Laboratory safety certification method and apparatus based on lims system

Abstract

The application discloses a laboratory safety authentication method and device based on a Lims system, relates to the technical field of safety authentication, and solves the technical problem that the safety authentication system in the prior art is difficult to adapt to the safety authentication of specific conditions of relevant personnel, thereby reducing the efficiency of safety authentication. The method comprises the following steps: obtaining reservation information of a laboratory, generating a reference threshold value based on reservation item information in the reservation information; obtaining verification information corresponding to a pilot verification item and a plurality of verification items of the laboratory; generating a pilot verification result based on the verification information corresponding to the pilot verification item and an influence coefficient corresponding to each verification item; generating an item matching degree based on the verification information corresponding to each verification item; and generating a final authentication result based on the item matching degree, the influence coefficient and an item reference threshold value of each verification item. The accuracy of safety authentication is improved, and the efficiency of safety authentication is improved.

Description

Technical Field

[0001] This application belongs to the field of safety certification technology, specifically a laboratory safety certification method and device based on the LIMS system. Background Technology

[0002] To ensure the safety of the experimental environment and the standardization of management, laboratories typically employ security authentication systems. These systems reliably and securely protect people's lives and property, enabling strict control over designated areas through methods such as card swiping, fingerprints, passwords, or facial recognition. Access control systems allow for real-time monitoring of the laboratory, effectively controlling access for relevant personnel and preventing unauthorized entry, thus ensuring laboratory security. Laboratories often store valuable equipment, hazardous chemicals, or important research data; stringent access control measures prevent theft, vandalism, or accidents.

[0003] Existing technology (invention patent with announcement number CN116416726B) discloses a high-security access control identification method and system based on multi-feature verification, including the following steps: S1 voice data acquisition, S2 entrance personnel data acquisition, S3 determining the personnel to be compared, and S4 verification step; the present invention proposes a high-security access control identification method and system based on multi-feature verification. By using the vector length as a reference value, the initial comparison target and the final comparison target are determined by the voice content and voice features respectively. This can quickly determine possible comparison objects, simplifying the original comparison method that requires traversing the entire database to a method that only requires comparing a few targets, greatly improving the comparison efficiency.

[0004] The aforementioned safety authentication system increases the accuracy of recognition through multiple voice analyses. However, due to the unique nature of laboratories, different experiments may present different hazards, necessitating appropriate attire such as masks, goggles, and protective clothing before entry. In such cases, relying solely on voice recognition is hampered by the masks and protective clothing, which significantly alter the voice characteristics of personnel. This reduces the accuracy of voice-based authentication, making it difficult to adapt to specific situations and thus reducing efficiency. Therefore, a laboratory safety authentication method and device based on a Lims system is needed. Summary of the Invention

[0005] This application provides a laboratory safety certification method and apparatus based on the LIMS system, which solves the technical problem that existing safety certification systems are difficult to adapt to the specific circumstances of relevant personnel for safety certification, resulting in reduced efficiency of safety certification.

[0006] To achieve the above objectives, this application adopts the following technical solution: Firstly, it provides a laboratory safety certification method based on the LIMS system, including: Obtain the laboratory's reservation information, which includes reservation time slot information, reservation personnel information, and reservation project information. The reservation project information includes experimental project information and experimental requirements information, including dress requirements. A reference threshold is generated based on the reservation item information in the reservation information; the reference threshold includes the item reference threshold for each verification item. Obtain the pilot verification project and verification information corresponding to several verification projects in the laboratory; the pilot verification project is the overall recognition image; the verification projects include face recognition, voice recognition, and handwriting recognition. Based on the verification information corresponding to the pilot verification projects, generate pilot verification results and the impact coefficients corresponding to each verification project; Generate project matching degree based on the verification information corresponding to each verification project; The final certification result is generated based on the project matching degree, impact coefficient, and project reference threshold of each verification project.

[0007] Based on the above technical solution, the laboratory safety certification method and apparatus based on the LIMS system provided in this application obtains the laboratory's reservation information and generates a reference threshold based on the reservation project information in the reservation information; obtains the laboratory's pilot verification projects and verification information corresponding to several verification projects; generates pilot verification results and influence coefficients corresponding to each verification project based on the verification information corresponding to the pilot verification projects; generates project matching degree based on the verification information corresponding to each verification project; and generates the final certification result based on the project matching degree, influence coefficient, and project reference threshold of each verification project. The threshold required for certification is adaptively set according to the specific circumstances of relevant personnel, and multiple projects are used for multiple certifications. The adaptive setting of the certification analysis scheme increases the accuracy of safety certification, thereby increasing the efficiency of safety certification.

[0008] In conjunction with the first aspect above, in one possible implementation, generating the reference threshold based on the reservation item information in the reservation information includes: Extract the dress code information from the appointment project information, and based on the dress code in the dress code information, query the influence coefficient of each verification project under the dress code in the influence coefficient table; Obtain the project threshold for each validation project; adjust the corresponding project threshold based on the impact coefficient of the validation project to generate a project reference threshold; integrate each validation project and its corresponding project reference threshold into a reference threshold.

