Artificial Intelligence-Based Hazardous Substance Identification Method, Device, Equipment, and Storage Medium

Abstract

This application relates to the field of computer technology, and provides a method, device, equipment and storage medium for identifying dangerous goods based on artificial intelligence. The method includes: for each pixel of the image to be identified, if the target object corresponding to the pixel is a suspicious dangerous good, using the category code of the suspicious dangerous good corresponding to the target object to label the pixel to obtain a target image; for each category code labeled on the target image, obtaining a target dangerous good identification model in a preset database based on the category code, and inputting the target image into the target dangerous good identification model to obtain the dangerous good identification information in the package to be detected; generating a dangerous good identification result in the package to be detected based on all the dangerous good identification information. This method can improve the accuracy of the dangerous good identification result.

Description

Technical Field

[0001] This application relates to the field of computer technology, and particularly to a method, device, equipment and storage medium for identifying dangerous goods based on artificial intelligence. Background Art

[0002] When the existing X-ray security inspection machine identifies dangerous goods in the items in a package, it usually scans the package through the X-ray security inspection machine to obtain a scanned image, and inputs the scanned image into a preset dangerous goods identification model to identify whether the items in the package contain dangerous goods. This method still has the problem of low identification accuracy. Summary of the Invention

[0003] This application provides a method, device, equipment and storage medium for identifying dangerous goods based on artificial intelligence to solve the problems raised in the above background art.

[0004] In a first aspect, this application provides a method for identifying dangerous goods based on artificial intelligence, including:

[0005] Obtaining an image to be identified; the X-ray imager of the security inspection system scans the package to be detected to obtain the image to be identified;

[0006] Respectively extracting the chromaticity values of each pixel of the image to be identified based on a preset chromaticity extraction model;

[0007] For each of the pixels, judging whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel;

[0008] For each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, labeling the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image;

[0009] For each of the category codes labeled on the target image, obtaining a target dangerous goods identification model in a preset database based on the category code, and inputting the target image into the target dangerous goods identification model to obtain the dangerous goods identification information in the package to be detected;

[0010] Generating a dangerous goods identification result in the package to be detected based on all the dangerous goods identification information.

[0011] In a second aspect, this application provides a device for identifying dangerous goods based on artificial intelligence, including:

[0012] A first acquisition module, configured to obtain an image to be identified; the X-ray imager of the security inspection system scans the package to be detected to obtain the image to be identified;

[0013] An extraction module, configured to extract the chromaticity values of each pixel of the image to be recognized respectively based on a preset chromaticity extraction model;

[0014] A judgment module, configured to judge whether the target object is a suspicious dangerous good for each of the pixels based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel;

[0015] A labeling module, configured to, for each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image;

[0016] A second acquisition module, configured to, for each of the category codes labeled on the target image, acquire a target dangerous good recognition model in a preset database based on the category code, and input the target image into the target dangerous good recognition model to obtain the dangerous good recognition information in the package to be detected;

[0017] A generation module, configured to generate a dangerous good recognition result in the package to be detected based on all the dangerous good recognition information.

[0018] In a third aspect, the present application provides an X-ray security inspection machine device, where the X-ray security inspection machine device includes a processor, a memory, and a computer program stored on the memory and executable by the processor, and when the computer program is executed by the processor, the above-mentioned artificial intelligence-based dangerous good recognition method is implemented.

[0019] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned artificial intelligence-based dangerous good recognition method is implemented.

[0020] This embodiment provides a method, apparatus, device, and storage medium for identifying dangerous goods based on artificial intelligence. The method includes obtaining an image to be identified; scanning a package to be detected by an X-ray imager of a security inspection system to obtain the image to be identified; respectively extracting the chromaticity values of each pixel of the image to be identified based on a preset chromaticity extraction model; for each pixel, determining whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel, where the target object is the object corresponding to the pixel; for each pixel, if the target object corresponding to the pixel is a suspicious dangerous good, labeling the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image; for each category code labeled on the target image, obtaining a target dangerous good identification model based on the category code in a preset database, and inputting the target image into the target dangerous good identification model to obtain the dangerous good identification information in the package to be detected; generating the dangerous good identification result in the package to be detected based on all the dangerous good identification information. This method improves the accuracy of dangerous good identification. Description of the Drawings

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0022] Figure 1 It is a schematic flowchart of the method for identifying dangerous goods based on artificial intelligence provided by the embodiment of the present application;

[0023] Figure 2 It is a schematic block diagram of the structure of the device for identifying dangerous goods based on artificial intelligence provided by the embodiment of the present application;

[0024] Figure 3 It is a schematic block diagram of the structure of the security inspection machine device provided by the embodiment of the present application. Detailed Embodiments

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, rather than all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the protection scope of the present invention.

