal materials, pharmaceutical factories need to identify the types of medicinal materials. Most pharmaceutical factories will identify the types of medicinal materials through manual inspection and identification. In order to meet the efficiency of automated production, a small number of pharmaceutical factories will introduce recognition systems to identify medicinal materials. The current recognition system needs to synchronously identify all the appearance features of the medicinal materials in the image based on the image of the medicinal materials. This analysis method will cause greater analysis pressure and difficulty for the system. At the same time, for some medicinal materials with unique characteristics, their types cannot be quickly identified. Summary of the invention
[0003] In view of the deficiencies in the prior art, the object of the present invention is to provide an intelligent medicinal material identification system for use in pharmaceutical factories.
[0004] To achieve the above object, the present invention provides the following technical solutions:
[0005] A medicinal material intelligent identification system applied to pharmaceutical factories, comprising a medicinal material feature order formulation module, a medicinal material feature identification module, and a medicinal material identification and determination module;
[0006] The medicinal material characteristic order formulation module is used to formulate the medicinal material characteristic identification order of the pharmaceutical factory;
[0007] The medicinal material feature recognition module numbers each medicinal material when identifying the medicinal material, takes a photo of the medicinal material, obtains the medicinal material photo, and performs feature recognition on the medicinal material features in the medicinal material photo in the order of medicinal material feature recognition;
[0008] After identifying a medicinal material feature, the medicinal material identification and determination module obtains medicinal material identification values of various medicinal materials, and then obtains identification significance values, and determines whether medicinal material identification is completed based on the comparison result of the identification significance values and their thresholds.
[0009] Furthermore, the formulation of the drug feature recognition order of the pharmaceutical factory is as follows: obtain all feature matching records of the pharmaceutical factory before the current time of the system, obtain the drug features of the feature matching records, obtain the matching stability value of each drug type in the same drug feature and the total number of matching stability of each drug type, sum up the matching stability values of all drug types with the same drug feature and take the average to obtain the matching stability mean Fsg, sum up the total number of matching stability of all drug types with the same drug feature to obtain the total number of matching stability Rhj, and use the formula The identification matching value Dk of the medicinal material feature is obtained, where t1 is the matching stable mean coefficient, and t2 is the matching stable total number coefficient. All medicinal material features are sorted in descending order according to the numerical value of the identification matching value, and then the medicinal material feature recognition order of the pharmaceutical factory is obtained.
[0010] Furthermore, the matching stability value of each medicinal material type in the same medicinal material feature and the total number of matching stability of each medicinal material type are obtained, specifically: the feature matching records of the same medicinal material feature are marked as matching records with the same characteristics, and the medicinal material types of the matching records with the same characteristics are obtained, and the matching records with the same characteristics of the same medicinal material type are marked as matching records with the same characteristics of the same kind, and all matching records with the same characteristics of the same kind are sorted in the order of the time of record generation, and the feature matching values of the two adjacent matching records with the same characteristics of the same kind after sorting are summed up to obtain the matching stability value of the medicinal material type in the same medicinal material feature, and a matching stability threshold is set. When the matching stability value is greater than or equal to the matching stability threshold, the number of matching stability times is increased by one, and when the matching stability value is less than the matching stability threshold, no processing is performed, and the number of matching stability times is summed up to obtain the total number of matching stability times of the medicinal material type in the same medicinal material feature.
[0011] Furthermore, the characteristics of the medicinal materials in the medicinal material photos are identified, specifically:
[0012] A feature data set of the medicinal material characteristic is obtained based on the medicinal material photo, a feature recognition model of the medicinal material characteristic in various medicinal materials is obtained, and the feature recognition models of the medicinal material characteristic in various medicinal materials are marked as independent feature models. Feature recognition values of the medicinal material characteristic in various medicinal materials are obtained according to each independent feature model, and a feature recognition threshold is obtained. The feature recognition values of the medicinal material characteristic in various medicinal materials are compared with the feature recognition threshold. When the feature recognition value is greater than the feature recognition threshold, a feature matching record is generated. When the feature recognition value is less than or equal to the feature recognition threshold, no processing is performed.
[0013] Furthermore, the feature matching record includes the medicinal material number, medicinal material type, medicinal material characteristics, feature matching value, and record generation time.
[0014] Furthermore, the feature fit value of the feature fit record is obtained by performing a difference calculation between the feature identification value and the feature identification boundary value to obtain the feature fit value.