[0009] In conjunction with the first aspect above, one possible implementation of the influence coefficient table includes: Obtain the matching degree of several items corresponding to each verification item under the dress code in the dress requirement information; obtain the matching degree of standard items corresponding to each verification item under the standard dress requirement; Calculate the average matching degree of several items under the dress label for the verification item, and record the ratio of the average matching degree of the average to the standard matching degree of the verification item as the influence coefficient of the verification item under the dress label; obtain the influence coefficient of each verification item under the dress label in sequence; Obtain the influence coefficients of each verification item under each clothing label in sequence; integrate several clothing labels, verification items and their corresponding influence coefficients into an influence coefficient table.

[0010] In conjunction with the first aspect above, in one possible implementation, the project reference threshold is generated by adjusting the project threshold based on the influence coefficient, including: Obtain the project threshold and impact coefficient for each verification project, and use the product of the impact coefficient and the project threshold as the project reference threshold for the verification project.

[0011] In conjunction with the first aspect above, in one possible implementation, generating the pilot verification results and the influence coefficients corresponding to each verification item based on the verification information corresponding to the pilot verification item includes: Extract the recognition image from the verification information corresponding to the pilot verification project, and input the recognition image into the clothing recognition model to obtain the clothing label. If the dress code does not match the dress code in the dress requirement information, the preliminary verification result is recorded as dress abnormal; otherwise, the preliminary verification result is recorded as dress normal. Obtain the dress code, and based on the dress code, query the influence coefficient table for each verification item under the dress code; The pilot validation results and the impact coefficients of each validation item are integrated into the pilot validation results.

[0012] In conjunction with the first aspect above, in one possible implementation, one training method for the clothing recognition model includes: Acquire several recognition images and corresponding clothing tags for the recognition images. The clothing tags are corresponding tags set by professionals for the clothing of the people in the recognition images, including ordinary clothing, ordinary protective clothing, heat-insulating protective clothing, and cleanroom protective clothing, etc.; integrate the several recognition images and their corresponding clothing tags into several training data and test data. The artificial intelligence model is trained using training data and tested using testing data. The final result is a clothing recognition model with the input being a recognition image and the output being the clothing label corresponding to the recognition image. The artificial intelligence model includes recurrent neural network models, etc.

[0013] In conjunction with the first aspect above, in one possible implementation, generating the project matching degree based on the verification information corresponding to each verification project includes: The verification item is extracted from the facial data in the verification information corresponding to the facial item; the facial data includes facial image data taken during the verification of the corresponding person; the reference facial data in the appointment information is extracted from the appointment information; the facial matching degree corresponding to the facial item is generated based on the reference facial data and the facial data; The audio data corresponding to the verification item is extracted from the verification information; the audio data includes audio data recorded by the corresponding personnel during verification; reference audio data is extracted from the appointment information of the appointment personnel; and the audio matching degree corresponding to the sound item is generated based on the reference audio data and the audio data. The handwriting data corresponding to the verification item is extracted and verified; the handwriting data includes the signature data of the corresponding personnel during verification; reference handwriting data is extracted from the appointment information of the appointment personnel; and the handwriting matching degree corresponding to the handwriting item is generated based on the reference handwriting data and the handwriting data. The project matching degree includes face matching degree, audio matching degree and handwriting matching degree.

[0014] In conjunction with the first aspect above, in one possible implementation, generating the face matching degree based on reference face data and face data includes: Extract the reference face image from the reference face data and the face image to be identified from the face data. Align the face image to be identified with the reference face image. Input the face image to be identified into the occlusion recognition model to obtain the occlusion recognition image. The occlusion recognition image is formed by marking the occluded area in the face image. Several key detection points are set in the occlusion recognition image. Extract features from the reference face image at each detection key point and integrate them into a reference feature vector in a set order; extract features from the face image to be identified at each detection key point and integrate them into a recognition feature vector in a set order. Calculate the cosine similarity between the identified feature vector and the reference feature vector, and use the cosine similarity as the face matching degree.

[0015] In conjunction with the first aspect above, in one possible implementation, the key detection points are set in the following way: The occlusion recognition image is rasterized to obtain several grids; the area of ​​the occlusion region in each grid is obtained; the area is substituted into the key point number adjustment function to obtain the number of key points K corresponding to the grid; Obtain the minimum Euclidean distance between each pixel of the grid and the occluded area; take the pixel with the largest minimum Euclidean distance as the core pixel of the grid; uniformly set K detection key points in the grid with the core pixel as the center; set the detection key points of each grid in sequence.

