An information recognition system for detecting an instrument logistics warehouse

Abstract

The application provides an information identification system for detecting instrument logistics warehouses, and relates to the technical fields of logistics storage automation and multi-source information identification, and comprises the following steps: a multi-source information acquisition module is used to acquire and preprocess radio frequency tags, two-dimensional codes, optical character information and length-width-height-weight data of detection instruments; a fusion confidence identification module is used to match the multi-source identity identification with a feature database, calculate the multi-identification fusion confidence, and determine whether the first-level identification is successful; a failure trigger analysis module is used to calculate an identification failure degree index when the first-level identification fails, and determine whether to trigger the second-level identification; and an ontology feature matching module is used to calculate the similarity based on normalized ontology feature data when the second-level identification is triggered, and output the second-level identification result. The application realizes accurate identification when the local identification information is missing, reduces the identification failure rate, and improves the system robustness and intelligent level through the multi-source identification fusion and ontology feature complementary hierarchical identification mechanism.

Description

Technical Field

[0001] This invention relates to the field of logistics and warehousing automation and multi-source information identification technology, specifically to an information identification system for a logistics warehouse for testing instruments. Background Technology

[0002] In the logistics warehouse for testing instruments, a large number of different types and specifications of testing instruments are handled daily for warehousing, outbound operations, and inventory checks. The accuracy and efficiency of the information identification process directly determine the overall operational efficiency of warehousing and logistics. Testing instruments are typically characterized by high value, complex models, and high similarity in appearance. Furthermore, a single instrument often carries multiple identification carriers, including RFID tags, QR code nameplates, and laser-engraved or affixed character sequences (Optical Character Recognition, OCR). Ideally, a single identification information can complete instrument identification. However, in the actual warehouse operating environment, RFID tags may fail to be read due to electromagnetic interference, metal shelf shielding, or tag aging; QR codes may be undecoded due to oil stains, wear, uneven lighting, or partial obstruction; and optical characters may be misidentified due to blurry fonts, tilted distortion, or insufficient contrast. When a particular identification information fails, if the system lacks an alternative identification channel, it will directly lead to identification interruption, requiring manual intervention for verification, severely impacting the logistics cycle and increasing labor costs. Therefore, how to achieve reliable and robust identification by detection instruments under the condition of partial loss or conflict of multi-source identity information has become an urgent technical problem to be solved in the intelligent upgrading of logistics warehouses.

[0003] In the prior art, patent CN107545443A discloses a warehouse management method and system based on anti-counterfeiting information codes. This technology can automatically manage the warehouse information of enterprise products. Before the goods are put into the warehouse, the product's identity information can be set to an active state by scanning the product's unique identification code, product's logistics identification code, or the product's anti-counterfeiting code and other corresponding identity information. It can also automatically repair the product's unique identification code, product's logistics identification code, or product's anti-counterfeiting code through a preset formula when the product's unique identification code, product's logistics identification code, or product's anti-counterfeiting code is incorrect, so as to prevent the inbound and outbound processes from being chaotic due to incorrect identity information.

[0004] However, the aforementioned existing technologies rely solely on single identification information. When this information becomes unavailable due to tag damage, signal interference, or parsing failure, the identification process is directly interrupted, lacking effective remediation and tiered identification mechanisms. Furthermore, this technology does not integrate the physical characteristics of the detection instrument (such as length, width, height, and weight), making secondary verification through physical features impossible when identification information is missing or conflicting, resulting in poor robustness of the identification system. In addition, the lack of quantitative evaluation indicators and intelligent triggering conditions for identification failures makes it difficult to distinguish between minor and severe failures, easily leading to unnecessary duplicate identifications or false alarms requiring manual processing, thus reducing the overall accuracy and operational efficiency of logistics warehouse information identification.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide an information identification system for a logistics warehouse of testing instruments, thereby solving the problems mentioned in the background art. This invention acquires RFID tags, QR codes, optical characters, and length, width, height, and weight data of testing instruments through a multi-source information acquisition module. A fusion confidence identification module calculates the fusion confidence of multiple identifiers and determines the first level of identification. When the first level of identification fails, a failure trigger analysis module calculates the identification failure degree index to determine whether to trigger the second level of identification. Then, an ontology feature matching module calculates the similarity based on normalized ontology features and outputs the second level of identification result, thereby improving the accuracy and robustness of identification, reducing the identification failure rate, and achieving intelligent hierarchical identification and efficient resource utilization.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] An information identification system for a testing instrument logistics warehouse includes the following functional modules:

