Intelligent analysis method and device for security check and intelligent security check system

Abstract

The present disclosure provides an intelligent analysis method, device, system and storage medium for security inspection, the intelligent analysis method comprising: detecting a generation event of a newly added original X-ray image in a security inspection device, wherein the security inspection device generates an original X-ray image each time a scanning operation is performed; in response to the generation event, reading the newly added original X-ray image from the security inspection device through a standard file access interface; and inputting the read newly added original X-ray image into an intelligent recognition model to obtain a security recognition result corresponding to the newly added original X-ray image. The method effectively avoids the problems of repeated frames, missing frames and cross frames, does not require video decoding and frame extraction, reduces the demand for computing power, while ensuring the quality of the original image, improving the recognition accuracy and system stability.

Description

Technical Field

[0001] This disclosure relates to the field of image recognition technology, and in particular to an intelligent analysis method, device and intelligent security inspection system for security inspection. Background Technology

[0002] In public places such as airports and train stations, a large number of security screening devices are deployed. However, these traditional security screening devices usually only have image acquisition and display functions, requiring a lot of manual identification work. They lack intelligent recognition capabilities, have limited accuracy, and are extremely costly in terms of labor. In order to enable automatic recognition capabilities without replacing the deployed security screening devices, related technologies usually use external analysis equipment to capture or decode and extract frames of the video output from the security screening devices to obtain security images. These security images are then input into deep learning models for recognition.

[0003] However, in practical applications, such schemes are prone to problems such as duplicate frames, missing frames, and cross-frames (combined frames of two images) captured during image switching, which may cause the recognition results to not accurately correspond to the original scanning operation performed by the security inspection equipment. In addition, the image quality may be significantly reduced after video encoding and scaling, resulting in lower accuracy of the model's output and making it difficult to achieve good recognition results. Summary of the Invention

[0004] In view of this, the present disclosure aims to provide an intelligent analysis method, device, intelligent security inspection system, and storage medium for security inspection, in order to solve the problem that the identification results cannot accurately correspond to the original scan and the identification effect is poor in the related art.

[0005] In a first aspect, this disclosure provides an intelligent analysis method for security inspection, the method comprising: detecting a generation event of a newly added original X-ray image in a security inspection device, wherein the security inspection device generates an original X-ray image each time a scanning operation is performed; in response to the generation event, reading the newly added original X-ray image from the security inspection device through a standard file access interface; and inputting the read newly added original X-ray image into an intelligent recognition model to obtain a security recognition result corresponding to the newly added original X-ray image.

[0006] In conjunction with the first aspect, in some implementations, the security inspection equipment is configured with a target storage directory for storing raw X-ray images. The detection of the generation event of a newly added raw X-ray image in the security inspection equipment includes: continuously monitoring or periodically polling the target storage directory to monitor changes in the target storage directory; when a newly added raw X-ray image is detected in the target storage directory, it is determined that a generation event has occurred in the security inspection equipment.

[0007] In conjunction with the first aspect, some implementations include, before reading the newly added raw X-ray image from the security inspection equipment through the standard file access interface, the following steps are also included: confirming whether the newly added raw X-ray image has been completely written to the target storage directory; when it is confirmed that the newly added raw X-ray image has been completely written to the target storage directory, the step of inputting the read newly added raw X-ray image into the intelligent recognition model is performed.

[0008] In conjunction with the first aspect, in some implementation methods, the original X-ray images have their own image identifiers. The intelligent analysis method also includes: assigning a result identifier to the security identification result, wherein the result identifier of the security identification result has a unique mapping relationship with the image identifier of the newly added original X-ray image.

[0009] In conjunction with the first aspect, in some implementations, the image identifier includes the timestamp information of the corresponding original X-ray image, which is used to characterize the time when the original X-ray image was generated.

[0010] In conjunction with the first aspect, in some implementations, detecting the generation event of newly added original X-ray images in security inspection equipment includes: detecting the generation event of newly added original X-ray images in multiple security inspection equipment, wherein the image identifier includes the device information of the corresponding original X-ray image, and the device information is used to indicate the security inspection equipment that generated the original X-ray image among the multiple security inspection equipment.

[0011] In conjunction with the first aspect, in some implementations, the standard file access interface is a network file sharing protocol, including at least one of Windows Shared Directory, FTP, SFTP, and SMB.

[0012] Secondly, this disclosure provides an intelligent analysis device for security inspection, the intelligent analysis device comprising: a memory and a processor; the memory for storing relevant program code; and the processor for calling the program code to execute the intelligent analysis method in the first aspect or any possible implementation thereof.

[0013] Thirdly, this disclosure provides an intelligent security inspection system, which includes: a security inspection device for performing security inspection scans and generating corresponding raw X-ray images; and an intelligent analysis device, as described in the second aspect, for reading the raw X-ray images from the security inspection device via a standard file access interface for security identification.

[0014] Fourthly, this disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the intelligent analysis method described in the first aspect or any possible implementation thereof.