[0026] The flowcharts shown in the accompanying drawings are merely illustrative examples, and do not necessarily include all content and operations / steps, nor do they necessarily need to be executed in the order described. For example, some operations / steps can be decomposed, combined, or partially merged, so the actual execution order may change according to the actual situation.

[0027] It should also be understood that the terms used in the specification of this application are merely for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification of this application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0028] It should be further understood that the term "and / or" used in the specification of this application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0029] When an existing X-ray security inspection machine identifies dangerous goods in the items in a package, it usually scans the package with the X-ray security inspection machine to obtain a scanned image, and inputs the scanned image into a preset dangerous goods identification model to identify whether the items in the package contain dangerous goods. This method still has the problem of low identification accuracy. For this reason, this application provides a method, device, equipment, and storage medium for identifying dangerous goods based on artificial intelligence to solve the above problems.

[0030] The following will describe in detail some embodiments of this application with reference to the accompanying drawings. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0031] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of the method for identifying dangerous goods based on artificial intelligence provided by the embodiment of this application. As Figure 1 shown, the method for identifying dangerous goods based on artificial intelligence provided by the embodiment of this application includes steps S100 to S600.

[0032] Step S100: Obtain an image to be identified; the X-ray imager of the security inspection system scans the package to be detected to obtain the image to be identified.

[0033] Step S200: Extract the chromaticity values of each pixel of the image to be identified based on a preset chromaticity extraction model.

[0034] Step S300: For each of the pixels, determine whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel.

[0035] Step S400: For each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image.

[0036] Step S500: For each of the category codes labeled on the target image, obtain a target dangerous good recognition model from a preset database based on the category code, and input the target image into the target dangerous good recognition model to obtain dangerous good recognition information inside the package to be detected.

[0037] Step S600: Generate a dangerous good recognition result inside the package to be detected based on all the dangerous good recognition information.

[0038] In this embodiment, for obtaining the image to be recognized as described in step S100 above, specifically, when the sensing device of the security inspection system senses the package to be detected, control the X-ray imager to scan the package to be detected to obtain the image to be recognized, and obtain the image to be recognized through the X-ray imager.

[0039] For respectively extracting the chromaticity values of each pixel of the image to be recognized as described in step S200 above, specifically, the chromaticity extraction model includes a pixel segmentation module and a chromaticity extraction module. The pixel segmentation module performs pixel segmentation on the image to be recognized to obtain a pixel segmentation network corresponding to the image to be recognized, and the chromaticity extraction module respectively extracts the chromaticity values of each pixel of the pixel segmentation network.

[0040] For each of the pixels as described in step S300 above, determine whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel. Specifically, for each of the pixels, traverse a preset chromaticity value-category code mapping table to determine whether there is a chromaticity value corresponding to the pixel in the chromaticity value-category code mapping table. If there is a chromaticity value corresponding to the pixel in the chromaticity value-category code mapping table, the target object is a suspicious dangerous good; if there is no chromaticity value corresponding to the pixel in the chromaticity value-category code mapping table, the target object is not a suspicious dangerous good. Among them, the target object is the object corresponding to the pixel.

[0041] For each of the pixels as described in step S400 above, if the target object corresponding to the pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image. Specifically, for each of the pixels, first, determine the category code corresponding to the chromaticity value in the chromaticity value-category code mapping table using the chromaticity value of the pixel, and then label the category code on the pixel.

[0042] For each category code labeled on the target image as described in step S500 above, obtain a target dangerous goods recognition model from a preset database based on the category code, and input the target image into the target dangerous goods recognition model to obtain the dangerous goods recognition information in the package to be detected. Specifically, for each category code labeled on the target image, first, generate a decryption password based on the category code, then, perform decryption processing on each of the dangerous goods recognition models in the preset database based on the decryption password, and use the successfully decrypted dangerous goods recognition model as the target dangerous goods recognition model. Finally, input the target image into the target dangerous goods recognition model to obtain the dangerous goods recognition information in the package to be detected. It can be understood that when there are multiple dangerous goods in the package to be detected and the category codes corresponding to each dangerous good are different from each other, multiple category codes will be labeled on the target image. It can also be understood that one category code corresponds to at least one suspicious dangerous good, and the chromaticity values of the scanned images obtained after the X-ray imager scans each of the suspicious dangerous goods corresponding to one category code are the same, and one category code corresponds to one target dangerous goods recognition model.