[0009] Multi-source information acquisition module: Acquires multi-source identification information and body feature data of the detection instrument entering the recognition area. The multi-source identification information includes radio frequency tag information, QR code information, and optical character recognition information. The body feature data includes the length data, width data, height data, and weight data of the detection instrument. The module also preprocesses the acquired body feature data.

[0010] Fusion Confidence Recognition Module: Constructs a feature database for the detection instrument, matches the collected multi-source identity information with the feature database for the detection instrument, determines the matching result of each identity information, calculates the multi-identity fusion confidence based on the matching result of each identity information, and determines whether the first-level recognition of the detection instrument is successful based on the multi-identity fusion confidence.

[0011] Failure Trigger Analysis Module: When the first-level identification of the detection instrument fails, the module calculates the identification failure degree index based on the multi-identifier fusion confidence level, and determines whether to trigger the second-level identification process based on the identification failure degree index.

[0012] Ontology feature matching module: When the second-level recognition process is triggered, based on the collected ontology feature data, the similarity between the current detection instrument and each detection instrument in the detection instrument feature database is calculated, and the second-level recognition result is output based on the similarity.

[0013] Furthermore, the method for preprocessing the collected ontological feature data is as follows:

[0014] The preprocessing includes data cleaning and data normalization.

[0015] The data cleaning includes the handling of outliers and missing values. Statistical methods are used to identify outliers in the length, width, height, and weight data of the testing instrument, and outliers in the length, width, height, and weight data of the testing instrument are deleted. Missing values ​​in the ontological feature data are filled with the mean, median, or mode of the ontological feature data of similar testing instruments.

[0016] The data normalization process is as follows: the minimum-maximum normalization method is used on the cleaned ontology feature data to map the length, width, height and weight data to the range of [0,1], so that the ontology feature data have the same scale.

[0017] Furthermore, the method for constructing the feature database of the detection instrument is as follows:

[0018] Each testing instrument is assigned a unique identifier, and multi-source identification information and ontological feature data of each testing instrument are collected. A corresponding mapping relationship is established between the unique identifier, multi-source identification information and ontological feature data of each testing instrument, and the data is stored in the testing instrument feature database in the form of a structured data table.

[0019] Furthermore, the method for determining the matching results of each identity identifier is as follows:

[0020] The total number of types of multi-source identification information collected by the detection instruments is set to be , No. The matching result of the identity information is denoted as ;

[0021] When the first [data] is successfully collected and parsed If the parsed identity information is completely consistent with the corresponding identity information of a certain testing instrument in the testing instrument feature database, the matching result of this identity information is considered a successful match. ;

[0022] When the first [data] is successfully collected and parsed If the parsed identity information is inconsistent with the corresponding identity information of all detection instruments in the detection instrument feature database, the matching result of this identity information is judged as a failure. ;

[0023] When the first [item] could not be collected When the collected identity information cannot be parsed to obtain valid identity information, or when the collected identity information cannot be parsed to obtain valid identity information. .

[0024] Furthermore, the formula used to calculate the multi-identifier fusion confidence score is as follows:

[0025]

[0026] in, For multi-identifier fusion confidence;

[0027] For the first Preset weighting coefficients corresponding to various identity information.

[0028] Furthermore, the determination logic for whether the first-level identification of the detection instrument is successful is as follows:

[0029] when ≥ When the first level of identification is successful, the multi-source identity information of the detection instrument corresponding to the matching result is output from the detection instrument feature database.

[0030] when < If this occurs, the first level of recognition is deemed to have failed.

[0031] in, The preset confidence threshold is used, and >0.

