Lost article information management device and method and electronic equipment
By automating information collection and generating structured documents through large-scale model analysis, the problems of low efficiency and low accuracy in information collection in lost and found management have been solved, enabling efficient search and management of lost and found items.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- INST OF COMPUTING TECH CHINA ACAD OF RAILWAY SCI
- Filing Date
- 2026-02-04
- Publication Date
- 2026-05-12
AI Technical Summary
The existing lost and found management system suffers from low information collection efficiency, inconsistent information quality, and a lack of structured data organization, resulting in low search efficiency and accuracy.
The system automatically collects information on lost items using an information collection module, generates structured files through an information processing module, and stores them using a storage module. Combined with auxiliary equipment such as image acquisition components, text recognition scanners, and microphones, it uses a large model to analyze and generate descriptive information, forming a standardized JSON format file.
It improves the efficiency and accuracy of lost item information collection, reduces human error, and enables efficient search and management of lost items.
Smart Images
Figure CN122019808A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data management, and more specifically, to an information management device, method, and electronic device for lost items. Background Technology
[0002] Currently, the management of lost and found items in public places faces problems such as low efficiency in information collection, inconsistent information quality, and insufficient data storage and utilization. Traditional lost and found management models rely on manual recording and visual judgment, which are highly susceptible to human factors, resulting in highly subjective information descriptions, low data standardization, and difficulty in forming effective data support, thus affecting the accuracy of finding lost items. Existing management systems often store collected information only in the form of scattered text or images, lacking structured data organization, making it difficult to effectively utilize the information and reducing the efficiency of finding lost items.
[0003] Therefore, improving the efficiency and accuracy of finding lost items is a technical problem that needs to be solved by those skilled in the art. Summary of the Invention
[0004] To address one or more deficiencies in the existing technology, the present invention provides an information management device, method, and electronic device for lost items.
[0005] A lost item information management device includes an information acquisition module, an information processing module, and a storage module, wherein: The information collection module is connected to the information processing module and is used to collect the item information of the lost item and input the item information into the information processing module; The information processing module is used to process the item information and generate a structured file; The storage module is connected to the information processing module and is used to store the structured file.
[0006] Optionally, the information acquisition module includes an image acquisition component, which includes: The shooting compartment has a backdrop on its inner wall; A shelf located inside the shooting compartment is used to place the lost items; A first camera, located on the top side of the shooting compartment, is used to capture first image information of the lost item.
[0007] Optionally, the shelf is a transparent, rotatable shelf; The image acquisition component also includes: A second camera, located at the bottom of the shooting compartment, is used to capture second image information of the lost item; Lighting equipment located at the bottom of the shooting chamber.
[0008] Optionally, the information acquisition module further includes an auxiliary information acquisition component, which includes: A text recognition scanner is used to collect text records related to the process of picking up the lost item; A microphone is used to collect audio recordings related to the retrieval process of the lost item.
[0009] A method for managing information on lost items, applied to the apparatus described in any one of the above claims, the method comprising: Collect information about lost items; The information of the lost items is analyzed based on a pre-set large model to form descriptive information about the lost items; A structured file is generated and stored based on the item information and the description information.
[0010] Optionally, the collection of lost item information includes: The camera captures image information of the lost items; The proportion of the lost item in the image is determined based on the image information; Control the camera to zoom and adjust so that the proportion of the lost item in the image meets the preset proportion requirements; Control the camera to focus so that the image information of the lost item meets the preset clarity requirements.
[0011] Optionally, the analysis of the item information based on a preset large model to form descriptive information about the lost item includes: Visual features of the image information are extracted using a preset visual encoder; The image information and visual features are analyzed using the large model to form first descriptive information; the first descriptive information is used to describe the appearance attributes of the lost item.
[0012] Optionally, the collection of lost item information includes: Audio recordings related to the retrieval process of the lost item were collected using a microphone; The analysis of the item information based on a preset large model generates descriptive information about the lost item, including: The audio recording information is converted into audio features using a preset audio encoder; The audio features are analyzed using the large model to form second descriptive information; the second descriptive information is used to describe the process of picking up lost items.
[0013] Optionally, the collection of lost item information includes: Scan the paper record sheet to generate a text record of the lost items; The analysis of the item information based on a preset large model generates descriptive information about the lost item, including: The large model is used to analyze the text record information to form a description of the process of picking up the lost item.
[0014] An electronic device, comprising: A processor and a memory, the memory being used to store at least one instruction, which, when loaded and executed by the processor, implements the information management method for lost items as described in any of the preceding embodiments.