[0015] Furthermore, the feature recognition value of the medicinal material feature in various medicinal materials is obtained according to each independent feature model, specifically: the feature data set of the medicinal material feature is input into each independent feature model, each independent feature model outputs the feature recognition value of the medicinal material feature in one medicinal material, and then the feature recognition value of the medicinal material feature in various medicinal materials is obtained.
[0016] Furthermore, the medicinal material identification value of the medicinal material is obtained by the following method: obtaining all the feature matching records of the numbered medicinal material, obtaining the medicinal material type of the feature matching record, marking the feature matching records of the same medicinal material type as the same medicinal material feature matching records, marking the total number of the same medicinal material feature matching records as RVep, obtaining the feature matching value of the same medicinal material feature matching records, summing up the feature matching values of all the same medicinal material feature matching records and taking the average value to obtain the same medicinal material feature matching mean value LSt, setting the feature matching threshold value, when the feature matching value of the same medicinal material feature matching record is greater than or equal to the feature matching threshold value, increasing the number of highlighted matching times by one, when the feature matching value of the same medicinal material feature matching record is less than the feature matching threshold value, no processing is performed, summing up the number of highlighted matching times to obtain the total number of highlighted matching times EWy, and using the formula The medicinal material identification value SN of the medicinal material is obtained, where g1 is the coefficient of the total number of records matching the same medicinal material characteristics, g2 is the mean coefficient of the matching of the same medicinal material characteristics, and g3 is the coefficient of the total number of prominent matching times.
[0017] Furthermore, the recognition significance value is obtained in the following manner: various medicinal materials are sorted from large to small according to the numerical values of the medicinal material identification values, the medicinal material type ranked first in the sorting is marked as the target type, and the difference between the identification values of the first two medicinal materials in the sorting is calculated to obtain the recognition significance value.
[0018] Furthermore, based on the comparison result between the recognition significance value and its threshold, it is determined whether the medicinal material identification is completed, specifically: the recognition significance threshold is set, when the recognition significance value is greater than or equal to the recognition significance threshold, the medicinal material is identified as the target type, when the recognition significance value is less than the recognition significance threshold, the next medicinal material feature of the medicinal material photo is feature identified in the order of medicinal material feature recognition.
[0019] Compared with the prior art, the present invention has the following beneficial effects:
[0020] 1. Set up a medicinal material feature sequence formulation module. It is no longer necessary to identify all the appearance features of the medicinal materials simultaneously. By sequentially identifying the appearance features of the medicinal materials, it not only improves the system's recognition efficiency for the medicinal materials, but also reduces the pressure and difficulty of the system to simultaneously identify and analyze all the appearance features of the medicinal materials;
[0021] 2. Setting up a medicinal material feature recognition module and a medicinal material identification and judgment module can conduct a three-dimensional comprehensive analysis of the identification results of the medicinal materials after identifying each appearance feature of the medicinal materials, identify the types of medicinal materials in a timely manner, and reasonably control the number of appearance features of the medicinal materials that need to be analyzed on the basis of ensuring high-precision identification of the medicinal materials. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1This is a flowchart of a medicine intelligent identification system used in pharmaceutical factories;
[0023] Figure 2 A flowchart for identifying the characteristics of medicinal materials in medicinal material photos;
[0024] Figure 3 The module block diagram of a medicinal material intelligent identification system used in pharmaceutical factories. DETAILED DESCRIPTION
[0025] Reference Figures 1 to 3 , an intelligent medicinal material identification system used in pharmaceutical factories, including a medicinal material feature order formulation module, a medicinal material feature identification module, and a medicinal material identification and judgment module.
[0026] Medicinal material characteristic sequence formulation module: formulate the medicinal material characteristic identification sequence of the pharmaceutical factory (the medicinal material characteristic identification sequence is to rank the appearance characteristics of various aspects of the medicinal materials in order, where the aspects of the medicinal material characteristics include but are not limited to morphological characteristics, surface characteristics, cross-sectional characteristics, and detail characteristics).