[0032] Furthermore, the formula used to calculate the failure rate index is as follows:

[0033]

[0034] in, To identify the degree of failure index.

[0035] Furthermore, the determination logic for whether the second-level identification process is triggered is as follows:

[0036] When 0 < ≤ When this occurs, the second-level recognition process is triggered;

[0037] when > If the second-level recognition process is not triggered, the first manual processing signal is output.

[0038] in, The preset failure threshold is 0 < <1.

[0039] Furthermore, the formula used to calculate the similarity between the current detection instrument and each detection instrument in the detection instrument feature database is as follows:

[0040] , in, The first in the current database of testing instruments and their characteristics Similarity of the testing instruments;

[0041] This refers to the index number of the testing instrument in the testing instrument feature database;

[0042] , , , These are the normalized values ​​of the length, width, height, and weight data of the current testing instrument.

[0043] , , , These are the first items in the detection instrument feature database. The length, width, height, and weight data of the testing instrument are normalized values.

[0044] Furthermore, the method used to output the second-level recognition result based on similarity is as follows:

[0045] Calculate the similarity between the current testing instrument and each testing instrument in the testing instrument feature database. And the calculated maximum similarity is denoted as ;

[0046] when ≥ At that time, the detection instrument with the highest similarity in the detection instrument feature database is used as the output result of the second-level identification, and the multi-source identity information of the detection instrument with the highest similarity is output.

[0047] when < If the second-level identification fails to match, a second manual processing signal is output.

[0048] in, This is the preset similarity threshold.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] This invention collects multi-source identification information and ontological feature data of testing instruments through a multi-source information acquisition module. It then calculates the multi-identifier fusion confidence level using a fusion confidence recognition module for first-level identification. When first-level identification fails, a failure trigger analysis module calculates the identification failure degree index and intelligently determines whether to trigger a second-level identification process. If triggered, an ontological feature matching module calculates the similarity based on normalized length, width, height, and weight data and outputs the second-level identification result. This significantly improves the accuracy and robustness of information identification in the logistics warehouse of testing instruments, reduces the identification failure rate caused by missing, incorrect, or unresolved single identification information, and avoids unnecessary computational resource consumption through a hierarchical identification strategy. For severe failures, it promptly outputs manual processing signals, reducing the risk of misjudgment in automated identification, improving the overall system operating efficiency and intelligence level, and achieving efficient fusion and complementary utilization of multi-source heterogeneous information. Attached Figure Description

[0051] Figure 1 This is a block diagram of an information identification system for a logistics warehouse of testing instruments.

[0052] Figure 2 This is a schematic diagram of the operation process of an information identification system for a logistics warehouse of a testing instrument. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0054] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0055] Example:

[0056] Please see Figures 1-2 The present invention provides a technical solution:

[0057] An information identification system for a testing instrument logistics warehouse includes the following functional modules:

[0058] Multi-source information acquisition module: This module acquires multi-source identification information and body feature data of the detection instruments entering the identification area. The multi-source identification information includes RFID tag information, QR code information, and optical character recognition information. The RFID tag information consists of the unique serial number and encoded data such as instrument model and production batch number stored in the RFID tag attached or embedded on the detection instrument. The QR code information consists of the instrument identification, serial number, and verification data encoded by the QR code graphic printed or pasted on the surface of the detection instrument. The optical character recognition information consists of visually identifiable text information such as the instrument number, model code, and production date presented in character form on the instrument's casing or nameplate. The body feature data includes the length, width, height, and weight data of the detection instrument. Specifically, it consists of the external dimensions measured along three mutually perpendicular directions in a standard posture, as well as the overall mass of the instrument. The acquired body feature data is preprocessed to improve data quality and consistency.

[0059] The method for preprocessing the collected ontological feature data is as follows:

[0060] The preprocessing includes data cleaning and data normalization.