[0043] Generate the dangerous goods recognition result in the package to be detected based on all the dangerous goods recognition information as described in step S600 above. Specifically, analyze and sort out all the dangerous goods recognition information to obtain the dangerous goods recognition result.

[0044] The method provided in this embodiment determines whether the target object corresponding to each pixel of the image to be recognized is a suspicious dangerous good through the chromaticity values corresponding to each pixel of the image to be recognized, and when the target object corresponding to any pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the pixel to obtain a target image. Furthermore, for each category code labeled on the target image, obtain a target dangerous goods recognition model from a preset database based on the category code, and input the target image into the target dangerous goods recognition model to obtain the dangerous goods recognition information in the package to be detected, improving the accuracy of the dangerous goods recognition result.

[0045] In some embodiments, the chromaticity extraction model includes a pixel segmentation module and a chromaticity extraction module. The steps of respectively extracting the chromaticity values of each pixel of the image to be recognized based on a preset chromaticity extraction model are as follows:

[0046] Perform pixel segmentation on the image to be recognized based on the pixel segmentation module to obtain a pixel segmentation network corresponding to the image to be recognized;

[0047] Respectively extract the chromaticity values of each pixel of the pixel segmentation network based on the chromaticity extraction module.

[0048] In this embodiment, first, the pixel segmentation module performs pixel segmentation on the image to be recognized to obtain the pixel segmentation network corresponding to the image to be recognized. Specifically, the pixel segmentation module segments each pixel of the image to be recognized to obtain the pixel segmentation network corresponding to the image to be recognized.

[0049] Then, the chromaticity extraction module extracts the chromaticity values of each pixel of the pixel segmentation network respectively. Specifically, the chromaticity extraction module traverses each pixel of the pixel segmentation network to extract the chromaticity values of each pixel of the pixel segmentation network.

[0050] The method provided in this embodiment helps to further improve the accuracy of the dangerous goods recognition result.

[0051] In some embodiments, the preset database includes multiple dangerous goods recognition models, and each of the dangerous goods recognition models is provided with an encryption password. Obtaining the target dangerous goods recognition model from the preset database based on the category code includes the following steps:

[0052] Generate a decryption password based on the category code;

[0053] Perform decryption processing on each of the dangerous goods recognition models based on the decryption password, and use the successfully decrypted dangerous goods recognition model as the target dangerous goods recognition model.

[0054] In this embodiment, first, generate a decryption password based on the category code. Specifically, first, generate a target table based on the category code, then select target characters in the target table, and finally, arrange the target characters in sequence based on the positions of the target characters in the target table to obtain the decryption password.

[0055] Then, perform decryption processing on each of the dangerous goods recognition models based on the decryption password, and use the successfully decrypted dangerous goods recognition model as the target dangerous goods recognition model. Specifically, use the decryption password to perform decryption processing on each of the dangerous goods recognition models one by one, and use the successfully decrypted dangerous goods recognition model as the target dangerous goods recognition model.

[0056] The method provided in this embodiment ensures the security of each of the dangerous goods recognition models by setting an encryption password for each of the dangerous goods recognition models, and can prevent unauthorized personnel from tampering with the dangerous goods recognition models, thereby affecting the dangerous goods recognition result.

[0057] In some embodiments, generating the decryption password based on the category code includes the following steps:

[0058] Calculate the hash values corresponding to each encoded character of the category code respectively based on a preset hash algorithm; wherein, the number of characters of each of the hash values is the same, and the hash algorithm is any one of MD5, SHA-1, and SHA-256;

[0059] Determine a target hash value among all the hash values; wherein, the target hash value has no characters in common with the category code;

[0060] For each of the target hash values, calculate the variance between each of the numbers in the target hash value;

[0061] Construct a blank table; wherein, the number of rows of the blank table is equal to the number of target hash values, and the number of columns of the blank table is equal to the number of characters of the target hash value;

[0062] Arrange each of the target hash values in the blank table in sequence from top to bottom to obtain a target table; wherein, the variances corresponding to each of the target hash values in the target table increase in sequence from top to bottom;

[0063] For each character of the target table, calculate the target number corresponding to the character; wherein, the target number is the sum of the row number and the column number corresponding to the character in the target table;

[0064] Determine a target character among all the characters of the target table; wherein, the target number corresponding to the target character is a prime number;

[0065] Arrange each of the target characters in sequence based on the positions of the target characters in the target table to obtain the decryption password.