[0016] In conjunction with the first aspect above, in one possible implementation, generating the audio matching degree based on the reference audio data and the audio data includes: Extract the reference audio from the reference audio data and the audio to be identified from the audio data; calculate the cosine similarity between the audio to be identified and the reference audio, and use the cosine similarity as the audio matching degree.

[0017] In conjunction with the first aspect above, in one possible implementation, generating the final authentication result based on the project matching degree, influence coefficient, and project reference threshold of each verification project includes: S1: Extract the face matching score, audio matching score, and handwriting matching score from the item matching score; and the item reference threshold corresponding to each verification item in the reference threshold; S2: Based on face matching, audio matching, and handwriting matching, and the reference thresholds for each verification item in the reference thresholds, the authentication results for each verification item are generated; S3: If the certification results for each verification item are all passed, proceed to S6; otherwise, proceed to S4. S4: Obtain the impact coefficient of each validation item; process the impact coefficient of each validation item using the weight generation function to obtain the weight coefficient corresponding to each validation item; perform a weighted summation of the item matching degree of each validation item based on each weight coefficient to obtain the comprehensive matching degree; perform a weighted summation of the item reference threshold of each validation item based on each weight coefficient to obtain the comprehensive reference threshold; S5: Determine whether the overall matching degree is greater than the overall reference threshold. If yes, proceed to S6; otherwise, proceed to S7. S6: Obtain the current time and the reservation time period in the reservation information; if the current time is within the reservation time period, set the final authentication result to pass; otherwise, proceed to S7; S7: Set the final authentication result to failed.

[0018] Secondly, a laboratory safety certification device based on the LIMS system is provided, including: a data acquisition unit, a data processing unit, and a display unit; Data acquisition unit: used to acquire the laboratory's reservation information, as well as the laboratory's pilot validation project and validation information for several validation projects; Data processing unit: used to generate reference thresholds based on the reservation item information in the reservation information; the reference thresholds include item reference thresholds for each verification item; and to generate preliminary verification results and influence coefficients for each verification item based on the verification information corresponding to the preliminary verification item; to generate item matching degree based on the verification information corresponding to each verification item; and to generate the final authentication result based on the item matching degree, influence coefficient, and item reference threshold of each verification item. Display unit: Used to display the final authentication result and the preliminary verification result.

[0019] Thirdly, this application provides a LIMS-based laboratory safety authentication device, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is configured to execute the instructions to implement the methods described in the first aspect and any possible implementation thereof. This LIMS-based laboratory safety authentication device can be an electronic device or a chip within an electronic device.

[0020] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a LIMS-based laboratory safety certification device, cause the LIMS-based laboratory safety certification device to perform the methods described in the first aspect and any possible implementation thereof.

[0021] Fifthly, this application provides a computer program product containing instructions that, when run on a LIMS-based laboratory safety certification device, causes the LIMS-based laboratory safety certification device to perform the methods described in the first aspect and any possible implementation thereof.

[0022] This application provides a laboratory safety certification method and apparatus based on a LIMS system. It can acquire laboratory reservation information and generate reference thresholds based on the reservation project information; acquire pilot verification projects and corresponding verification information for several verification projects; generate pilot verification results and influence coefficients for each verification project based on the verification information for the pilot verification projects; generate project matching degrees based on the verification information for each verification project; and generate the final certification result based on the project matching degrees, influence coefficients, and reference thresholds for each verification project. The method adaptively sets the required certification thresholds according to the specific circumstances of relevant personnel, and employs multiple projects for multiple certifications, adaptively setting the certification analysis scheme to increase the accuracy and efficiency of safety certification.

[0023] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram illustrating the steps of the security authentication method in this application; Figure 2 This is a flowchart illustrating the process of generating the authentication results in this application. Figure 3 This is a schematic diagram of the unit connections of the security authentication device in this application. Detailed Implementation

[0026] The technical solutions of this application will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0027] Please see Figure 1 The first aspect of this application provides a laboratory safety certification method based on a LIMS system, comprising: Obtain laboratory reservation information, which includes the reservation time slot, the person making the reservation, and the reservation project information. The reservation project information includes the experimental project information and experimental requirements information, including dress requirements. The reservation information comes from the relevant database provided by the LIMS system. All data generated in this embodiment will be uploaded and stored in the database. Reference thresholds are generated based on the reservation item information in the reservation information; the reference thresholds include the item reference thresholds for each verification item. The process involves acquiring pilot validation projects and validation information corresponding to several validation projects in the laboratory; the pilot validation project is a recognition image of the whole recognition; the validation projects include face recognition, voice recognition, and handwriting recognition; generating pilot validation results and influence coefficients corresponding to each validation project based on the validation information corresponding to the pilot validation projects; and generating project matching degrees based on the validation information corresponding to each validation project. The final certification result is generated based on the project matching degree, impact coefficient, and project reference threshold of each verification project.

[0028] It is worth noting that the specific management environment of the laboratory's safety certification method is based on the Lims system, and the actual calculations and program operations are all carried out in the Lims system; all kinds of instructions generated are also sent by the Lims system.