[0061] The data cleaning process includes handling outliers and missing values. Statistical methods are used to identify outliers in the length, width, height, and weight data of the testing instruments. Specifically, the three-standard-deviation criterion is employed: the mean and standard deviation of each dimension are calculated, and data deviating from the mean by more than three standard deviations are identified as outliers and deleted. This ensures that subsequent analysis is not affected by data with extreme deviations. Simultaneously, for missing values ​​in the ontological feature data, the system searches the database for ontological feature data of testing instruments belonging to the same category and calculates the mean, median, or mode of these similar data to fill in the current missing values. The mean is preferred when the data distribution is roughly symmetrical and there are no significant outliers; the median is preferred when the data distribution is skewed or there are a few residual outliers; and the mode is preferred when the data is discrete or has multiple clustered categories. Through these methods, the completeness and reliability of the ontological feature data can be effectively improved, laying a consistent data foundation for subsequent identification and calculation.

[0062] The data normalization process is as follows: The cleaned ontology feature data is normalized using a min-max normalization method, mapping length, width, height, and weight data to the range [0,1], ensuring all ontology feature data have the same scale. This normalization process first determines the actual minimum and maximum values ​​of the four dimensions (length, width, height, and weight) in the cleaned dataset. Then, a linear transformation is performed on each original value of each dimension, i.e., the value is subtracted from the minimum value of the dimension, and then divided by the difference between the maximum and minimum values ​​of that dimension, thereby proportionally compressing the original values ​​to [the specified scale]. Within the range, after transformation, features that were originally incomparable due to different units (e.g., length, width, and height in meters or centimeters, weight in kilograms or grams) or vastly different numerical distribution ranges (e.g., length might be a decimal, weight might be tens to hundreds) are all unified to a dimensionless scale with consistent amplitude. Therefore, each dimension of ontological feature data makes a balanced contribution to subsequent similarity calculations or distance measurements, avoiding the problem of features with large numerical ranges dominating the results while features with small numerical ranges are ignored, thus improving the consistency and comparability of feature representation.

[0063] Fusion Confidence Recognition Module: Constructs a feature database for detection instruments, that is, pre-establishes a structured record for each identifiable detection instrument, including its multiple identity information and ontological features, to ensure that the feature database covers all possible types of detection instruments and their corresponding multi-source identity information and ontological features. The collected multi-source identity information is matched with the feature database of detection instruments to determine the matching results of each identity information. Based on the matching results of each identity information, the multi-identifier fusion confidence score, which can comprehensively reflect the overall matching confidence, is calculated. Based on the multi-identifier fusion confidence score, it is determined whether the first-level recognition of the detection instrument is successful.

[0064] The method for constructing the feature database of the detection instrument is as follows:

[0065] Each testing instrument is assigned a unique identifier, which is globally unique and can be generated sequentially according to the instrument's entry order, model classification, or custom coding rules, ensuring that each instrument can be independently distinguished and traced. Subsequently, the system collects multi-source identification information for each testing instrument, including but not limited to the electronic code stored in the RFID tag, the string sequence carried by the QR code, and the instrument nameplate or engraved characters obtained through optical character recognition; simultaneously, it collects the instrument's physical characteristics data, namely its length, width, height, weight, and other physical dimensions and mass attributes. After acquiring all the above information, the system establishes a corresponding mapping relationship between each testing instrument's unique identifier, multi-source identification information, and physical characteristics data, ensuring accurate association and no confusion among the three. Finally, these mapping relationships are stored in the testing instrument characteristic database in the form of a structured data table. Each row of the data table corresponds to one testing instrument, and each column records the instrument's unique identifier, the specific values ​​of various identification information, and the values ​​of various physical characteristics data, thus forming a standardized, complete, and easily searchable and matching instrument characteristic information database.

[0066] The method used to determine the matching results of each identity identifier is as follows:

[0067] The total number of types of multi-source identification information collected by the detection instruments is set to be , No. The matching result of the identity information is denoted as ;

[0068] When the first [data] is successfully collected and parsed If the parsed identity information is completely consistent with the corresponding identity information of a certain testing instrument in the testing instrument feature database, the matching result of this identity information is considered a successful match. This situation indicates that the identification information is not only effectively obtained, but also precisely corresponds to the record in the feature database of the detection instrument, and can independently support identification;

[0069] When the first [data] is successfully collected and parsed If the parsed identity information is inconsistent with the corresponding identity information of all detection instruments in the detection instrument feature database, the matching result of this identity information is judged as a failure. This situation indicates that the identification information itself is valid, but a matching object cannot be found in the detection instrument's feature database, which may be due to reasons such as incorrect information, missing records, or the instrument not being registered.