[0066] The method provided in this embodiment, by calculating the hash values corresponding to each encoded character of the category code respectively based on a preset hash algorithm, generating a target table based on the category code and each of the hash values, and generating the decryption password based on the target table, improves the cracking difficulty of the decryption password and further improves the security of each of the dangerous goods identification models.

[0067] In some embodiments, the target dangerous goods identification model includes an input layer, a target pixel extraction layer, a smoothing processing layer, a feature extraction layer, a pooling layer, a fully connected layer, and an output layer. The step of inputting the target image into the target dangerous goods identification model to obtain the dangerous goods identification information in the package to be detected includes the following steps:

[0068] Input the target image through the input layer;

[0069] Extract the target area corresponding to the target image through the target pixel extraction layer; wherein, each pixel in the target area is labeled with the category code corresponding to the target dangerous goods recognition model.

[0070] Smooth the contour of the target area through the smoothing processing layer to obtain the contour area corresponding to the target image.

[0071] Extract features from the contour area through the feature extraction layer to obtain the feature information corresponding to the contour area.

[0072] Reduce the dimension of the feature information through the pooling layer to obtain the target feature information.

[0073] Perform deep learning on the target feature information through the fully connected layer to obtain the dangerous goods recognition information corresponding to the target dangerous goods recognition model.

[0074] Output the dangerous goods recognition information through the output layer.

[0075] It can be understood that the target dangerous goods recognition models corresponding to each category code labeled on the target image all execute the steps in this embodiment.

[0076] The method provided in this embodiment extracts the target area corresponding to the target image through the target pixel extraction layer, and smooths the contour of the target area through the smoothing processing layer to obtain the contour area corresponding to the target image, which helps to further improve the accuracy of the dangerous goods recognition result.

[0077] Please refer to Figure 2 , Figure 2 which is a schematic block diagram of the structure of the dangerous goods recognition device 100 based on artificial intelligence provided in the embodiment of the present application. As Figure 2 shown, the dangerous goods recognition device 100 based on artificial intelligence includes:

[0078] A first acquisition module 110, configured to acquire an image to be recognized; the X-ray imager of the security inspection system scans the package to be detected to obtain the image to be recognized.

[0079] An extraction module 120, configured to extract the chromaticity values of each pixel of the image to be recognized based on a preset chromaticity extraction model.

[0080] A judgment module 130, configured to judge whether the target object is a suspicious dangerous good for each of the pixels based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel.

[0081] The annotation module 140 is configured to, for each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, annotate the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image.

[0082] The second acquisition module 150 is configured to, for each of the category codes annotated on the target image, obtain a target dangerous good recognition model from a preset database based on the category code, and input the target image into the target dangerous good recognition model to obtain the dangerous good recognition information in the package to be detected.

[0083] The generation module 160 is configured to generate a dangerous good recognition result in the package to be detected based on all the dangerous good recognition information.

[0084] It should be noted that those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working processes of the above-described device and each module can refer to the corresponding processes in the foregoing embodiments of the method for identifying dangerous goods based on artificial intelligence, and will not be elaborated herein.

[0085] The dangerous good recognition device 100 based on artificial intelligence provided in the above embodiment can be implemented in the form of a computer program, and the computer program can run on a security inspection machine device 200 as shown in Figure 3 shown.

[0086] Please refer to Figure 3 , Figure 3 which is a schematic block diagram of the structure of the security inspection machine device 200 provided in the embodiment of the present application. The security inspection machine device 200 includes a processor 201 and a memory 202. The processor 201 and the memory 202 are connected through a system bus 203. Among them, the memory 202 may include a non-volatile storage medium and an internal memory.

[0087] The non-volatile storage medium can store a computer program. The computer program includes program instructions, and when the program instructions are executed by the processor 201, the processor 201 can be made to execute any of the above methods for identifying dangerous goods based on artificial intelligence.