[0015] According to the intelligent analysis method, device, intelligent security inspection system, and storage medium for security inspection provided in this disclosure, in response to the generation event of a newly added original X-ray image in the security inspection device, the newly added original X-ray image is directly read from the security inspection device through a standard file access interface and input into an intelligent recognition model to obtain a security recognition result. This solution uses the generation event of the original X-ray image as a trigger mechanism to timely and accurately obtain the original X-ray image corresponding to the current scanning operation, effectively avoiding problems such as duplicate frames, missing frames, and overlapping frames. It achieves accurate correspondence between each original X-ray image and the recognition result, and has low computational requirements. Furthermore, since the original X-ray image has high image quality, this solution can also effectively improve the recognition accuracy. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 The diagram shown is an architectural schematic of an intelligent security inspection system provided in an embodiment of this disclosure.

[0018] Figure 2 The diagram shown is a flowchart of an intelligent analysis method for security inspection provided in an embodiment of this disclosure.

[0019] Figure 3 The diagram shown is a flowchart illustrating another intelligent analysis method for security checks provided in an embodiment of this disclosure.

[0020] Figure 4 The diagram shown is a structural schematic of an intelligent analysis device provided in an embodiment of this disclosure.

[0021] Figure 5 The diagram shown is a structural schematic of an intelligent analysis device provided in an embodiment of this disclosure. Detailed Implementation

[0022] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0023] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure. However, this disclosure may also be implemented in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this disclosure. Therefore, this disclosure is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of this disclosure. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] With the rapid development of artificial intelligence technology, intelligent recognition technology has been widely used in X-ray security inspection equipment in scenarios such as airports, train stations, subways, customs and logistics to automatically identify various prohibited items such as knives, firearms, liquids, and lithium batteries.

[0026] When upgrading traditional X-ray security inspection equipment with external analysis devices for intelligent recognition, the common practice is to access the video output signal of the security inspection equipment, capture frames from the displayed image, and then feed the captured single-frame image into a deep learning model for recognition. The working principle of this approach is that the security inspection equipment converts the scanned X-ray image into a video signal output, and the intelligent analysis device decodes and extracts frames from the video stream to obtain the image to be recognized from the continuous images. Because this method does not require modification to the original hardware and software structure of the security inspection equipment, and only requires the addition of external analysis equipment to achieve automatic recognition, the deployment cost is relatively low, making it widely adopted.

[0027] However, in actual security screening scenarios, when security equipment continuously performs scanning operations, especially with a large volume of luggage and high baggage throughput, the aforementioned method of video stream frame capture encounters a series of problems. Security equipment generates X-ray images as a series of scanning events, while the video stream is a continuous signal output at a fixed frame rate. Frame capture cannot reliably distinguish the boundaries of the original image from each scan. For example, when a piece of luggage is scanned and the image has just appeared on the screen without any changes, the video stream may still refresh with the original content, causing the same image to be captured and repeatedly recognized. Furthermore, when the security equipment quickly generates the next image, the frame capture moment may fall precisely during the image switching process, capturing a mixed and overlapping image, reducing recognition accuracy. Additionally, some images may not be captured at all, resulting in missed detections. Simultaneously, it is difficult to establish an accurate one-to-one correspondence between the recognition result and the original scanned image, causing inconvenience for subsequent verification.

[0028] The reason for the above problems is that the image acquisition of the video stream frame capture scheme depends on the display output and video encoding and decoding process. After the original X-ray image is generated inside the security inspection equipment, it needs to go through processes such as resolution scaling, video encoding, network transmission, decoding and extraction. This not only leads to image quality loss and wasted computing power, but also causes the one-to-one correspondence between each security scan action and the final acquired image to be lost.

[0029] In view of this, this disclosure proposes an intelligent analysis method for security inspection. By detecting the generation event of a newly added original X-ray image in the security inspection equipment, and promptly reading the newly added original X-ray image directly from the security inspection equipment through a standard file access interface when the generation event occurs, the method inputs the image into an intelligent recognition model to obtain the security recognition result. This effectively avoids the aforementioned problems in related technologies and improves the accuracy and temporal correspondence of the recognition result.

[0030] The technical solution of this disclosure will be described in detail below through specific embodiments.

[0031] Figure 1 This is a schematic diagram of the architecture of an intelligent security inspection system provided in an embodiment of this disclosure. Figure 1 As shown, the intelligent security inspection system includes at least one security inspection device (e.g., Figure 1 The system includes security inspection equipment 1, security inspection equipment 2... security inspection equipment N) and intelligent analysis equipment. The intelligent analysis equipment can read raw X-ray images from the security inspection equipment through a standard file access interface for security identification. The intelligent analysis equipment can connect to each security inspection equipment, for example, through a network or near-field communication, and communicate based on a standard file sharing protocol.