[0070] When the first [item] could not be collected When the collected identity information cannot be parsed to obtain valid identity information, or when the collected identity information cannot be parsed to obtain valid identity information. This situation indicates that the identity information is unavailable in the current identification process. It is neither considered a successful match nor a clear failure to match, but rather an intermediate ambiguous state that participates in the subsequent calculation of fusion confidence.

[0071] The formula used to calculate the confidence score of multi-identifier fusion is as follows:

[0072] , in, For multi-identifier fusion confidence;

[0073] For the first Preset weighting coefficients corresponding to various identity information, weighting coefficients The introduction of this feature allows different levels of importance to be assigned to different identification information: higher weights can be set for identification types with high reliability and strong anti-interference capabilities (such as RFID tags in environments without electromagnetic shielding), while lower weights can be set for identifications that are susceptible to environmental factors (such as optical characters in poor lighting conditions), so that the fusion confidence level is more in line with the confidence characteristics of different sensors in actual working scenarios.

[0074] Multi-identifier fusion confidence The larger the value, the higher the overall matching degree between the collected multi-source identity information and the feature database of the detection instrument, and the stronger the reliability of the identification; conversely, the smaller the value, the lower the value. The smaller the value, the more failed matches or unobtainable identifiers there are in the multi-source identity information, resulting in a lower overall confidence level.

[0075] The logic for determining whether the first-level identification of the detection instrument is successful is as follows:

[0076] when ≥ When the system detects that the current multi-source identity information collected matches the detection instrument feature database to a level of reliability recognized by the system, the first-level identification is determined to be successful, and the multi-source identity information of the detection instrument corresponding to the matching result in the detection instrument feature database is output.

[0077] when < If the current multi-source identity information collected fails to meet the preset reliability requirements in the feature database of the detection instrument, it means that the identity of the detection instrument cannot be uniquely and accurately determined based on the existing multi-source identity information, and the first-level identification is judged to have failed.

[0078] in, The preset confidence threshold is used, and >0.

[0079] Failure Trigger Analysis Module: When the first-level identification of the detection instrument fails, the failure index is calculated based on the confidence level of multi-identifier fusion. This failure index can reflect the difference between the first-level identification result and the confidence threshold, thereby providing a quantitative basis for whether to enable a more in-depth identification process. Based on the failure index, it is determined whether to trigger the second-level identification process.

[0080] The formula used to calculate the failure rate index is as follows:

[0081] , in, To identify the degree of failure index;

[0082] This item reflects the relative deviation between the multi-identifier fusion confidence level and the preset confidence threshold. The closer hour, A value close to 0 indicates a milder degree of failure in the first-level identification, only slightly below the reliability level determined by the system; when... much smaller hour, A value close to 1 indicates a severe failure rate, meaning the reliability of matching multi-source identity information with the detection instrument's feature database is extremely low.

[0083] The logic for determining whether the second-level identification process has been triggered is as follows:

[0084] When 0 < ≤ When the first level of identification fails, it indicates that the failure is within a controllable or remedial range, and the second level of identification process is triggered to attempt to complete the identification through other information channels.

[0085] when > When the first level of identification fails, it indicates that the failure is quite serious and exceeds the limit that can be remedied by subsequent identification processes. The second level of identification process is not triggered, and the first manual processing signal is output to prompt manual intervention to verify the collection status of multi-source identity information or the integrity of the identification carrier.

[0086] in, The preset failure threshold is 0 < <1.

[0087] Ontology feature matching module: When the second-level recognition process is triggered, based on the collected ontology feature data, the similarity between the current detection instrument and each detection instrument in the detection instrument feature database is calculated. This similarity comprehensively reflects the overall closeness between the current detection instrument and each instrument in the detection instrument feature database in terms of geometric size and weight characteristics, and outputs the second-level recognition result based on the similarity.