[0015] The lost item information management device provided in this invention includes an information acquisition module, an information processing module, and a storage module. The information acquisition module is connected to the information processing module and is used to collect item information of the lost item and input the item information into the information processing module. The information processing module processes the item information to generate a structured file. The storage module is connected to the information processing module and is used to store the structured file. This invention, by inputting item information into the information processing module through the information acquisition module, avoids information loss and errors during manual input and transmission, improving the efficiency and accuracy of item information collection. The information processing module processes the collected data and outputs structured data, facilitating subsequent storage and retrieval, thus improving the efficiency and accuracy of finding lost items. Attached Figure Description
[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 This is a schematic diagram of the structure of an information management device for lost items provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of another lost item information management device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an image acquisition component 110 provided in an embodiment of the present invention; Figure 4 This is a flowchart of a lost item information management method provided in an embodiment of the present invention. Detailed Implementation
[0018] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0019] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0020] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.
[0021] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.
[0022] Currently, the management of lost and found items in public places faces problems such as low efficiency in information collection, inconsistent information quality, and insufficient data storage and utilization. Traditional lost and found management models rely on manual recording and visual judgment, which are highly susceptible to human factors, resulting in highly subjective information descriptions, low data standardization, and difficulty in forming effective data support, thus affecting the accuracy of finding lost items. Existing management systems often store collected information only in the form of scattered text or images, lacking structured data organization, making it difficult to effectively utilize the information and reducing the efficiency of finding lost items.
[0023] Therefore, the present invention provides an information management device, method and electronic device for lost items to solve the above problems.
[0024] Please refer to Figure 1 The diagram below illustrates the structure of an information management device for lost items provided in an embodiment of the present invention. It includes an information acquisition module 100, an information processing module 200, and a storage module 300, wherein: The information collection module 100 is connected to the information processing module 200 and is used to collect the item information of the lost item and input the item information into the information processing module 200. The information processing module 200 is used to process item information and generate structured files; The storage module 300 is connected to the information processing module 200 and is used to store structured files.
[0025] Traditional lost and found management requires dedicated personnel to register, describe, and locate items. However, manual data entry is highly subjective, and descriptions of the same item can vary significantly depending on the person describing it, making subsequent searches difficult. Therefore, in this embodiment, the user simply places the found lost item into the information management device. The information collection module 100 and information processing module 200 within the device automatically collect and convert the item information, while the storage module 300 automatically stores the structured file. This embodiment transforms the traditional work model, reliant on manual observation, recording, and repeated communication, into a standardized operation model driven by equipment and assisted by personnel. This helps reduce labor costs, effectively decouples the data collection process from personnel status, avoids subjective description bias and data entry errors caused by fatigue, and improves the objectivity and consistency of item information collection.
[0026] In this embodiment, the information collection module 100 is used to collect various information about the lost item, including but not limited to images, voice descriptions (from the finder or witnesses), and possible written records such as passenger handover slips. By inputting the item information into the information processing module 200 through the information collection module 100, information loss and errors during manual entry and transmission can be avoided, improving the efficiency and accuracy of item information collection. The information collection module 100 inputs the collected item information into the information processing module 200, enabling the information processing module 200 to generate a structured file based on the item information, facilitating subsequent storage and retrieval, thereby improving the efficiency and accuracy of finding lost items.
[0027] In this embodiment, the information processing module 200 can utilize existing technologies to process the item information. For example, it can use existing data format conversion technologies to convert the item information into a structured file, such as a JSON data file, which facilitates system retrieval and subsequent searching and matching.
[0028] In this embodiment, the storage module 300 is used to store structured files. For example, a database can be used for storage. The choice of database can be evaluated based on actual needs; for example, a relational database or a NoSQL database can be selected. Relational databases have good data integrity and transaction processing capabilities, making them suitable for scenarios requiring strong consistency; NoSQL databases, on the other hand, have better scalability and flexibility, making them suitable for scenarios that need to process large amounts of unstructured data.
[0029] Based on the above technical solution, the lost item information management device provided in this embodiment of the invention includes an information acquisition module 100, an information processing module 200, and a storage module 300. The information acquisition module 100 is connected to the information processing module 200 and is used to collect item information of the lost item and input the item information into the information processing module 200. The information processing module 200 processes the item information to generate a structured file. The storage module 300 is connected to the information processing module 200 and is used to store the structured file. This invention, by inputting item information into the information processing module 200 through the information acquisition module 100, avoids information loss and errors during manual input and transmission, improving the efficiency and accuracy of item information collection. The information processing module 200 processes the collected data and outputs structured data, facilitating subsequent storage and retrieval, thus improving the efficiency and accuracy of finding lost items.