[0032] Security inspection equipment can be a traditional X-ray security inspection machine, used to scan objects with X-rays and generate corresponding raw X-ray images. Intelligent analysis equipment can be a computing device with processing capabilities. In practical applications, intelligent analysis equipment can be, for example, an industrial control computer or edge computing device with GPU or CPU computing power.

[0033] It should be noted that, although Figure 1 The intelligent analysis device is shown as a single block, but those skilled in the art will understand that it contains multiple functional modules deployed internally via software or firmware. These modules work together to implement the intelligent analysis method of this disclosure. Furthermore, the intelligent analysis device of this disclosure is not limited to a specific hardware form or internal module division method; any computing device capable of executing the intelligent analysis method described below can be used as the intelligent analysis device of this disclosure.

[0034] The following will combine Figure 2 The details are described above. Figure 1The system architecture shown illustrates the detailed process of the intelligent analysis method executed by the intelligent analysis device.

[0035] Figure 2 This is a flowchart illustrating an intelligent analysis method for security checks provided in an embodiment of this disclosure. This intelligent analysis method can be executed by a processor or server, for example, by... Figure 1 The intelligent analysis equipment in the intelligent security inspection system shown performs the operation. For example... Figure 2 As shown, the intelligent analysis method may include the following steps S210~S230.

[0036] Step S210: Detect the generation event of newly added raw X-ray images in the security inspection equipment.

[0037] When a security screening device performs a scan, it generates a raw X-ray image. At this time, a new raw X-ray image generation event is generated in the security screening device. In other words, the generation event can be understood as the occurrence of the fact that the security screening device generates a new raw X-ray image.

[0038] It should be noted that the newly added raw X-ray image refers to the X-ray image generated by the security inspection equipment after completing a single scan. This X-ray image is the raw data directly generated by the security inspection equipment, which is different from the secondary image obtained through video interception or format conversion in related technologies. For example, this X-ray image can be an X-ray image file written to a storage medium, which may include, but is not limited to, X-ray image files saved in formats such as BMP, PNG, JPEG, and DICOM.

[0039] Step S220: In response to the generation event, read the newly added raw X-ray image from the security inspection equipment through the standard file access interface.

[0040] Upon detecting the generation of a new raw X-ray image, the system responds by initiating a file read request to the corresponding security screening device via the network using a standard file access interface. This read process leverages the security screening device's native file sharing service, requiring no additional agents or drivers to be installed on the device. The intelligent analysis device directly reads the complete byte stream of the file, similar to accessing a local file system, thereby obtaining the raw X-ray image data without video compression, screen overlay, or format conversion.

[0041] It should be noted that the standard file access interface is a network file sharing protocol or data exchange mechanism, enabling intelligent analysis devices to directly read image data by accessing a remote file system. For example, this interface could be a Windows shared directory (SMB / CIFS protocol), FTP, SFTP, etc. Exemplarily, a connection can be established and the target file downloaded based on a pre-configured shared address and credentials, or the image byte stream can be read directly via network mounting. This method preserves the original resolution and pixel accuracy, reduces computational power requirements, and ensures image quality.

[0042] Step S230: Input the newly added original X-ray image into the intelligent recognition model to obtain the security recognition result corresponding to the newly added original X-ray image.

[0043] In the embodiments of this disclosure, the intelligent recognition model can be any computational model capable of processing input X-ray images to identify contraband or security-related items. It analyzes the morphology of items in the X-ray image to determine whether they contain contraband such as knives, firearms, or liquids, and outputs a security recognition result. For example, the intelligent recognition model can include, but is not limited to, deep learning-based convolutional neural networks, object detection models, image segmentation models, or image classification models, and can also be a Visual Large Language Model (VLM), etc.

[0044] Security identification results can be conclusions or datasets of security information output from the analysis of raw X-ray images. For example, this may include, but is not limited to, contraband category labels, target bounding boxes, confidence scores, alarm levels, or combinations thereof.

[0045] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, in response to the generation event of a newly added original X-ray image in the security inspection equipment, the newly added original X-ray image is directly read from the security inspection equipment through a standard file access interface and input into the intelligent recognition model to obtain the security recognition result. This solution uses the generation event of the original X-ray image as a trigger mechanism to timely and accurately obtain the original X-ray image corresponding to the current scanning operation, effectively avoiding problems such as duplicate frames, missing frames, and overlapping frames. It achieves accurate correspondence between each original X-ray image and the recognition result, and has low computational requirements. Furthermore, since the original X-ray image has high image quality, this solution can also effectively improve the recognition accuracy.

[0046] Optionally, in some embodiments, the security inspection equipment may be configured with a target storage directory for storing raw X-ray images. In order to detect new image generation events in the security inspection equipment in a timely and accurate manner, the above step S210 may specifically include: continuously monitoring or periodically polling the target storage directory to monitor changes in the target storage directory; when a new raw X-ray image is found in the target storage directory, it is determined that a generation event has occurred in the security inspection equipment.