[0029] Based on the above technical solution, the laboratory safety certification method and apparatus based on the LIMS system provided in this application obtains the laboratory's reservation information and generates a reference threshold based on the reservation project information in the reservation information; obtains the laboratory's pilot verification projects and verification information corresponding to several verification projects; generates pilot verification results and influence coefficients corresponding to each verification project based on the verification information corresponding to the pilot verification projects; generates project matching degree based on the verification information corresponding to each verification project; and generates the final certification result based on the project matching degree, influence coefficient, and project reference threshold of each verification project. The threshold required for certification is adaptively set according to the specific circumstances of relevant personnel, and multiple projects are used for multiple certifications. The adaptive setting of the certification analysis scheme increases the accuracy of safety certification, thereby increasing the efficiency of safety certification.

[0030] In one possible implementation, a reference threshold is generated based on the reservation item information in the reservation information, including: extracting the dress requirement information from the reservation item information, and querying the influence coefficient corresponding to each verification item under the dress label in the influence coefficient table based on the dress label in the dress requirement information. Obtain the project threshold for each validation item; the project threshold is a standard threshold set manually for the matching degree of each validation item. When the matching degree of a validation item exceeds the project threshold, it indicates that the validation item has passed the validation; adjust the corresponding project threshold based on the impact coefficient of the validation item to generate a project reference threshold; integrate each validation item and its corresponding project reference threshold into a reference threshold; specifically, adjusting the project threshold based on the impact coefficient to generate the project reference threshold includes: obtaining the project threshold and impact coefficient of each validation item, and using the product of the impact coefficient and the project threshold as the project reference threshold of the validation item; the reference threshold is the threshold corresponding to each validation item after correction according to the personnel dress requirements in the laboratory reservation project information.

[0031] Because different laboratories or different experiments require different safety protection measures for personnel, such as wearing masks, gloves, protective clothing, and dustproof clothing, these requirements can affect the verification of personnel's identity upon entering the laboratory. For example, wearing masks can affect the accuracy of facial and voice recognition, while wearing gloves can affect handwriting recognition. Therefore, this embodiment modifies the standard thresholds as required by the author to ensure that the thresholds for each item match the impact of the experiment on personnel's attire, thereby reducing the occurrence of misidentifications caused by the impact of attire on recognition accuracy.

[0032] In one possible implementation, one way to construct the influence coefficient table includes: obtaining the matching degree of several items corresponding to each verification item under the dress code in the dress requirement information; obtaining the standard matching degree of each verification item under the standard dress requirements; the standard matching degree is the matching degree of each verification item obtained after the person being identified collects and identifies the verification information without any interference, and multiple people are identified multiple times, and the average of the matching degrees is finally used as the standard matching degree of the verification item; such as the matching degree of items without wearing accessories such as jewelry, masks, gloves and protective clothing; Calculate the average matching degree of several items under the clothing label for the verification item, and record the ratio of the average value to the standard matching degree of the verification item as the influence coefficient of the verification item under the clothing label; obtain the influence coefficient of each verification item under the clothing label in turn; it can be understood that the standard matching degree is the maximum value of the matching degree of the corresponding verification item, and wearing accessories, protective clothing, etc. will cause the matching degree of some verification items to decrease; therefore, the value of the influence coefficient should be in the range of 0 to 1; obtain the influence coefficient of each verification item under each clothing label in turn; integrate several clothing labels, verification items and corresponding influence coefficients into an influence coefficient table.

[0033] In one possible implementation, the pilot validation results and the impact coefficients for each validation item are generated based on the validation information corresponding to the pilot validation item, including: Extract the recognition image from the verification information corresponding to the pilot verification project, input the recognition image into the clothing recognition model to obtain the clothing label. The clothing label is the clothing type corresponding to the relevant person obtained by recognizing the image, including labels corresponding to the body, hands, face and mouth; for example, if the hand is recognized as gloves, the corresponding clothing label is "wearing gloves". If the dress code does not match the dress code in the dress requirement information, the preliminary verification result is recorded as dress abnormal; otherwise, the preliminary verification result is recorded as dress normal. When the preliminary verification result is dress abnormal, it means that the relevant personnel did not dress in accordance with the dress requirements of the experiment, and the relevant personnel should be reminded to change their clothing. Obtain the clothing label, and based on the clothing label, query the influence coefficient table for the influence coefficient of each verification item under the clothing label.

[0034] In one possible implementation, a training method for the clothing recognition model includes: acquiring several recognition images and clothing tags corresponding to the recognition images, wherein the clothing tags are corresponding tags set by professionals for the clothing of the people in the recognition images, including ordinary clothing, ordinary protective clothing, heat-insulating protective clothing, and cleanroom protective clothing, etc.; and integrating the several recognition images and their corresponding clothing tags into several training data and test data. The artificial intelligence model is trained using training data and tested using testing data. The final result is a clothing recognition model with the input being the recognition image and the output being the clothing label corresponding to the recognition image. The artificial intelligence model includes recurrent neural network models, etc. The model training method used for image recognition is a relatively existing technology, which will not be elaborated on here.