[0088] The formula used to calculate the similarity between the current testing instrument and each testing instrument in the testing instrument feature database is as follows:

[0089] , in, The first in the current database of testing instruments and their characteristics Similarity of the testing instruments;

[0090] This refers to the index number of the testing instrument in the testing instrument feature database;

[0091] , , , These are the normalized values ​​of the length, width, height, and weight data of the current testing instrument.

[0092] , , , These are the first items in the detection instrument feature database. The normalized values ​​of the length, width, height, and weight data of the testing instrument.

[0093] This indicates the current testing instrument and the first item in the testing instrument feature database. The Euclidean distance between the two testing instruments in the four-dimensional feature space of the instrument. The smaller the Euclidean distance, the smaller the overall deviation between the two testing instruments in the four dimensions of length, width, height, and weight, and the closer their overall features are; conversely, the larger the Euclidean distance, the greater the overall deviation and the more significant the feature differences.

[0094] When the ontological features of the current detection instrument in all dimensions are completely consistent with the corresponding features of a certain detection instrument in the detection instrument feature database, the Euclidean distance is zero, and the similarity is zero. The Euclidean distance reaches its maximum value, indicating that the two instruments are completely identical in terms of ontological features, and the matching confidence is the highest. When there are deviations between the ontological features of the current detection instrument and the corresponding features of a certain detection instrument in the detection instrument feature database, the Euclidean distance is greater than zero, and the similarity is high. The similarity is less than 1, and the larger the Euclidean distance, the higher the similarity. The smaller the difference, the greater the difference in size or weight between the two instruments, and the lower the reliability of the match.

[0095] The method used to output the second-level recognition result based on similarity is as follows:

[0096] Calculate the similarity between the current testing instrument and each testing instrument in the testing instrument feature database. And select the maximum value from all calculated similarities, denoted as . ;

[0097] when ≥ At that time, the detection instrument with the maximum similarity in the detection instrument feature database is output as the most likely matching result of the second-level identification, and the multi-source identity information of the detection instrument with the maximum similarity is also output. It should be noted that the matching result is calculated based on the numerical similarity of the ontological features and is not an absolutely accurate judgment. In practical applications, it can be used as a high-confidence reference identification result for subsequent steps.

[0098] when < If the current detection instrument fails to meet the matching requirements, it indicates that even the most similar records in the detection instrument feature database cannot be matched. The current detection instrument cannot establish a reliable correspondence with any instrument in the detection instrument feature database through its own features. The second level of identification is determined to be unmatched, and a second manual processing signal is output, indicating that manual intervention is required to confirm whether the detection instrument belongs to a new type of detection instrument not included in the detection instrument feature database.

[0099] in, This is the preset similarity threshold.

[0100] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0101] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0102] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. An information identification system for a logistics warehouse for testing instruments, characterized in that, Includes the following functional modules: Multi-source information acquisition module: Acquires multi-source identification information and body feature data of the detection instrument entering the recognition area. The multi-source identification information includes radio frequency tag information, QR code information, and optical character recognition information. The body feature data includes the length data, width data, height data, and weight data of the detection instrument. The module also preprocesses the acquired body feature data. Fusion Confidence Recognition Module: Constructs a feature database for the detection instrument, matches the collected multi-source identity information with the feature database for the detection instrument, determines the matching result of each identity information, calculates the multi-identity fusion confidence based on the matching result of each identity information, and determines whether the first-level recognition of the detection instrument is successful based on the multi-identity fusion confidence. Failure Trigger Analysis Module: When the first-level identification of the detection instrument fails, the module calculates the identification failure degree index based on the multi-identifier fusion confidence level, and determines whether to trigger the second-level identification process based on the identification failure degree index. Ontology feature matching module: When the second-level recognition process is triggered, based on the collected ontology feature data, the similarity between the current detection instrument and each detection instrument in the detection instrument feature database is calculated, and the second-level recognition result is output based on the similarity.