[0047] In one embodiment, the target storage directory can be detected through continuous monitoring. For example, the target storage directory can be continuously monitored to detect changes, and a generation event is determined when a new image file is detected in the target storage directory. By utilizing the inotify or ReadDirectoryChangesW mechanisms provided by the operating system, a notification can be received immediately after a new file appears in the target storage directory, with a latency typically less than tens of milliseconds, providing good real-time performance.

[0048] Alternatively, in another embodiment, the target storage directory can be detected by periodic polling. For example, the file list in the target storage directory can be scanned at preset time intervals (e.g., 200ms or 500ms) and compared with the previous and current file lists. If a new file is found in the list, it is determined that a generation event has occurred.

[0049] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, by monitoring changes in the target storage directory of original X-ray images in the security inspection equipment, when a new original X-ray image is found in the target storage directory, a generation event is determined, thereby ensuring that every security inspection can be detected accurately and without omission, and providing accurate and reliable trigger signals for subsequent operations.

[0050] Figure 3 This disclosure provides a flowchart illustrating another intelligent analysis method for security checks.

[0051] In some cases, the writing process of newly added original X-ray images can be time-consuming. If reading begins before the images are fully written, the images will be incomplete, leading to recognition errors. To address this issue, in the embodiments of this disclosure, such as... Figure 3 As shown, Figure 2 Step S220 in the intelligent analysis method shown can be specifically implemented by including the following steps S310~320.

[0052] Step S310: Confirm that the newly added original X-ray image has been completely written to the target storage directory.

[0053] In practical applications, when security inspection equipment writes raw X-ray images to the target storage directory, the writing operation may not be completed instantaneously. Therefore, if the intelligent analysis device reads the image immediately after detecting the generation event, it may read incomplete image data that is still being written. Therefore, this embodiment adds an image file integrity verification step after detecting the generation event of a new raw X-ray image and before reading the new raw X-ray image from the security inspection equipment through the standard file access interface, so that the reading operation is performed only after confirming that the image has been completely written. This integrity verification step can, for example, determine whether the image file has been written completely by detecting the stability of the image file size (reading the file size twice within a preset time interval and determining whether the file size is the same and non-zero), or by obtaining the file close event issued by the operating system to determine that the file has been completely saved. This disclosure does not limit the specific verification method.

[0054] Step S320: When it is confirmed that the newly added raw X-ray image has been completely written to the target storage directory, the action of reading the newly added raw X-ray image from the security inspection equipment through the standard file access interface is executed.

[0055] After confirming integrity, the confirmed complete original X-ray image can be read from the security inspection equipment and input into the intelligent recognition process. Specifically, after confirming that the newly added original X-ray image has been completely written to the target storage directory through image file size stability detection or other integrity judgment methods, subsequent reading and recognition are triggered. In particular, when a readability signal is received, the complete original X-ray image is read from the target storage location through a standard file access interface, and its data is directly transmitted to the intelligent recognition model.

[0056] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, by adding a file integrity verification step after detecting the generation event of a new original X-ray image file and before reading the file through the standard file access interface, it is ensured that the image file read is a valid image file that has been completely written to the target storage location, thereby avoiding recognition errors or system anomalies caused by reading incomplete image files that are still being written, and improving the accuracy and reliability of the recognition results.

[0057] In an optional implementation, in order to establish a unique mapping relationship between the recognition result and the original image, the original X-ray images have their own image identifiers. The intelligent analysis method provided in the above embodiments may further include: assigning a result identifier to the security recognition result, wherein the result identifier of the security recognition result has a unique mapping relationship with the image identifier of the newly added original X-ray image.

[0058] It should be noted that the unique mapping relationship refers to the one-to-one correspondence between the image identifier and the result identifier, that is, a raw X-ray image is bound to only one security identification result, and the security identification result corresponds only to this one raw X-ray image.

[0059] Specifically, in this embodiment, the original X-ray image itself has an image identifier. This image identifier can be, for example, a filename automatically assigned by the security inspection equipment when generating the image file. The filename may contain a scan timestamp, equipment number, or serial number, or alternatively, the scan timestamp, equipment number, or serial number may be directly used as the image identifier. When reading the image through a standard file access interface, the image identifier can be acquired and recorded synchronously. Subsequently, after completing the analysis and outputting the security identification result, a unique result identifier (such as a database primary key, UUID, or auto-incrementing serial number, etc., which is not limited here) can be assigned to the result, and this result identifier can be bound to and stored with the aforementioned image identifier.

[0060] In some specific examples, the process of binding and storing result identifiers with image identifiers can be achieved in the following ways: for example, by creating a mapping table in a relational database, which includes fields such as image identifier, result identifier, recognition result content, recognition time, and security equipment number. A new record is inserted each time a recognition is completed, thereby ensuring that a unique mapping relationship is formed between each original image and its recognition result.