[0088] The processor 201 is configured to provide computing and control capabilities to support the operation of the entire security inspection machine device 200.

[0089] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 201, the processor 201 can be made to execute any of the above methods for identifying dangerous goods based on artificial intelligence.

[0090] Those skilled in the art can understand that Figure 3The structure shown is only a block diagram of some of the structures related to the solution of this application, and does not constitute a limitation on the security inspection machine device 200 involved in the solution of this application. Specifically, the security inspection machine device 200 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0091] It should be understood that the processor 201 may be a central processing unit (CPU), and the processor 201 may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0092] Among them, in some embodiments, the processor 201 is used to run a computer program stored in the memory to implement the following steps:

[0093] Obtain an image to be recognized; the X-ray imager of the security inspection system scans the package to be detected to obtain the image to be recognized;

[0094] Extract the chromaticity values of each pixel of the image to be recognized based on a preset chromaticity extraction model;

[0095] For each of the pixels, determine whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel;

[0096] For each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image;

[0097] For each of the category codes labeled on the target image, obtain a target dangerous good recognition model based on the category code in a preset database, and input the target image into the target dangerous good recognition model to obtain the dangerous good recognition information in the package to be detected;

[0098] Generate the dangerous goods identification result in the package to be detected based on all the above-mentioned dangerous goods identification information. It should be noted that those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working process of the security inspection machine device 200 described above can refer to the corresponding process of the aforementioned dangerous goods identification method based on artificial intelligence, which will not be elaborated here.

[0099] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by one or more processors, the one or more processors are caused to implement the dangerous goods identification method based on artificial intelligence provided by the embodiment of the present application.

[0100] Among them, the computer-readable storage medium may be an internal storage unit of the security inspection machine device 200 in the foregoing embodiment, such as the hard disk or memory of the security inspection machine device 200. The computer-readable storage medium may also be an external storage device of the security inspection machine device 200, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped with the security inspection machine device 200.

[0101] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or substitutions within the technical scope disclosed by the present application, and these modifications or substitutions should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for identifying dangerous goods based on artificial intelligence, characterized in that, Including: Obtain the image to be recognized; The X-ray imager of the security inspection system scans the package to be detected to obtain the image to be recognized; Extract the chromaticity values of each pixel of the image to be recognized based on a preset chromaticity extraction model; For each of the pixels, determine whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel; For each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image; For each category code labeled on the target image, obtain a target dangerous good recognition model from a preset database based on the category code, and input the target image into the target dangerous good recognition model to obtain the dangerous good recognition information in the package to be detected; Generate the dangerous good recognition result in the package to be detected based on all the dangerous good recognition information; Wherein, the preset database includes multiple dangerous good recognition models, and each of the dangerous good recognition models is provided with an encryption password. The obtaining of the target dangerous good recognition model from the preset database based on the category code includes: Calculate the hash values corresponding to each coding character of the category code respectively based on a preset hash algorithm; wherein, the number of characters of each of the hash values is the same; Determine a target hash value among all the hash values; wherein, the target hash value does not have the same characters as the category code; For each of the target hash values, calculate the variance between each number in the target hash value; Construct a blank table; wherein, the number of rows of the blank table is equal to the number of the target hash values, and the number of columns of the blank table is equal to the number of characters of the target hash value; Arrange each of the target hash values in the blank table in sequence from top to bottom to obtain a target table; wherein, the variance corresponding to each of the target hash values in the target table increases in sequence from top to bottom; For each character of the target table, calculate the target number corresponding to the character; wherein, the target number is the sum of the row number and the column number corresponding to the character in the target table; Determine a target character among all the characters of the target table; wherein, the target number corresponding to the target character is a prime number; Arrange each of the target characters in sequence based on the positions of the target characters in the target table to obtain a decryption password; Perform decryption processing on each of the dangerous good recognition models based on the decryption password, and use the successfully decrypted dangerous good recognition model as the target dangerous good recognition model.

2. The method for identifying dangerous goods based on artificial intelligence according to claim 1, wherein, Before obtaining the image to be recognized, the method further includes: When the induction device of the security inspection system senses the package to be detected, control the X-ray imager to scan the package to be detected to obtain the image to be recognized.