[0030] Please refer to Figure 2 This is a schematic diagram of another lost item information management device provided in an embodiment of the present invention. Figure 2 As shown, the information acquisition module 100 includes an image acquisition component 110. The image acquisition component 110 is used to acquire image information of lost items.
[0031] Please refer to Figure 3 This is a schematic diagram of the structure of an image acquisition component 110 provided in an embodiment of the present invention. The image acquisition component 110 may include: The shooting chamber 112 has a background panel 111 on its inner wall; A storage table 113 located inside the shooting compartment 112 is used to place lost items; A first camera 114 is installed on the top side of the shooting compartment 112 to capture first image information of lost items.
[0032] Traditional open shooting environments are easily affected by factors such as lighting, background, and angle, resulting in inconsistent image quality that requires manual adjustment and post-processing, leading to low efficiency. Therefore, in this embodiment, the image acquisition component 110 integrates the first camera 114 into the shooting chamber 112 and works with the storage table 113 to place lost items, standardizing the shooting process and improving the efficiency of item information acquisition while reducing human error. Users only need to place the lost item on the storage table 113 and press the start button to automatically complete image acquisition, without the need for complex adjustments and settings. This integrated design of the image acquisition component 110 not only improves work efficiency but also reduces operational difficulty.
[0033] In some embodiments, the image acquisition component 110 can also be equipped with different models of cameras to improve image resolution and clarity as needed; additional sensors, such as depth sensors, can also be added to acquire three-dimensional information of objects. This modular design ensures the flexibility and scalability of the device, enabling it to adapt to future application needs. Furthermore, the image acquisition component 110 should also have a certain fault tolerance mechanism, such as automatically switching to a backup camera when a camera malfunctions, ensuring the normal operation of the system.
[0034] In this embodiment, the shooting compartment 112 serves to provide a closed, standard shooting environment for the first camera 114. A background panel 111 is installed on the inner wall of the shooting compartment 112 to eliminate interference from external backgrounds and ensure a uniform background in the captured images. The background panel 111 can be pure white or light gray to minimize the effects of light reflection and shadows, making the images clearer and brighter. The shooting compartment 112 can be made of lightweight, easy-to-clean materials, such as plastic or metal sheets. The size of the shooting compartment 112 can be designed according to the actual application scenario to accommodate lost items of different sizes.
[0035] In this embodiment, the function of the shelf 113 is to provide a stable support platform for placing lost items, ensuring that the items remain horizontal and in a fixed position during the shooting process. The height of the shelf 113 can be adjusted according to the shooting angle of the first camera 114 to ensure that the captured image can include all the details of the lost items. The surface of the shelf 113 can be provided with an anti-slip texture to prevent the items from sliding.
[0036] In this embodiment, the first camera 114 is installed on the inner top side of the shooting compartment 112 so as to capture the overall shape and detailed features of the lost item from a top-down angle and obtain the first image information of the lost item.
[0037] The image acquisition component 110 of this embodiment improves the quality and efficiency of image acquisition of lost items by adopting the design of a shooting chamber 112 with an internal background plate 111 and a storage platform 113, and ensures the clarity and consistency of the acquired images.
[0038] In some embodiments, the shelf 113 is a transparent, rotatable shelf; Image acquisition component 110 also includes: A second camera 115 is installed at the bottom of the shooting compartment 112 to capture second image information of lost items; Lighting equipment 116 is located at the bottom of the shooting chamber 112.
[0039] In this embodiment, a transparent, rotatable shelf is used to solve the problem of traditional shelves obstructing the view and making it difficult to photograph the bottom of items. Using a transparent material (such as acrylic or glass) eliminates visual obstruction, allowing the second camera 115, mounted at the bottom of the shooting compartment 112, to clearly capture information about the bottom of the item, thus achieving omnidirectional image acquisition of lost items. The rotation function of the shelf 113 allows the first camera 114 to shoot from different angles, thereby constructing complete image data of the item.
[0040] In some embodiments, to prevent lost items from slipping during rotation, the surface of the shelf 113 may be provided with anti-slip textures or grooves.
[0041] In actual operation, the user only needs to place the lost item on the transparent rotatable shelf 113, start the rotation motor, and the first camera 114 and the second camera 115 will simultaneously collect the image information of the item, and finally obtain the first image information and the second image information containing all angle information of the lost item.