[0061] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, by establishing a unique mapping relationship, any recognition result can be directly associated with a unique original X-ray image. Security personnel or managers can quickly retrieve the image and its recognition result using conditions such as timestamps and device numbers. They can also retrieve the corresponding original image based on the recognition result at any time, or retrieve the recognition result based on the image, facilitating review, archiving, and statistical analysis. The mapping relationship is stored in a database, and even if the original image file is deleted due to storage expiration, the association information between the recognition result and the image is still retained, providing a data foundation for subsequent long-term analysis.

[0062] In one alternative implementation, the image identifier includes timestamp information of the corresponding original X-ray image, the timestamp information being used to characterize the time when the original X-ray image was generated.

[0063] Image identification includes the timestamp information of the corresponding original X-ray image. It should be noted that the timestamp information can refer to data recording the moment the original X-ray image was generated. For example, it may include, but is not limited to, the creation or modification time in the file system, the date and time string contained in the image filename, or the time field in the EXIF ​​information. The timestamp information is typically accurate to the second or millisecond level.

[0064] Specifically, the image file can be named 20260615120030.jpg, where 20260615120030 directly corresponds to the generation time, which is 12:00:30 on June 15, 2026. Using this as an image identifier can intuitively reflect the order of the image files, making it easy to sort and retrieve them by time.

[0065] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, by incorporating timestamp information into image identification, the chronological order of image files can be intuitively reflected, facilitating sorting and retrieval by time and enhancing the convenience of data management. More importantly, when retrieving the corresponding original image based on the identification results, accurate timestamp information is helpful in determining the accurate time location of security incidents or illegal items, making it easier for staff to implement corresponding measures.

[0066] In one optional implementation, in a scenario with multiple security screening devices, in order to achieve unified management of multiple independent security screening devices and clearly distinguish the source of each image, the generation event of a newly added original X-ray image in the security screening device is detected. This includes: detecting the generation event of a newly added original X-ray image in multiple security screening devices, wherein the image identifier includes the device information of the corresponding original X-ray image, and the device information is used to indicate the security screening device that generated the original X-ray image among the multiple security screening devices.

[0067] When multiple independent security screening devices need to be managed, by detecting the generation events of newly added original X-ray images of each security screening device and writing the device information into the image identifier, centralized management of multiple independent security screening devices can be achieved, and the source of each image file can be clearly distinguished.

[0068] Specifically, a centralized intelligent analyzer can be used to create separate detection tasks for different security inspection devices, each pointing to the target storage location of its respective device (e.g., the shared directory \\DeviceA\Images for security inspection device A, and the shared directory \\DeviceB\Images for security inspection device B). Each task independently monitors changes in its corresponding directory, and when a newly added raw X-ray image file is detected, it immediately captures the generation event.

[0069] Meanwhile, to clearly identify the source device for each image, this embodiment incorporates device information into the image identifier. This device information can be the security inspection device's serial number, IP address, or a preset device name. For example, when reading an image from security inspection device A, the device number "DEVICE_A" is combined with the timestamp in the filename to form an image identifier like DEVICE_A_20260615_120030_005.jpg (the image identifier DEVICE_A_20260615_120030_005.jpg means that the security inspection device with the serial number A generated the original X-ray image file during its 5th scan at 12:00:30 on June 15, 2026). Specifically, DEVICE_A indicates that the image comes from security scanner numbered A (device information); 20260615 represents the date, June 15, 2026; 120030 represents the time, 12:00:30; 005 indicates that the device performed its 5th scan that day; and .jpg represents the file format. In other words, by simply seeing this image identifier, one can know its device origin, generation time, and which scan it was performed on that day by the security scanner. Regardless of how it is subsequently stored, transmitted, or analyzed, each image can be accurately located and traced.

[0070] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, by centrally detecting multiple devices and identifying device information, the system can intuitively determine which security inspection device the result comes from through image identification, thereby achieving accurate data management in scenarios with multiple security inspection devices.

[0071] In some embodiments of this disclosure, to ensure broad compatibility, the standard file access interface may employ a network file sharing protocol, such as at least one of Windows Shared Directory, FTP, SFTP, and SMB.

[0072] In practical applications, security inspection equipment of different brands and models may run on different operating systems, and the file sharing protocols they support also vary. To ensure broad compatibility, this embodiment limits the standard file access interface to network file sharing protocols, including but not limited to: Windows Shared Directory: Most security screening devices are based on the Windows operating system, and their file system supports folder sharing. Simply enable sharing on the security screening device and set the directory used to store raw X-ray images as a shared folder, and it can be accessed directly via the local area network. This method does not require installing any additional software on the security screening device, nor does it require any technical support from the security screening device manufacturer.

[0073] FTP (File Transfer Protocol): A classic file transfer protocol. If the security inspection equipment runs on a Linux system or supports FTP service, an FTP client can connect to the FTP server of the security inspection equipment to browse and download newly added raw X-ray image files. The FTP protocol is mature, highly cross-platform, and suitable for scenarios where Windows sharing is not possible.