[0027] The formulation method of the drug feature recognition order of the pharmaceutical factory is as follows: obtain all feature matching records of the pharmaceutical factory before the current time of the system, obtain the drug features of the feature matching records, mark the feature matching records with the same drug features as the same feature matching records, obtain the drug types of the same feature matching records, mark the same feature matching records of the same drug type as the same type and same feature matching records, sort all the same type and same feature matching records in the order of record generation time, sum the feature matching values of the two adjacent same type and same feature matching records after sorting, obtain the matching stability value of the drug type in the same drug feature, and set The matching stability threshold (system threshold, which can be modified according to actual needs) is set. When the matching stability value is greater than or equal to the matching stability threshold, the matching stability times are increased once. When the matching stability value is less than the matching stability threshold, no processing is performed. The matching stability times are summed to obtain the total matching stability times of the medicinal material type with the same medicinal material characteristics. The matching stability values of all medicinal material types with the same medicinal material characteristics are summed and averaged to obtain the matching stability mean Fsg. The total matching stability times of all medicinal material types with the same medicinal material characteristics are summed to obtain the total matching stability times Rhj. The formula is used. The identification matching value Dk of the medicinal material feature is obtained, where t1 is the matching stable mean coefficient, t2 is the matching stable total number coefficient, the value of t1 is 0.62, and the value of t2 is 0.59. All medicinal material features are sorted in descending order according to the numerical value of the identification matching value, and then the medicinal material feature recognition order of the pharmaceutical factory is obtained.
[0028] The feature recognition value of the medicinal material feature in various medicinal materials is obtained according to each independent feature model, specifically: the feature data set of the medicinal material feature is input into each independent feature model, each independent feature model outputs the feature recognition value of the medicinal material feature in one medicinal material, and then the feature recognition value of the medicinal material feature in various medicinal materials is obtained.
[0029] Each medicinal material feature of each medicinal material needs to be trained by an independent deep learning model. The training methods of different deep learning models are the same. It is only necessary to replace the training data. This embodiment takes the morphological features of Angelica sinensis as an example to introduce the construction method of the feature recognition model: collect n groups of feature data sets of Angelica sinensis morphological features. The above feature data sets can be actually collected or virtually set. Construct a deep learning model, use the feature data sets of n groups of Angelica sinensis morphological features as training data of the deep learning model, assign feature recognition values to the training data, divide the training data into a training set and a validation set, and the division ratio of the training set and the validation set can be set according to demand. Iteratively train the training set and the validation set to obtain a feature recognition model. Among them, the value range of the feature recognition value is (1~3). The larger the value of the feature recognition value, the closer it is to the morphological features of Angelica sinensis, and the smaller the value of the feature recognition value, the less close it is to the morphological features of Angelica sinensis. The value range of the feature recognition values of different feature recognition models is the same, both (1~3).
[0030] By setting up a module for formulating the order of medicinal material features, it is no longer necessary to identify all the appearance features of the medicinal materials simultaneously. By sequentially identifying the appearance features of the medicinal materials, it not only improves the system's recognition efficiency for the medicinal materials, but also reduces the pressure and difficulty of the system in simultaneously identifying and analyzing all the appearance features of the medicinal materials.
[0031] Medicinal material feature recognition module: Each time a medicinal material is recognized, the medicinal material is numbered (each medicinal material corresponds to an independent number), and a photo of the medicinal material is taken to obtain the medicinal material photo, and the medicinal material features in the medicinal material photo are recognized in the order of medicinal material feature recognition.
[0032] The characteristics of the medicinal materials in the medicinal material photos are identified, specifically:
[0033] Based on the photos of medicinal materials, a feature data set of the medicinal material characteristics is obtained (the feature data set contains all the feature data of the medicinal material characteristics. Taking morphological characteristics as an example, the feature data of morphological characteristics include but are not limited to length data, diameter data, thickness data, and are obtained by image feature extraction), and a feature recognition model of the medicinal material characteristics in various medicinal materials is obtained (each aspect of the medicinal material characteristics of various medicinal materials corresponds to a feature recognition model. Taking three medicinal materials as an example, the three medicinal materials are ginseng, angelica, and astragalus. Taking morphological characteristics as an example, it is necessary to obtain the feature recognition model of ginseng morphological characteristics, the feature recognition model of angelica morphological characteristics, and the feature recognition model of astragalus morphological characteristics). The feature recognition model of the state feature of various medicinal materials is used to identify the feature recognition model of the medicinal material feature in various medicinal materials, and the feature recognition value of the medicinal material feature in various medicinal materials is obtained according to each independent feature model. The feature recognition threshold value (system threshold value, which can be modified according to actual needs) is obtained, and the feature recognition value of the medicinal material feature in various medicinal materials is compared with the feature recognition threshold value. When the feature recognition value is greater than the feature recognition threshold value, a feature matching record is generated. The feature matching record includes the medicinal material number, medicinal material type, medicinal material feature, feature matching value, and record generation time. When the feature recognition value is less than or equal to the feature recognition threshold value, no processing is performed.