[0035] In one possible implementation, generating the project matching degree based on the verification information corresponding to each verification item includes: The verification item is extracted from the facial data in the verification information corresponding to the facial item; the facial data includes facial image data taken during the verification of the corresponding person; the reference facial data in the appointment information is extracted from the appointment information; the facial matching degree corresponding to the facial item is generated based on the reference facial data and the facial data; The audio data corresponding to the verification item is extracted from the verification information; the audio data includes audio data recorded by the corresponding personnel during verification; reference audio data is extracted from the appointment information of the appointment personnel; and the audio matching degree corresponding to the sound item is generated based on the reference audio data and the audio data. For example, the person making the appointment can record a poem into the system in advance. When authenticating, the person makes the appointment and recites the poem into the microphone. The audio is then compared with the audio recorded in the system. If the similarity reaches a pre-set threshold of 98%, the authentication is successful.

[0036] The handwriting data corresponding to the verification item is extracted and verified; the handwriting data includes the signature data of the corresponding personnel during verification; reference handwriting data is extracted from the appointment information of the appointment personnel; and the handwriting matching degree corresponding to the handwriting item is generated based on the reference handwriting data and the handwriting data. Project matching accuracy includes facial matching accuracy, audio matching accuracy, and handwriting matching accuracy.

[0037] In one possible implementation, generating the face matching score based on reference face data and face data includes: Extract the reference face image from the reference face data and the face image to be identified from the face data; align the face image to be identified with the reference face image; input the face image to be identified into the occlusion recognition model to obtain the occlusion recognition image, which is formed by marking the occluded areas in the face image; rasterize the occlusion recognition image to obtain several grids; obtain the area of ​​the occluded area in each grid; substitute the area into the keypoint number adjustment function to obtain the number of keypoints K corresponding to the grid; one expression of the keypoint number adjustment function is:

[0038] Wherein, K is the number of key detection points in the grid, ZM is the area of ​​the occluded region in the grid, SM is the region of the grid, and YS is the initial number of detection points set in the grid; Obtain the minimum Euclidean distance between each pixel of the grid and the occluded area; take the pixel with the largest minimum Euclidean distance as the core pixel of the grid; uniformly set K detection key points in the grid with the core pixel as the center; set the detection key points of each grid in sequence; Extract features from the reference face image at each detection key point and integrate them into a reference feature vector in a set order; extract features from the face image to be identified at each detection key point and integrate them into a recognition feature vector in a set order. Calculate the cosine similarity between the identified feature vector and the reference feature vector, and use the cosine similarity as the face matching degree.

[0039] In this embodiment, key detection points are set in different grids in the manner described above. When the occlusion area in the grid is larger, the error caused by that area to the final recognition accuracy is larger. Therefore, fewer key detection points are set in that area, which effectively avoids the influence of the occluded area on the face recognition result. At the same time, by concentrating the key detection points in the unoccluded parts, the final recognition accuracy is further increased.

[0040] One training method for the occlusion recognition model includes: acquiring several face images, manually marking the occluded areas in the face images to obtain occlusion recognition images; integrating the several face images and their corresponding occlusion recognition images into several training data and verification data; using the training data to train the artificial intelligence model, and using the verification data to verify the trained artificial intelligence model, finally obtaining an occlusion recognition model whose input is a face image and whose output is an occlusion recognition image; the model training method for image recognition is a relatively existing technology, which will not be elaborated on here.

[0041] In one possible implementation, generating an audio matching degree based on reference audio data and audio data includes: extracting reference audio from the reference audio data and the audio to be identified from the audio data; calculating the cosine similarity between the audio to be identified and the reference audio, and using the cosine similarity as the audio matching degree.

[0042] In one possible implementation, generating the handwriting matching degree based on reference handwriting data and handwriting data includes: Extract handwriting images from reference handwriting data; the handwriting images are images of handwriting completed by relevant personnel, including electronic screen images or photographs of paper signatures; The handwriting image is converted to a grayscale image; the pixel values ​​of each pixel in the grayscale image are extracted, the number of pixels with the same pixel value is calculated, and a pressure feature distribution vector is generated based on the number of pixel values ​​of each pixel; specifically, in this embodiment, the pressure feature distribution vector is expressed as follows: using the formula

[0043] The feature value TZi of each pixel is calculated; where Ii is the pixel value of the pixel corresponding to i; k is the pixel value corresponding to the signature background; when the signature background is pure white, k=255, and when the signature background is pure black, k=0; it can be understood that k is generally the pixel value with the most pixels in the grayscale image; the feature values ​​other than the feature value corresponding to the pixel value of k are integrated into a pressure feature distribution vector in a set order; it can be understood that in general electronic signatures or paper signatures, the handwriting will have variations in darkness, and the variations in darkness reflect the pressure at different positions of the handwriting; Obtain the reference pressure distribution feature vector from the reference handwriting data, calculate the cosine similarity between the pressure distribution feature vector and the reference pressure distribution feature vector, and use the cosine similarity as the handwriting matching degree.