[0074] SFTP (SSH File Transfer Protocol): An encrypted file transfer protocol based on SSH, offering enhanced security. It is suitable for scenarios with high data transfer security requirements, especially those requiring data transfer over the internet or untrusted networks.

[0075] SMB (Server Message Block Protocol): SMB is a protocol natively supported by Windows systems and is also widely used in Linux devices that support SAMBA services. Through the SMB protocol, intelligent analysis devices can directly manipulate files in the shared directory of security inspection equipment as if they were local files, achieving efficient and real-time file reading.

[0076] In actual implementation, multiple protocols mentioned above can be supported simultaneously, and the most suitable protocol can be automatically selected for connection based on the actual capabilities of the security inspection equipment. For example, it may prioritize SMB connection and automatically switch to FTP if that fails.

[0077] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, the standard file access interface is limited to network file sharing protocols, among which Windows shared directories, FTP, SFTP, SMB, and other protocols are relatively basic and universal file sharing methods in the computer field, and can be supported by most operating systems and devices. This solution is compatible with security inspection equipment of different brands, models, and operating systems, without requiring the installation of any agents, plugins, or drivers on the security inspection equipment, nor does it require customized development for specific devices.

[0078] In a preferred embodiment, the intelligent analysis device also supports linkage with audible and visual alarms, monitoring platforms, and business systems.

[0079] Specifically, when the security identification result output by the intelligent identification model indicates the presence of suspected contraband (for example, the confidence level exceeds a preset threshold), an alarm command will be automatically generated and a trigger signal will be sent to the on-site audible and visual alarm via a network (such as TCP / IP) or a dedicated I / O interface (such as a relay) to achieve immediate alarm and remind security personnel to pay attention.

[0080] Simultaneously, the original X-ray images, security identification results (including contraband categories, confidence levels, and bounding box information), image identifiers, equipment information, and timestamps can be packaged into a structured data packet and pushed to the remote monitoring platform in real time via standard network communication protocols. The image interpreters on the monitoring platform can view the high-resolution original images and analysis results immediately for manual review and confirmation.

[0081] Furthermore, intelligent analysis equipment can interface with airport baggage handling systems, security information management systems, or customs passenger clearance systems via standardized API interfaces. When prohibited items are identified, alarm information can be automatically linked to the baggage's barcode information and the passenger's identification information. For example, if suspected liquid is detected in checked baggage, the system can automatically sort the baggage to the baggage inspection channel and push relevant information to the security officer's terminal device, thereby improving security efficiency and the accuracy of incident handling.

[0082] According to the intelligent analysis method for security inspection provided in the above embodiments of this disclosure, by linking the intelligent analysis device with the sound and light alarm, the monitoring platform and the business system, security personnel do not need to switch back and forth between different systems, thereby improving the actual work efficiency and response speed of security personnel.

[0083] The intelligent analysis method of this disclosure will be illustrated below through a specific embodiment: In the baggage security area of ​​a major international airport, twelve traditional X-ray security scanners from different manufacturers, all of which have been in use for many years, are placed. These scanners originally only had image acquisition and display functions and lacked the ability to automatically identify prohibited items. To achieve a rapid and low-cost upgrade, the technical solution disclosed in this paper is adopted, which adds an intelligent analysis device without requiring any software or hardware modifications to any of the scanners, and connects all the scanners and the intelligent analysis device to the same gigabit local area network.

[0084] Security personnel maintain consistent operating procedures on each security screening machine. When a piece of passenger luggage is sent into any security screening machine lane, the machine performs a scan and automatically generates a raw X-ray image file in its local target storage directory. The file name includes the device number, date, and timestamp, for example, DEVICE03_20251214143022.jpg. Simultaneously, the security screening machine exposes this target storage directory to the public through a pre-enabled Windows shared directory service, providing a standard file access interface.

[0085] The intelligent analysis equipment utilizes the file system event listening mechanism provided by the operating system to continuously monitor twelve target storage directories shared by twelve security inspection machines. When a new image file appears in a directory of a security inspection machine, and the write operation is completed while the file size remains unchanged for 200 milliseconds, it confirms that the file has been completely written. Immediately, it transmits the generated event over the network and, based on pre-configured sharing credentials, directly reads the newly added raw X-ray image byte stream via the SMB protocol to obtain a full-resolution image without video encoding / decoding or scaling.

[0086] The raw X-ray images are fed into a deep learning recognition model in an intelligent analysis device. This model, trained on a large number of X-ray contraband samples, can analyze the shape, density, and texture features of objects in the image in real time, automatically determining whether they contain knives, firearms, explosives, liquids, or other contraband. After recognition, the model outputs a security identification result, including the contraband category, confidence level, and location information within the image. Upon receiving the identification result, the image file name is used as an image identifier to assign a result identifier, establishing a unique mapping between the two. For example, for the image file DEVICE03_20251214143022.jpg, the identification result is marked with the same identifier string, ensuring that the original image can be directly retrieved based on any result, or the identification conclusion can be looked up from the image.