[0034] The feature match value of the feature match record is obtained by performing a difference calculation between the feature identification value and the feature identification threshold value to obtain the feature match value.
[0035] Medicinal material identification and judgment module: After identifying a medicinal material feature, obtain the medicinal material identification values of various medicinal materials, sort the various medicinal materials from large to small according to the medicinal material identification value, mark the medicinal material type ranked first as the target type, and perform difference calculation on the medicinal material identification values of the first two medicinal materials (subtract the medicinal material identification value of the first medicinal material from the medicinal material identification value of the second medicinal material), obtain the identification significance value, set the identification significance threshold (system threshold, which can be modified according to actual needs), when the identification significance value is greater than or equal to the identification significance threshold, identify the medicinal material as the target type (for example, if the target type is Astragalus, when the identification significance value is greater than or equal to the identification significance threshold, the medicinal material is identified as Astragalus), when the identification significance value is less than the identification significance threshold, perform feature identification on the next medicinal material feature in the medicinal material photo according to the medicinal material feature recognition order.
[0036] The medicinal material identification value of the medicinal material is obtained in the following way: obtain all the feature matching records of the medicinal material with the same number, obtain the medicinal material type of the feature matching record, mark the feature matching records of the same medicinal material type as the same medicinal material feature matching records, mark the total number of the same medicinal material feature matching records as RVep, obtain the feature matching value of the same medicinal material feature matching records, sum and average the feature matching values of all the same medicinal material feature matching records, obtain the same medicinal material feature matching mean LSt, set the feature matching threshold (system threshold, which can be modified according to actual needs), when the feature matching value of the same medicinal material feature matching record is greater than or equal to the feature matching threshold, increase the number of highlighted matches by one, when the feature matching value of the same medicinal material feature matching record is less than the feature matching threshold, do not process it, sum the number of highlighted matches, obtain the total number of highlighted matches EWy, and use the formula The medicinal material identification value SN of the medicinal material is obtained, where g1 is the total number coefficient of the same medicinal material feature matching records, g2 is the mean coefficient of the same medicinal material feature matching, and g3 is the total number of prominent matching coefficients. The value of g1 is 1.98, the value of g2 is 1.85, and the value of g3 is 2.31;
[0037] By setting up a medicinal material feature recognition module and a medicinal material identification and judgment module, after identifying each appearance feature of the medicinal material, a three-dimensional comprehensive analysis of the identification results of the medicinal material can be performed, and the type of the medicinal material can be identified in time. On the basis of ensuring high-precision identification of the medicinal material, the number of appearance features of the medicinal material that need to be analyzed can be reasonably controlled.
[0038] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.
[0039] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0040] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0041] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0042] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0043] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0044] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.