[0044] Wearing gloves creates a certain thickness between the hand and the pen, which can distort the handwriting. Simply identifying handwriting from images is not very accurate. Gloves only reduce or increase the pressure during writing, and have little impact on the distribution of pressure in the handwriting. By generating a pressure distribution feature vector for the handwriting, the accuracy of handwriting recognition is further improved.

[0045] Please see Figure 2 In conjunction with the first aspect above, in one possible implementation, generating the final authentication result based on the project matching degree, influence coefficient, and project reference threshold of each verification project includes: S1: Extract the face matching score, audio matching score, and handwriting matching score from the item matching score; and the item reference threshold corresponding to each verification item in the reference threshold; S2: Based on face matching degree, audio matching degree, and handwriting matching degree; and the reference thresholds corresponding to each verification item in the reference thresholds, generate the authentication result for each verification item; specifically, when the face matching degree is greater than the reference threshold corresponding to the face item, mark the authentication result corresponding to the face item as passed; otherwise, mark the authentication result corresponding to the face item as failed; when the audio matching degree is greater than the reference threshold corresponding to the audio item, mark the authentication result corresponding to the audio item as passed; otherwise, mark the authentication result corresponding to the audio item as failed; when the handwriting matching degree is greater than the reference threshold corresponding to the handwriting item, mark the authentication result corresponding to the handwriting item as passed; otherwise, mark the authentication result corresponding to the handwriting item as failed. S3: If the certification results for each verification item are all passed, proceed to S6; otherwise, proceed to S4. S4: Obtain the impact coefficient of each validation item; process the impact coefficient of each validation item using a weighting generation function to obtain the corresponding weight coefficient for each validation item; perform a weighted summation of the item matching degrees of each validation item based on each weight coefficient to obtain the comprehensive matching degree; perform a weighted summation of the item reference thresholds of each validation item based on each weight coefficient to obtain the comprehensive reference threshold; one expression of the weighting generation function is:

[0046] Where QXi is the weight coefficient corresponding to the verification item numbered i; YXi is the influence coefficient corresponding to the verification item numbered i; when i=1, YX1 is the influence coefficient corresponding to the face item; when i=2, YX2 is the influence coefficient corresponding to the voice item; when i=3, YX3 is the influence coefficient corresponding to the handwriting item; it can be understood that the verification items can include other items, not limited to face, voice and handwriting. This embodiment calculates the weight coefficient of each reference item using the above formula. When a certain item is greatly affected by clothing, the weight value of the final recognition result of that item is reduced to reduce the impact of that item on the final recognition result. At the same time, in conjunction with an adaptive threshold, the accuracy of the final authentication result is effectively improved. S5: Determine whether the overall matching degree is greater than the overall reference threshold. If yes, proceed to S6; otherwise, proceed to S7. S6: Obtain the current time and the reservation time period in the reservation information; if the current time is within the reservation time period, set the final authentication result to pass; otherwise, proceed to S7; S7: Set the final authentication result to failed.

[0047] Secondly, please refer to Figure 3 A laboratory safety certification device based on a Lims system is provided, comprising: a data acquisition unit, a data processing unit, and a display unit; Data acquisition unit: used to acquire the laboratory's reservation information, as well as the laboratory's pilot validation project and validation information for several validation projects; Data processing unit: used to generate reference thresholds based on the reservation item information in the reservation information; the reference thresholds include item reference thresholds for each verification item; and to generate preliminary verification results and influence coefficients for each verification item based on the verification information corresponding to the preliminary verification item; to generate item matching degree based on the verification information corresponding to each verification item; and to generate the final authentication result based on the item matching degree, influence coefficient, and item reference threshold of each verification item. Display unit: Used to display the final authentication result and the preliminary verification result.

[0048] Thirdly, this application provides a LIMS-based laboratory safety authentication device, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is configured to execute the instructions to implement the methods described in the first aspect and any possible implementation thereof. This LIMS-based laboratory safety authentication device can be an electronic device or a chip within an electronic device.

[0049] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a LIMS-based laboratory safety certification device, cause the LIMS-based laboratory safety certification device to perform the methods described in the first aspect and any possible implementation thereof.

[0050] Fifthly, this application provides a computer program product containing instructions that, when run on a LIMS-based laboratory safety certification device, causes the LIMS-based laboratory safety certification device to perform the methods described in the first aspect and any possible implementation thereof. Some data in the above formulas are obtained by removing dimensions and calculating numerical values; the formulas are derived from a large amount of collected data through software simulation to obtain a formula that most closely approximates the real situation; the preset parameters and preset thresholds in the formulas are set by those skilled in the art based on actual conditions or obtained through simulation with a large amount of data.