[0042] In this embodiment, the second camera 115 is used to capture the bottom information of the lost item (i.e., the second image information) to supplement the image information captured by the first camera 114, thereby achieving omnidirectional image acquisition of the item.
[0043] In this embodiment, the lighting device 116 located at the bottom of the shooting chamber 112 provides uniform illumination to the shooting chamber 112 and compensates for the influence of bottom shadows, ensuring that the image captured by the second camera 115 is clear and bright. The lighting device 116 can use LED light strips or ring lights as light sources. The position and angle of the lighting device 116 are precisely adjusted to avoid unnecessary shadows and reflections, ensuring uniform brightness of the captured image.
[0044] In some embodiments, in order to achieve automated lighting adjustment, the lighting device 116 may be equipped with a light sensor and an automatic brightness control system to automatically adjust the brightness according to the light intensity in the shooting chamber 112.
[0045] In some embodiments, the information acquisition module 100 further includes an auxiliary information acquisition component 120, which includes: A text recognition scanner is used to collect text records related to the process of finding lost items; A microphone is used to collect audio recordings related to the process of retrieving lost items.
[0046] In this embodiment, the text recognition scanner is used to collect text records related to the lost item retrieval process, such as passenger handover slips, leftover notes, or memos. This text information typically contains key information such as the type of lost item, the location where it was retrieved, and the time, which is crucial for subsequent matching of lost items and owner recognition.
[0047] In practice, staff only need to place the text record within the field of view of the text recognition scanner and start the scanner to automatically collect the data.
[0048] In this embodiment, the microphone is used to collect audio recordings related to the process of retrieving lost items, such as verbal descriptions from the person retrieving the item. This audio information often contains details that are difficult to express visually, such as special markings on the item and features of the surrounding environment, which are crucial for the accurate identification of lost items.
[0049] Please refer to Figure 4 The flowchart below illustrates a method for managing information on lost items provided in an embodiment of the present invention. Applied to the apparatus described in any of the above embodiments, the method includes the following steps: Step S01: Collect information about the lost items.
[0050] In this embodiment, collecting information about lost items refers to obtaining information describing the various attributes or characteristics of the lost items. This information may include, but is not limited to, multi-view images, auxiliary information (such as handover forms or verbal descriptions), and covers the item's appearance, possible markings, the circumstances under which it was found, and any clues that can help with identification and matching. By collecting this information, a complete lost item file can be established, facilitating subsequent intelligent identification and matching, and ultimately helping the owner recover their lost items.
[0051] In some embodiments, item information may include image information. Based on this, to improve the stability and standardization of the image information acquisition process, avoid image blurring or poor composition due to differences in item size, reduce the need for manual intervention, and increase the automation level of image acquisition, proportion and clarity control based on image information can be introduced into the image acquisition process for lost items. Specifically, as mentioned in step S01, collecting item information for lost items may include the following steps: Step S11: Collect image information of the lost item using a camera.
[0052] In this embodiment, image information of the lost item is captured by a camera. This image information forms the basis for subsequent image analysis and includes visual features such as the item's color, shape, and texture. The image information records the visual state of the lost item at a specific moment, providing raw data for subsequent proportion determination and sharpness adjustment.
[0053] Step S12: Determine the proportion of the lost item in the image based on the image information.
[0054] In this embodiment, determining the proportion of the lost item in the image based on image information refers to using image processing algorithms to analyze image data and calculate the area or pixel ratio occupied by the lost item in the image. Proportion is a numerical indicator that measures the size of the space occupied by the item in the image, used for subsequent zoom adjustments to ensure that the item's proportion in the image meets preset requirements, facilitating subsequent image analysis.
[0055] Determining the proportion of the object is crucial to ensuring that subsequent zoom adjustments accurately place the object within the appropriate image area. If the object is too small or too large in the image, it may affect the accuracy of subsequent sharpness adjustments and image analysis. By determining the proportion, a clear reference point is provided for zooming, making zoom adjustments more precise.
[0056] In some embodiments, the proportion of a lost item in an image can be determined using an object detection algorithm (e.g., YOLO v8 / YOLO-NAS). The object detection algorithm can identify lost items in an image and output bounding boxes that define the location and size of the lost items. The size information of the bounding boxes can be used to calculate the proportion of the item in the image; for example, the area of the bounding box divided by the total area of the image.