3. The method for identifying dangerous goods based on artificial intelligence according to claim 1, wherein The chromaticity extraction model includes a pixel segmentation module and a chromaticity extraction module. The extracting of the chromaticity values of each pixel of the image to be recognized based on a preset chromaticity extraction model includes: Performing pixel segmentation on the image to be recognized based on the pixel segmentation module to obtain a pixel segmentation network corresponding to the image to be recognized; Extracting the chromaticity values of each pixel of the pixel segmentation network respectively based on the chromaticity extraction module.

4. The method for identifying dangerous goods based on artificial intelligence according to claim 1, wherein The determining whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel includes: Traversing a preset chromaticity value - category code mapping table to determine whether there is a chromaticity value corresponding to the pixel in the chromaticity value - category code mapping table.

5. The method for identifying dangerous goods based on artificial intelligence according to claim 1, wherein The target dangerous good recognition model includes an input layer, a target pixel extraction layer, a smoothing processing layer, a feature extraction layer, a pooling layer, a fully connected layer, and an output layer. Inputting the target image into the target dangerous good recognition model to obtain the dangerous good recognition information in the package to be detected includes: Inputting the target image through the input layer; Extracting a target area corresponding to the target image through the target pixel extraction layer; wherein, each pixel of the target area is labeled with a category code corresponding to the target dangerous good recognition model; Performing smoothing processing on the contour of the target area through the smoothing processing layer to obtain a contour area corresponding to the target image; Performing feature extraction on the contour area through the feature extraction layer to obtain feature information corresponding to the contour area; Performing dimensionality reduction processing on the feature information through the pooling layer to obtain target feature information; Performing deep learning on the target feature information through the fully connected layer to obtain the dangerous good recognition information corresponding to the target dangerous good recognition model; Outputting the dangerous good recognition information through the output layer.

6. A dangerous goods identification device based on artificial intelligence, characterized in that, Including: A first acquisition module, configured to acquire an image to be recognized; Scanning the package to be detected by an X - ray imager of the security inspection system to obtain the image to be recognized; An extraction module, configured to extract the chromaticity values of each pixel of the image to be recognized respectively based on a preset chromaticity extraction model; A judgment module, configured to, for each of the pixels, judge whether the target object is a suspicious dangerous good based on the chromaticity value corresponding to the pixel; wherein, the target object is the object corresponding to the pixel; A labeling module, configured to, for each of the pixels, if the target object corresponding to the pixel is a suspicious dangerous good, label the pixel with the category code of the suspicious dangerous good corresponding to the target object to obtain a target image; A second acquisition module, configured to, for each category code labeled on the target image, acquire a target dangerous good recognition model in a preset database based on the category code, and input the target image into the target dangerous good recognition model to obtain the dangerous good recognition information in the package to be detected; A generation module, configured to generate a dangerous good recognition result in the package to be detected based on all the dangerous good recognition information; Wherein, the preset database includes multiple dangerous good recognition models, and each of the dangerous good recognition models is provided with an encryption password. The acquiring a target dangerous good recognition model in the preset database based on the category code includes: Calculating the hash value corresponding to each coding character of the category code respectively based on a preset hash algorithm; wherein, the number of characters of each of the hash values is the same; Determine a target hash value among all the hash values; wherein, there are no identical characters between the target hash value and the category code. For each of the target hash values, calculate the variance between each pair of digits in the target hash value. Construct a blank table; wherein, the number of rows of the blank table is equal to the number of target hash values, and the number of columns of the blank table is equal to the number of characters in the target hash values. Arrange each of the target hash values in the blank table sequentially from top to bottom to obtain a target table; wherein, the variances corresponding to each of the target hash values in the target table increase sequentially from top to bottom. For each character in the target table, calculate the target number corresponding to the character; wherein, the target number is the sum of the row number and the column number corresponding to the character in the target table. Determine a target character among all the characters in the target table; wherein, the target number corresponding to the target character is a prime number. Based on the positions of each of the target characters in the target table, arrange each of the target characters sequentially to obtain a decryption password. Based on the decryption password, perform decryption processing on each of the dangerous goods identification models, and use the successfully decrypted dangerous goods identification models as the target dangerous goods identification models.

7. An X-ray security inspection device, characterized in that, The security inspection machine device includes a processor, a memory, and a computer program stored on the memory and executable by the processor. When the computer program is executed by the processor, it implements the artificial intelligence-based dangerous goods identification method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, it implements the artificial intelligence-based dangerous goods identification method according to any one of claims 1 to 5.

Images