[0051] How this application works: By acquiring laboratory reservation information, reference thresholds are generated based on the reservation project information. The laboratory's pilot validation projects and corresponding validation information for several validation projects are acquired. Pilot validation results and impact coefficients for each validation project are generated based on the validation information for the pilot validation projects. Project matching degrees are generated based on the validation information for each validation project. The final certification result is generated based on the project matching degree, impact coefficient, and reference threshold for each validation project. The required certification thresholds are adaptively set according to the specific circumstances of relevant personnel. Multiple projects are used for multiple certifications, and the adaptive certification analysis scheme increases the accuracy and efficiency of security certification.

[0052] The above embodiments are only used to illustrate the technical methods of this application and are not intended to limit it. Although this application has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of this application without departing from the spirit and scope of the technical methods of this application.

Claims

1. A laboratory safety certification method based on the LIMS system, characterized in that, include: Obtain laboratory appointment information; A reference threshold is generated based on the reservation item information in the reservation information; The reference threshold includes the project reference threshold for each verification item corresponding to the security certification; including: extracting the dress requirement information from the appointment project information; querying the influence coefficient corresponding to each verification item under the second dress tag in the influence coefficient table based on the second dress tag in the dress requirement information; obtaining the project threshold for each verification item; adjusting the corresponding project threshold based on the influence coefficient of the verification item to generate the project reference threshold; and integrating each verification item and its corresponding project reference threshold into a reference threshold; The process involves: acquiring pilot validation projects and validation information corresponding to several validation projects in the laboratory; generating pilot validation results and target influence coefficients for each validation project based on the validation information corresponding to the pilot validation projects; including: extracting recognition images from the validation information corresponding to the pilot validation projects, inputting the recognition images into a clothing recognition model to obtain a first clothing label; when the first clothing label is different from the second clothing label in the clothing requirement information, the pilot validation result is recorded as clothing abnormal; otherwise, the pilot validation result is recorded as clothing normal; acquiring a second clothing label, querying the influence coefficient table based on the second clothing label for the influence coefficients corresponding to each validation project under the second clothing label, and recording them as target influence coefficients; the target influence coefficients represent the degree of influence of the second clothing label on the accuracy of validation project recognition; Generate project matching degree based on the verification information corresponding to each verification project; The final certification result is generated based on the project matching degree, target impact coefficient, and project reference threshold of each verification project; including: S1: Extract the face matching score, audio matching score, and handwriting matching score from the item matching score; and the item reference threshold corresponding to each verification item in the reference threshold; S2: Based on face matching, audio matching, and handwriting matching, and the reference thresholds for each verification item in the reference thresholds, the authentication results for each verification item are generated; S3: If the certification results for each verification item are all passed, proceed to S6; otherwise, proceed to S4. S4: Obtain the target impact coefficient corresponding to each verification project; process the impact coefficient of each verification project using the weight generation function to obtain the weight coefficient corresponding to each verification project; perform a weighted summation of the project matching degree of each verification project based on each weight coefficient to obtain the comprehensive matching degree; perform a weighted summation of the project reference threshold of each verification project based on each weight coefficient to obtain the comprehensive reference threshold; S5: Determine whether the overall matching degree is greater than the overall reference threshold. If yes, proceed to S6; otherwise, proceed to S7. S6: Obtain the current time and the reservation time slot in the reservation information; if the current time is within the reservation time slot, set the final authentication result to pass; otherwise, proceed to S7; S7: Set the final authentication result to failed.

2. The laboratory safety certification method based on the LIMS system according to claim 1, characterized in that, One training method for the clothing recognition model includes: Acquire several recognition images and corresponding clothing tags for the recognition images; integrate the several recognition images and their corresponding clothing tags into several training data and test data; The artificial intelligence model is trained using training data and tested using validation data. The final result is a clothing recognition model with the recognition image as input and the clothing label corresponding to the recognition image as output.

3. The laboratory safety certification method based on the LIMS system according to claim 1, characterized in that, One method for constructing the influence coefficient table includes: Obtain the matching degree of several items corresponding to each verification item under the second dress tag in the dress requirement information; obtain the standard item matching degree of each verification item corresponding to the standard dress requirement; the standard dress requirement is not wearing any accessories. Calculate the average matching degree of the verification item under the second dress label for several items, and record the ratio of the average matching degree of the verification item to the standard matching degree of the verification item as the influence coefficient of the verification item under the second dress label; Obtain the influence coefficients corresponding to each verification item under each second dress label in sequence; integrate several second dress labels, verification items and corresponding influence coefficients into an influence coefficient table.