[0057] Step S13: Control the camera to zoom and adjust so that the proportion of the lost item in the image meets the preset proportion requirements.
[0058] In this embodiment, by changing the focal length of the camera, the magnification or reduction of the lost item is adjusted so that the proportion of the lost item in the image reaches a preset value. The preset proportion requirement ensures that the lost item has an appropriate size in the image, facilitating subsequent clarity adjustment and analysis.
[0059] In some embodiments, the proportion of the lost item in the image can be compared with a preset proportion (e.g., a preset proportion of 40%). If the proportion does not meet the requirements, the system adjusts the camera focus according to the deviation until the proportion meets the requirements. For example, if the proportion is too small, the image needs to be enlarged; if the proportion is too large, the image needs to be reduced.
[0060] Step S14: Control the camera to focus so that the image information of the lost item meets the preset clarity requirements.
[0061] In this embodiment, the camera is controlled to focus, ensuring that the image of the lost item meets a preset clarity requirement. Clarity is a crucial indicator of image quality; a clear image provides more detailed information, which is beneficial for subsequent analysis. Adjusting the camera's focus distance ensures that the lost item in the image appears clear, meeting the preset clarity standard.
[0062] In some embodiments, focus adjustment can be based on an autofocus algorithm (such as a hybrid autofocus algorithm). The camera continuously adjusts the lens focus by detecting the sharpness of the image until the image reaches a preset sharpness requirement.
[0063] The criterion for sharpness can be set to ensure that the sharpness does not decrease within N frames when rotating and acquiring the image of the object. The sharpness can be measured using the Tenengrad gradient function, and the specific calculation method is as follows: Where M is the number of pixel columns in the image information of the lost item, and N is the number of pixel rows in the image information of the lost item. To represent the convolution result of pixels in the horizontal direction with the Sobel operator, This represents the convolution result of pixels in the vertical direction with the Sobel operator.
[0064] Step S02: Analyze the item information based on the preset large model to form descriptive information about the lost items.
[0065] Simply collecting item information cannot be directly used for retrieval and matching; it needs to be transformed into standardized descriptive information. However, traditional manual descriptions are easily influenced by subjective factors, resulting in inconsistencies and hindering efficient retrieval.
[0066] Therefore, in this embodiment, the item information is analyzed based on a pre-set large model to generate descriptive information for the lost item. This descriptive information is the extraction and summary of item characteristics, such as color, material, shape, brand logo, and special markings. This descriptive information can more accurately represent the item's characteristics, aiding in subsequent intelligent image and text retrieval and matching. Using a large model for analysis avoids subjective bias, extracts more objective and comprehensive item characteristics, and generates more accurate descriptive information, improving recognition accuracy and retrieval efficiency.
[0067] In some embodiments, item information may include image information. Based on this, to achieve a precise description of the appearance attributes of the lost item, improve the expressiveness and accuracy of the descriptive information, reduce the subjectivity of manual description, and enhance the consistency and standardization of information, this can be achieved by combining visual features extracted by a visual encoder with a pre-set large model. That is, as mentioned in step S02, analyzing item information based on a pre-set large model to form descriptive information about the lost item may specifically include the following steps: Step S21: Extract visual features of image information using a preset visual encoder.
[0068] Since relying solely on raw image data makes efficient analysis and recognition difficult, it is necessary to convert image information into numerical features that are easy for computers to process. Therefore, in this embodiment, a pre-trained visual encoder (e.g., Vision Transformer, ViT) is used to analyze the acquired image information and convert it into a series of numerical visual features. These visual features include information such as the position, shape, color, and texture of objects in the image, which serve as the foundational data for subsequent large-scale model analysis and provide key information for describing the appearance attributes of lost items.
[0069] Step S22: Analyze the image information and visual features using a large model to form the first descriptive information.
[0070] The first descriptive information is used to describe the appearance attributes of the lost item.
[0071] In this embodiment, a first descriptive information is formed by comprehensively analyzing the visual features and image information extracted by the visual encoder using a large model. This first descriptive information may include, but is not limited to, dominant color, surface material (such as metal, fabric, or plastic), shape and outline, prominent markings (such as a brand logo), unique scratches, or decorations. For example, for a smartphone, it can not only describe its color and material, but also identify complex spatial relationships such as "the camera module is located in a matrix arrangement in the upper left corner of the back panel," and infer detailed information such as "slight scratches exist on the screen edge, possibly indicating signs of use." This process effectively solves the problems of strong subjectivity and inconsistent dimensions in traditional methods of describing items, providing standardized, high-information-density textual evidence for subsequent retrieval.