[0087] The entire identification process takes less than 200 milliseconds. If the identification result is deemed safe, the record is stored for future reference; if a suspected contraband item is identified, an alert can be issued via an audible and visual alarm within less than one second, and the original image, identification result, and information are simultaneously pushed to the security inspector's display terminal. The inspector can directly view the high-resolution original X-ray image and the marked suspicious item area on the integrated interface, and make an on-the-spot decision to open the bag for inspection or allow passage.

[0088] In this scenario, this disclosure improves the implementation speed and system stability of intelligent security inspection transformation, realizes a one-to-one correspondence between each original X-ray image and the recognition result, effectively avoids the problems of duplicate frames, missing frames and cross frames, eliminates the need for video decoding and frame extraction, reduces computing power requirements, and at the same time ensures the quality of the original image, improves recognition accuracy and system stability.

[0089] Figure 4 The diagram shown is a structural schematic of the intelligent analysis device provided in an embodiment of this disclosure. Based on the same concept as the intelligent analysis methods provided in the above embodiments, referencing... Figure 4 This disclosure also provides an intelligent analysis device 400.

[0090] like Figure 4As shown, the intelligent analysis device 400 may include: a detection module 410, used to detect the generation event of a newly added original X-ray image in the security inspection equipment, wherein the security inspection equipment generates one original X-ray image each time a scanning operation is performed; a reading module 420, used to read the newly added original X-ray image from the security inspection equipment through a standard file access interface in response to the generation event; and a recognition module 430, used to input the read newly added original X-ray image into an intelligent recognition model to obtain a security recognition result corresponding to the newly added original X-ray image.

[0091] According to the intelligent analysis device provided in the above embodiments, in response to the generation event of a newly added original X-ray image in the security inspection equipment, the device directly reads the newly added original X-ray image from the security inspection equipment through a standard file access interface and inputs it into the intelligent recognition model to obtain the security recognition result. This solution uses the generation event of the original X-ray image as a trigger mechanism to timely and accurately obtain the original X-ray image corresponding to the current scanning operation, effectively avoiding problems such as duplicate frames, missing frames, and overlapping frames. It achieves accurate correspondence between each original X-ray image and the recognition result, and has low computational requirements. Furthermore, because the original X-ray image has high image quality, this solution can also effectively improve the recognition accuracy.

[0092] In some embodiments of this disclosure, the security inspection equipment is configured with a target storage directory for storing raw X-ray images. The detection module 410 is also used to continuously monitor or periodically poll the target storage directory to monitor changes in the target storage directory. When a newly added raw X-ray image is found in the target storage directory, it is determined that a generation event has occurred in the security inspection equipment.

[0093] In some embodiments of this disclosure, before reading the newly added original X-ray image from the security inspection device through the standard file access interface, the reading module 420 is specifically used to: confirm whether the newly added original X-ray image has been completely written to the target storage directory; when it is confirmed that the newly added original X-ray image has been completely written to the target storage directory, perform the action of reading the newly added original X-ray image from the security inspection device through the standard file access interface.

[0094] In some embodiments of this disclosure, the original X-ray images have their own image identifiers, and the recognition module 430 is further configured to: assign a result identifier to the security recognition result, wherein the result identifier of the security recognition result has a unique mapping relationship with the image identifier of the newly added original X-ray image.

[0095] In some embodiments of this disclosure, the image identifier includes timestamp information of the corresponding original X-ray image, which is used to characterize the time when the original X-ray image was generated.

[0096] In some embodiments of this disclosure, the detection module 410 is further configured to detect the generation event of a newly added original X-ray image among multiple security inspection devices, wherein the image identifier includes device information of the corresponding original X-ray image, and the device information is used to indicate the security inspection device among the multiple security inspection devices that generated the original X-ray image.

[0097] In some embodiments of this disclosure, the standard file access interface is a network file sharing protocol, including at least one of Windows Shared Directory, FTP, SFTP, and SMB.

[0098] It should be noted that the explanations and descriptions of the technical effects of the aforementioned intelligent analysis method embodiments also apply to the intelligent analysis device in this embodiment, and will not be repeated here.

[0099] Figure 5 The diagram shown is a structural schematic of the intelligent analysis device provided in an embodiment of this disclosure. Figure 5 The intelligent analysis device 500 shown includes a memory 510, a processor 520, a communication interface 530, and a bus 540. The memory 510, processor 520, and communication interface 530 are interconnected via the bus 540.

[0100] It should be noted that the explanations and descriptions of the aforementioned intelligent analysis method embodiments also apply to the intelligent analysis device of this embodiment, and will not be repeated here.

[0101] The memory 510 may be a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 510 may store a program, and when the program stored in the memory 510 is executed by the processor 520, the processor 520 and the communication interface 530 are used to execute the various steps of the intelligent analysis method of the embodiments of this disclosure.

[0102] The processor 520 may be a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), graphics processing unit (GPU), or one or more integrated circuits, used to execute relevant programs to achieve the functions required by the units in the intelligent analysis method of this disclosure embodiment.