[0045] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. An intelligent identification system for medicinal materials used in pharmaceutical factories. It is characterized in that It includes a medicinal material characteristic order formulation module, a medicinal material characteristic recognition module, and a medicinal material recognition and determination module; The medicinal material characteristic order formulation module is used to formulate the medicinal material characteristic identification order of the pharmaceutical factory; The medicinal material feature recognition module numbers each medicinal material when identifying the medicinal material, takes a photo of the medicinal material, obtains the medicinal material photo, and performs feature recognition on the medicinal material features in the medicinal material photo in the order of medicinal material feature recognition; After identifying a medicinal material feature, the medicinal material identification and determination module obtains medicinal material identification values of various medicinal materials, and then obtains identification significance values, and determines whether medicinal material identification is completed based on the comparison result of the identification significance values and the threshold value; The formulation of the drug feature recognition order of the pharmaceutical factory is as follows: obtain all feature matching records of the pharmaceutical factory before the current time of the system, obtain the drug features of the feature matching records, obtain the matching stability value of each drug type in the same drug feature and the total number of matching stability of each drug type, sum up the matching stability values of all drug types with the same drug feature and take the average to obtain the matching stability mean Fsg, sum up the total number of matching stability of all drug types with the same drug feature to obtain the total number of matching stability Rhj, and use the formula The identification match value Dk of the medicinal material feature is obtained, where t1 is the matching stable mean coefficient, and t2 is the matching stable total number coefficient. All medicinal material features are sorted in descending order according to the value of the identification match value, and then the medicinal material feature recognition order of the pharmaceutical factory is obtained; Obtain the matching stability value of each medicinal material type in the same medicinal material feature and the total number of matching stability of each medicinal material type, specifically: mark the feature matching record of the same medicinal material feature as the same-feature matching record, obtain the medicinal material type of the same-feature matching record, mark the same-feature matching record of the same medicinal material type as the same-type and same-feature matching record, sort all the same-type and same-feature matching records in the order of record generation time, sum the feature matching values of the two adjacent same-type and same-feature matching records after sorting, obtain the matching stability value of the medicinal material type in the same medicinal material feature, set the matching stability threshold, when the matching stability value is greater than or equal to the matching stability threshold, increase the matching stability times by one, when the matching stability value is less than the same-feature matching stability threshold, do not process, sum the matching stability times, and obtain the total number of matching stability times of the medicinal material type in the same medicinal material feature; The characteristics of the medicinal materials in the medicinal material photos are identified, specifically: Based on the medicinal material photos, a feature data set of the medicinal material characteristics is obtained, a feature recognition model of the medicinal material characteristics in various medicinal materials is obtained, and the feature recognition models of the medicinal material characteristics in various medicinal materials are all marked as independent feature models, and feature recognition values of the medicinal material characteristics in various medicinal materials are obtained according to each independent feature model, and a feature recognition threshold is obtained, and the feature recognition values of the medicinal material characteristics in various medicinal materials are compared with the feature recognition threshold. When the feature recognition value is greater than the feature recognition threshold, a feature matching record is generated, and when the feature recognition value is less than or equal to the feature recognition threshold, no processing is performed; The characteristic matching record includes the medicinal material number, medicinal material type, medicinal material characteristics, characteristic matching value, and record generation time; The feature match value of the feature match record is obtained by: performing a difference calculation between the feature identification value and the feature identification threshold value to obtain the feature match value; Obtaining the feature recognition value of the feature of the medicinal material in various medicinal materials according to each independent feature model, specifically: inputting the feature data set of the medicinal material feature into each independent feature model, each independent feature model outputting the feature recognition value of the medicinal material feature in one medicinal material, and then obtaining the feature recognition value of the medicinal material feature in various medicinal materials; The medicinal material identification value of the medicinal material is obtained in the following way: obtain all the feature matching records of the medicinal material with the same number, obtain the medicinal material type of the feature matching record, mark the feature matching records of the same medicinal material type as the same medicinal material feature matching records, mark the total number of the same medicinal material feature matching records as RVep, obtain the feature matching value of the same medicinal material feature matching records, sum up the feature matching values of all the same medicinal material feature matching records and take the average value to obtain the same medicinal material feature matching mean value LSt, set the feature matching threshold value, when the feature matching value of the same medicinal material feature matching record is greater than or equal to the feature matching threshold value, increase the number of highlighted matching times by one, when the feature matching value of the same medicinal material feature matching record is less than the feature matching threshold value, do not process it, sum up the number of highlighted matching times to obtain the total number of highlighted matching times EWy, and use the formula The medicinal material identification value SN of the medicinal material is obtained, where g1 is the total number coefficient of the same medicinal material feature matching records, g2 is the mean coefficient of the same medicinal material feature matching records, and g3 is the total number of prominent matching coefficients; The recognition significance value is obtained in the following manner: various medicinal materials are sorted from large to small according to the numerical value of the medicinal material identification value, the medicinal material type ranked first in the sorting is marked as the target type, and the difference between the identification values of the first two medicinal materials in the sorting is calculated to obtain the recognition significance value.
2. According to claim 1, a medicinal material intelligent identification system used in a pharmaceutical factory, It is characterized in that Based on the comparison result between the recognition significance value and its threshold, it is determined whether the medicinal material identification is completed. Specifically, the recognition significance threshold is set. When the recognition significance value is greater than or equal to the recognition significance threshold, the medicinal material is identified as the target type. When the recognition significance value is less than the recognition significance threshold, the next medicinal material feature of the medicinal material photo is feature identified in the order of medicinal material feature recognition.