[0072] In some embodiments, to expand the dimensions of lost item information management and provide more comprehensive clues for identifying and recovering lost items, the collection and analysis of audio recording information can be introduced. Specifically, the collection of item information for lost items mentioned in step S01 can include: Audio recordings related to the retrieval process of lost items were collected using a microphone.
[0073] In this embodiment, collecting audio recordings related to the process of retrieving lost items via microphone means converting the voice information that occurs when retrieving the lost item into digital audio data using a microphone. The audio recordings contain key information such as the finder's description of the lost item, the location of the retrieval, and the time of retrieval, providing a more comprehensive picture of the lost item's condition.
[0074] Based on this, step S02, which involves analyzing the item information using a pre-defined large model to generate descriptive information about the lost item, may specifically include the following steps: Step S31: Convert the audio recording information into audio features using a preset audio encoder.
[0075] Since raw audio data is difficult to analyze and identify directly, it needs to be converted into numerical features that are easy for computers to process. Therefore, in this embodiment, a pre-trained audio encoder is used to analyze the acquired audio recordings, converting the audio information into a series of numerical audio features. These audio features include information such as frequency, pitch, and rhythm, and serve as the foundational data for subsequent large-scale model analysis.
[0076] Step S32: Analyze the audio features using a large model to form second descriptive information.
[0077] The second descriptive information describes the process of retrieving the lost item.
[0078] In this embodiment, a pre-defined large model is used to analyze the audio features extracted by the audio encoder to form second descriptive information. The large model can associate audio features with contextual information, perform semantic understanding, and form second descriptive information. The second descriptive information may include key information such as the pickup location, time, and description of the pickup personnel, and uses natural language to describe the details of the pickup process.
[0079] In some embodiments, the large model can perform error correction and de-colloquialization processing on the initial text collected by the microphone (such as eliminating pause words like "uh" and "um"), and then perform Named Entity Recognition (NER) and relation extraction based on the understanding of the context. Taking a lost item scenario at a railway station as an example, from a jumbled dialogue description by the person picking up the item, "a black laptop bag with the brand name XX" can be accurately identified as the item type in the second description information, and "left on a massage chair in the waiting room on the second floor around 3 pm" can be identified as the picking time and picking location in the second description information.
[0080] In some embodiments, to improve the efficiency and accuracy of lost item management and avoid errors and subjective biases from manual entry, paper records can be scanned and analyzed using a large model to digitize the information. Specifically, as mentioned in step S01, collecting information about the lost items can include: Scan the paper record sheet to create a text record of the lost item information.
[0081] In this embodiment, the paper record sheet is scanned to generate a text record of the lost item. This text record contains key information from the retrieval process, such as the item number, the person who retrieved it, the location of retrieval, and the registration time. Paper record sheets are a commonly used information carrier in lost and found management; digitizing them directly avoids errors and delays caused by manual entry and improves information entry efficiency.
[0082] Based on this, as mentioned in step S02, the analysis of item information based on a preset large model to form descriptive information about the lost items can specifically include: The text record information is analyzed using the large model to form third descriptive information.
[0083] In this embodiment, a pre-defined large model is used to perform semantic analysis on the text record information, extract key information, and generate third descriptive information. The third descriptive information may include information about the person who picked up the item, a description of the item, the location where it was picked up, and the registration time.
[0084] In some embodiments, the large model can receive text record information as input, utilize its internal neural network structure to analyze the text information, identify key entities (e.g., picker, item name, location) and relationships (e.g., pick-up time, registration time), and transform the extracted information into a structured text description. For example, the large model can identify "Picker: Mr. Zhang, Item: Black wallet, Pick-up location: Waiting room".
[0085] Step S03: Generate and store a structured file based on the item information and description information.
[0086] In this embodiment, the collected item information and the descriptive information generated through large-scale model analysis are integrated to generate a structured data file according to a predetermined format (e.g., JSON format), and this file is stored in the system. Linking the descriptive information with the item information and storing it in a structured form facilitates subsequent retrieval and management, providing data support for the entire lost and found management system.
[0087] In some embodiments, a large model can be used to integrate item information and description information into a structured file in JSON format. For example, the JSON file may contain a unique identifier for the item, a visual description, an audio summary, supplementary information, and an index associated with multi-view images.