[0103] The processor 520 can also be an integrated circuit chip with signal processing capabilities. In implementation, each step of the intelligent analysis method of this disclosure can be completed by the integrated logic circuits in the hardware of the processor 520 or by instructions in software form. The processor 520 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this disclosure can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory 510. The processor 520 reads the information in the memory 510 and, in conjunction with its hardware, performs the functions required by the units included in the intelligent analysis method of this disclosure embodiment, or executes the intelligent analysis method of this disclosure method embodiment.

[0104] The communication interface 530 uses a transceiver device, such as, but not limited to, a transceiver, to enable communication between the intelligent analysis device 500 and other devices or communication networks. For example, sensor data can be acquired through the communication interface 530.

[0105] Bus 540 may include a pathway for transmitting information between various components of the intelligent analysis device 500 (e.g., memory 510, processor 520, communication interface 530).

[0106] It should be noted that, although Figure 5 The illustrated intelligent analysis device 500 only shows the memory, processor, and communication interface. However, those skilled in the art should understand that in specific implementations, the intelligent analysis device 500 may also include other devices necessary for normal operation. Furthermore, depending on specific needs, those skilled in the art should understand that the intelligent analysis device 500 may also include hardware devices for implementing other additional functions. Moreover, those skilled in the art should understand that the intelligent analysis device 500 may only include the devices necessary for implementing the embodiments of this disclosure, and may not necessarily include... Figure 5 All the devices shown.

[0107] In addition to the methods, apparatus, and devices described above, embodiments of this disclosure may also be computer program products, including computer program instructions that, when executed by a processor, cause the processor to perform the steps of the intelligent analysis methods provided in the various embodiments of this disclosure.

[0108] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0109] Furthermore, embodiments of this disclosure may also be computer-readable storage media storing a computer program that, when executed by a processor, causes the processor to perform the various steps of the intelligent analysis methods provided in the various embodiments of this disclosure.

[0110] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0111] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0112] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0113] In the several embodiments provided in this disclosure, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0114] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0115] In addition, the functional units in the various embodiments of this disclosure can be integrated into a similar region segmentation unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0116] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0117] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. An intelligent analysis method for security inspection, characterized in that, include: The event of generating a new original X-ray image in the security inspection equipment is detected, wherein the security inspection equipment generates one original X-ray image each time a scanning operation is performed; In response to the generation event, the newly added raw X-ray image is read from the security inspection equipment via a standard file access interface; The newly added original X-ray image is input into the intelligent recognition model to obtain a security recognition result corresponding to the newly added original X-ray image.

2. The intelligent analysis method according to claim 1, characterized in that, The security inspection equipment is configured with a target storage directory for storing the original X-ray images. The event of generating a newly added original X-ray image in the security inspection equipment includes: The target storage directory is continuously monitored or periodically polled to detect changes in the target storage directory; When the newly added original X-ray image is found in the target storage directory, it is determined that the generation event has occurred in the security inspection equipment.

3. The intelligent analysis method according to claim 2, wherein, Before reading the newly added raw X-ray image from the security screening device via a standard file access interface, the method further includes: Confirm that the newly added original X-ray image has been completely written to the target storage directory; Once it is confirmed that the newly added raw X-ray image has been completely written to the target storage directory, the action of reading the newly added raw X-ray image from the security inspection device through the standard file access interface is executed.

4. The intelligent analysis method according to claim 1, characterized in that, The original X-ray images each have their own image identifiers, and the intelligent analysis method further includes: The security identification result is assigned a result identifier, wherein the result identifier of the security identification result has a unique mapping relationship with the image identifier of the newly added original X-ray image.

5. The intelligent analysis method according to claim 4, characterized in that, The image identifier includes the timestamp information of the corresponding original X-ray image, which is used to characterize the time when the original X-ray image was generated.

6. The intelligent analysis method according to claim 4, characterized in that, The event of generating newly added raw X-ray images in the security inspection equipment includes: Detect the generation events of newly added raw X-ray images from multiple security screening devices. The image identifier includes device information corresponding to the original X-ray image, and the device information is used to indicate the security inspection device that generated the original X-ray image among the plurality of security inspection devices.

7. The intelligent analysis method according to claim 1, wherein, The standard file access interface is a network file sharing protocol, including at least one of Windows shared directories, FTP, SFTP, and SMB.

8. An intelligent analysis device for security inspection, characterized in that, include: Memory and processor; The memory is used to store the relevant program code; The processor is used to call the program code to execute the intelligent analysis method according to any one of claims 1 to 7.

9. An intelligent security inspection system, characterized in that, include: Security screening equipment is used to perform security checks and scans and generate corresponding raw X-ray images; The intelligent analysis device as described in claim 8 is used to read the original X-ray image from the security inspection device via a standard file access interface for security identification.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the intelligent analysis method according to any one of claims 1 to 7.

Images