4. The laboratory safety certification method based on the LIMS system according to claim 1, characterized in that, The process of generating a project matching degree based on the verification information corresponding to each verification item includes: The verification items are extracted from the facial data in the verification information corresponding to the facial items; the reference facial data in the appointment information is extracted from the appointment information; and the facial matching degree corresponding to the facial items is generated based on the reference facial data and the facial data. Extract audio data from the verification information corresponding to the sound item; extract reference audio data from the appointment information of the appointment personnel; generate the audio matching degree corresponding to the sound item based on the reference audio data and the audio data. Extract handwriting data from the verification information corresponding to the handwriting item; extract reference handwriting data from the appointment information of the appointment person; generate the handwriting matching degree corresponding to the handwriting item based on the reference handwriting data and the handwriting data. The project matching degree includes face matching degree, audio matching degree and handwriting matching degree.

5. The laboratory safety certification method based on the LIMS system according to claim 4, characterized in that, The face matching score is generated based on the reference face data and the face data, including: Extract the reference face image from the reference face data and the face image to be identified from the face data, align the face image to be identified with the reference face image, and input the face image to be identified into the occlusion recognition model to obtain the occlusion recognition image; and set several key detection points in the occlusion recognition image. Extract features from the reference face image at each detection key point and integrate them into a reference feature vector in a set order; extract features from the face image to be identified at each detection key point and integrate them into a recognition feature vector in a set order. Calculate the cosine similarity between the identified feature vector and the reference feature vector, and use the cosine similarity as the face matching degree.

6. The laboratory safety certification method based on the LIMS system according to claim 5, characterized in that, One method for setting the key detection points includes: The occlusion recognition image is rasterized to obtain several grids; the area of ​​the occlusion region in each grid is obtained; the area is substituted into the key point number adjustment function to obtain the number of key points K corresponding to the grid; Obtain the minimum Euclidean distance between each pixel of the grid and the occluded area; take the pixel with the largest minimum Euclidean distance as the core pixel of the grid; uniformly set K detection key points in the grid with the core pixel as the center; set the detection key points of each grid in sequence.

7. A laboratory safety certification device based on a Lims system, comprising the application of the laboratory safety certification method based on a Lims system as described in any one of claims 1 to 6; characterized in that, include: Data acquisition unit, data processing unit, and display unit; Data acquisition unit: used to acquire the laboratory's reservation information, as well as the laboratory's pilot validation project and validation information for several validation projects; Data processing unit: used to generate reference thresholds based on the reservation item information in the reservation information; The reference threshold includes the project reference threshold for each verification item corresponding to the security certification; including: extracting the dress requirement information from the appointment project information; querying the influence coefficient corresponding to each verification item under the second dress tag in the influence coefficient table based on the second dress tag in the dress requirement information; obtaining the project threshold for each verification item; adjusting the corresponding project threshold based on the influence coefficient of the verification item to generate the project reference threshold; and integrating each verification item and its corresponding project reference threshold into a reference threshold; Furthermore, based on the verification information corresponding to the pilot verification items, a pilot verification result and a target influence coefficient corresponding to each verification item are generated; including: extracting the recognition image from the verification information corresponding to the pilot verification item, inputting the recognition image into the clothing recognition model to obtain a first clothing label; when the first clothing label is different from the second clothing label in the clothing requirement information, the pilot verification result is recorded as clothing abnormal; otherwise, the pilot verification result is recorded as clothing normal; obtaining the second clothing label, querying the influence coefficient table based on the second clothing label for the influence coefficient corresponding to each verification item under the second clothing label, and recording it as the target influence coefficient; the target influence coefficient represents the degree of influence of the second clothing label on the accuracy of the verification item recognition; The project matching degree is generated based on the verification information corresponding to each verification project; the final certification result is generated based on the project matching degree, target impact coefficient and project reference threshold of each verification project. Display unit: Used to display the final authentication result and the preliminary verification result, including: S1: Extract the face matching score, audio matching score, and handwriting matching score from the item matching score; and the item reference threshold corresponding to each verification item in the reference threshold; S2: Based on face matching, audio matching, and handwriting matching, and the reference thresholds for each verification item in the reference thresholds, the authentication results for each verification item are generated; S3: If the certification results for each verification item are all passed, proceed to S6; otherwise, proceed to S4. S4: Obtain the target impact coefficient of each verification project; process the target impact coefficient of each verification project using the weight generation function to obtain the weight coefficient corresponding to each verification project; perform a weighted summation of the project matching degree of each verification project based on each weight coefficient to obtain the comprehensive matching degree; perform a weighted summation of the project reference threshold of each verification project based on each weight coefficient to obtain the comprehensive reference threshold; S5: Determine whether the overall matching degree is greater than the overall reference threshold. If yes, proceed to S6; otherwise, proceed to S7. S6: Obtain the current time and the reservation time slot in the reservation information; if the current time is within the reservation time slot, set the final authentication result to pass; otherwise, proceed to S7; S7: Set the final authentication result to failed.

Images