2. The information identification system for a logistics warehouse of testing instruments according to claim 1, characterized in that: The method for preprocessing the collected ontological feature data is as follows: The preprocessing includes data cleaning and data normalization. The data cleaning includes the handling of outliers and missing values. Statistical methods are used to identify outliers in the length, width, height, and weight data of the testing instrument, and outliers in the length, width, height, and weight data of the testing instrument are deleted. Missing values ​​in the ontological feature data are filled with the mean, median, or mode of the ontological feature data of similar testing instruments. The data normalization process is as follows: the minimum-maximum normalization method is used on the cleaned ontology feature data to map the length, width, height and weight data to the range of [0,1], so that the ontology feature data have the same scale.

3. The information identification system for a logistics warehouse of testing instruments according to claim 1, characterized in that: The method for constructing the feature database of the detection instrument is as follows: Each testing instrument is assigned a unique identifier, and multi-source identification information and ontological feature data of each testing instrument are collected. A corresponding mapping relationship is established between the unique identifier, multi-source identification information and ontological feature data of each testing instrument, and the data is stored in the testing instrument feature database in the form of a structured data table.

4. The information identification system for a logistics warehouse of testing instruments according to claim 3, characterized in that: The method used to determine the matching results of each identity identifier is as follows: The total number of types of multi-source identification information collected by the detection instruments is set to be , No. The matching result of the identity information is denoted as ; When the first [data] is successfully collected and parsed If the parsed identity information is completely consistent with the corresponding identity information of a certain testing instrument in the testing instrument feature database, the matching result of this identity information is considered a successful match. ; When the first [data] is successfully collected and parsed If the parsed identity information is inconsistent with the corresponding identity information of all detection instruments in the detection instrument feature database, the matching result of this identity information is judged as a failure. ; When the first [item] could not be collected When the collected identity information cannot be parsed to obtain valid identity information, or when the collected identity information cannot be parsed to obtain valid identity information. .

5. The information identification system for a logistics warehouse of testing instruments according to claim 4, characterized in that: The formula used to calculate the confidence score of multi-identifier fusion is as follows: in, For multi-identifier fusion confidence; For the first Preset weighting coefficients corresponding to various identity information.

6. The information identification system for a logistics warehouse of testing instruments according to claim 5, characterized in that: The logic for determining whether the first-level identification of the detection instrument is successful is as follows: when ≥ When the first level of identification is successful, the multi-source identity information of the detection instrument corresponding to the matching result is output from the detection instrument feature database. when < If this occurs, the first level of recognition is deemed to have failed. in, The preset confidence threshold is used, and >

0.

7. The information identification system for a logistics warehouse of testing instruments according to claim 6, characterized in that: The formula used to calculate the failure rate index is as follows: in, To identify the degree of failure index.

8. The information identification system for a logistics warehouse of testing instruments according to claim 7, characterized in that: The logic for determining whether the second-level identification process has been triggered is as follows: When 0 < ≤ When this occurs, the second-level recognition process is triggered; when > If the second-level recognition process is not triggered, the first manual processing signal is output. in, The preset failure threshold is 0 < <1.

9. The information identification system for a logistics warehouse of testing instruments according to claim 2, characterized in that: The formula used to calculate the similarity between the current testing instrument and each testing instrument in the testing instrument feature database is as follows: in, The first in the current database of testing instruments and their characteristics Similarity of the testing instruments; This refers to the index number of the testing instrument in the testing instrument feature database; , , , These are the normalized values ​​of the length, width, height, and weight data of the current testing instrument. , , , These are the first items in the detection instrument feature database. The length, width, height, and weight data of the testing instrument are normalized values.

10. The information identification system for a logistics warehouse of testing instruments according to claim 9, characterized in that: The method used to output the second-level recognition result based on similarity is as follows: Calculate the similarity between the current testing instrument and each testing instrument in the testing instrument feature database. And the calculated maximum similarity is denoted as ; when ≥ At that time, the detection instrument with the highest similarity in the detection instrument feature database is used as the output result of the second-level identification, and the multi-source identity information of the detection instrument with the highest similarity is output. when < If the second-level identification fails to match, a second manual processing signal is output. in, This is the preset similarity threshold.

Images