[0088] Based on the above technical solution, the lost item information management method provided in this embodiment of the invention collects item information of lost items; analyzes the item information based on a preset large model to form descriptive information of the lost items; and generates and stores a structured file based on the item information and descriptive information. This lost item information management method, by introducing a preset large model to analyze the collected lost item information, can generate more accurate and richer descriptive information, effectively improving the quality and standardization of lost item information. Compared with the traditional method relying on manual description, this method can reduce subjective bias, achieve a deep understanding of item characteristics, and lay a more reliable foundation for subsequent intelligent retrieval and matching.
[0089] For example, at a certain train station, at 10:30 AM on a certain day, staff found a dark blue canvas backpack on the luggage rack in carriage 3 of train G102. The person who found it, Mr. Zhang (contact number: [phone number missing]), was notified. Hand the backpack to the lost and found and verbally describe: "Um, I found a backpack on the luggage rack in the third carriage of the G102 train. It is about 30×20×15 cm in size. After we opened it as required, we found a laptop and that's about it inside. We need to register it."
[0090] (1) Information Collection Stage: The staff places the backpack on the transparent rotating platform of the image acquisition component 110. After the device is started, the item information is collected through the following steps: Autofocus and zoom: The system uses the YOLOv8 target detection algorithm to identify the position of the backpack in the image, outputs a bounding box to determine the backpack area, and automatically adjusts the focus so that the backpack occupies 40% of the image size; Multi-view image acquisition: The transparent rotating platform rotates at a speed of 30 degrees / second, and high-definition cameras at the top, side and bottom simultaneously acquire images from a total of 6 angles (front, left side, right side, back, top and bottom). Background uniformity processing: A pure white background ensures that all images have a consistent background, avoiding cluttered backgrounds that could affect subsequent analysis; Assisted information entry: The staff recorded Mr. Zhang's voice description through the microphone: "Well, I found a backpack on the luggage rack in the third carriage of the G102 train. The size is about 30×20×15. After we opened it as required, we found a laptop computer inside. That's about it. We need to register it."
[0091] (2) Information processing stage Large-scale models are used for data collection and analysis.
[0092] Image content analysis: The visual encoder extracts depth features from multi-view images; the large language model generates first descriptive information, such as "dark blue canvas backpack, cuboid shape, with a frosted texture on the surface, the xxx brand logo on the zipper, and a scratch about 3 cm long in the lower right corner"; the internal outline of the backpack is identified, and it is inferred that it may contain a laptop and documents.
[0093] The second descriptive information extraction: The audio encoder processes the audio recording information and converts it into feature vectors; the large language model performs de-colloquialization processing to eliminate pause words such as "uh" and "um"; after key information extraction (NER) and information summarization, the lost location, item size and internal details information such as "luggage rack in carriage 3 of train G102", "30×20×15", and "one laptop computer in the bag" are obtained.
[0094] Third, information extraction: Text information from the passenger handover form is extracted using OCR technology, and semantic understanding is performed using a large language model. Key fields are structured, including pickup time (xxxx-xx-xxT10:30:00), train number (G102), seat number (12A), pickup location (luggage rack in carriage 3), and registration information (Mr. Zhang). ).
[0095] (3) Information modeling stage for lost items The system integrates the processed multimodal information into a standardized JSON format file, the details of which are as follows: { "$schema": ". / schemas / item-schema.json", "item_id": "ITEM-xxxxxxxx-001", "visual_description": { "color": "dark blue", "material": "canvas" "shape": "rectangular prism", "dimensions": { "length": 30, "width": 20, "height": 15 }, "texture": "frosted", "brand": "xxx", "special_marks": ["There is a scratch about 3 cm long in the bottom right corner", "The zipper has the 'xxx' brand logo"] }, "audio_summary": { "key_points": "A backpack, approximately 30×20×15 cm in size, was found on the luggage rack in carriage 3 of train G102. A laptop computer was found inside the backpack." "audio_source": "Live recording", "timestamp": 24.5 }, "document_info": { "pickup_time": "2025-12-22T10:30:00+08:00", "train_number": "G102", "seat_number": "12A", "pickup_location": "Luggage rack in carriage 3", "reporter_info": { "name": "Zhang" ", "contact": " " } }, "image_references": [ { "image_id": "IMG-20251222-001-A", "view_angle": "Front view", "url": "https: / / example.com / images / ITEM-20251222-001-front.jpg", "upload_time": "2025-12-22T11:00:00+08:00" }, { "image_id": "IMG-20251222-001-B", "view_angle": "side view", "url": "https: / / example.com / images / ITEM-20251222-001-side.jpg", "upload_time": "2025-12-22T11:00:00+08:00" }, { "image_id": "IMG-20251222-001-C", "view_angle": "top", "url": "https: / / example.com / images / ITEM-20251222-001-top.jpg", "upload_time": "2025-12-22T11:00:00+08:00" }, { "image_id": "IMG-20251222-001-D", "view_angle": "bottom", "url": "https: / / example.com / images / ITEM-20251222-001-bottom.jpg", "upload_time": "2025-12-22T11:00:00+08:00" }, { "image_id": "IMG-20251222-001-E", "view_angle": "Label close-up", "url": "https: / / example.com / images / ITEM-20251222-001-logo.jpg", "upload_time": "2025-12-22T11:00:00+08:00" }, { "image_id": "IMG-20251222-001-F", "view_angle": "Close-up of the damage", "url": "https: / / example.com / images / ITEM-20251222-001-scratch.jpg", "upload_time": "2025-12-22T11:00:00+08:00" } ], "metadata": { "version": "1.0", "last_updated": "xxxx-xx-xxT12:00:00+08:00", "data_source": "Lost and Found System V3" } } This embodiment provides an electronic device, including a processor and a memory. The memory is used to store at least one instruction. When the instruction is loaded and executed by the processor, it implements the above-mentioned method for managing information on lost items. Its execution method and beneficial effects are similar and will not be described again here.
[0096] It should be noted that although the steps are described in a specific order above, it does not mean that the steps must be executed in the above specific order. In fact, some of these steps can be executed concurrently, or even in a different order, as long as the required function can be achieved.
[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A lost item information management device, characterized in that, It includes an information acquisition module, an information processing module, and a storage module, among which: The information collection module is connected to the information processing module and is used to collect the item information of the lost item and input the item information into the information processing module; The information processing module is used to process the item information and generate a structured file; The storage module is connected to the information processing module and is used to store the structured file.
2. The apparatus according to claim 1, characterized in that, The information acquisition module includes an image acquisition component, which includes: The shooting compartment has a backdrop on its inner wall; A shelf located inside the shooting compartment is used to place the lost items; A first camera, located on the top side of the shooting compartment, is used to capture first image information of the lost item.
3. The apparatus according to claim 2, characterized in that, The shelf is a transparent, rotatable shelf; The image acquisition component also includes: A second camera, located at the bottom of the shooting compartment, is used to capture second image information of the lost item; Lighting equipment located at the bottom of the shooting chamber.
4. The apparatus according to claim 1, characterized in that, The information acquisition module further includes an auxiliary information acquisition component, which includes: A text recognition scanner is used to collect text records related to the process of picking up the lost item; A microphone is used to collect audio recordings related to the retrieval process of the lost item.
5. A method for managing information on lost items, characterized in that, Applied to the apparatus of any one of claims 1-4, the method comprises: Collect information about lost items; The information of the lost items is analyzed based on a pre-set large model to form descriptive information about the lost items; A structured file is generated and stored based on the item information and the description information.
6. The method according to claim 5, characterized in that, The information collected on lost items includes: The camera captures image information of the lost items; The proportion of the lost item in the image is determined based on the image information; Control the camera to zoom and adjust so that the proportion of the lost item in the image meets the preset proportion requirements; Control the camera to focus so that the image information of the lost item meets the preset clarity requirements.
7. The method according to claim 6, characterized in that, The analysis of the item information based on a preset large model generates descriptive information about the lost item, including: Visual features of the image information are extracted using a preset visual encoder; The image information and visual features are analyzed using the large model to form first descriptive information; the first descriptive information is used to describe the appearance attributes of the lost item.
8. The method according to claim 5, characterized in that, The information collected on lost items includes: Audio recordings related to the retrieval process of the lost item were collected using a microphone; The analysis of the item information based on a preset large model generates descriptive information about the lost item, including: The audio recording information is converted into audio features using a preset audio encoder; The audio features are analyzed using the large model to form second descriptive information; the second descriptive information is used to describe the process of picking up lost items.
9. The method according to claim 5, characterized in that, The information collected on lost items includes: Scan the paper record sheet to generate a text record of the lost items; The analysis of the item information based on a preset large model generates descriptive information about the lost item, including: The text record information is analyzed using the large model to form third descriptive information.
10. An electronic device, characterized in that, include: A processor and a memory, the memory being used to store at least one instruction, which, when loaded and executed by the processor, implements the information management method for lost items as described in any one of claims 5-9.