Intelligent archive shelving guiding method and device, electronic equipment and storage medium

Abstract

The invention provides an intelligent archive shelving guiding method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining a target archiving position of a to-be-archived archive and an initial pose output by a user through fusion of a camera on augmented reality equipment and an inertial sensor; generating an initial navigation path based on the target archiving position, the initial pose and a pre-constructed archive three-dimensional map, and rendering and superposing the initial navigation path to a real picture currently shot by the camera in a virtual arrow sequence form; when the user advances along the initial navigation path, continuously tracking the current real-time pose of the user, and calculating the deviation distance between the position of the current real-time pose and the initial navigation path; and when the deviation distance is greater than a preset deviation range, taking the position of the current real-time pose as a new starting point, re-planning a navigation path to the target archiving position, and updating the virtual arrow sequence. According to the invention, while the racking efficiency is improved, the accuracy of the whole guiding process is enhanced.

Description

Technical Field

[0001] This invention relates to the field of warehousing technology, and more specifically, to a method, apparatus, electronic device, and storage medium for intelligent shelving guidance of archives. Background Technology

[0002] In large archives, libraries, and corporate archives centers, a large number of physical files need to be accurately placed into designated cabinets and storage locations in the storage room every day. Archive storage rooms are typically vast and complex, with thousands of cabinets and storage locations arranged densely. Staff need to locate the corresponding storage location based on the file's management number and then accurately shelve the file. This process is an extremely important yet error-prone step in archive management.

[0003] In existing technologies, manual shelving is inefficient and prone to errors; while semi-automated methods, such as installing LED indicator lights in each storage location to guide users by illuminating the target location, can accurately locate the final storage location, but cannot provide path guidance from the current location to the target shelf. In large warehouses, just finding the shelf can take a significant amount of time. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method, device, electronic device and storage medium for intelligent file shelving guidance, so as to provide an intelligent guidance path and improve the efficiency and accuracy of file shelving operation.

[0005] Firstly, a method for guiding intelligent file shelving is provided, including: Obtain the target archiving location of the archive to be archived, as well as the initial pose output by the user through the fusion of the camera and inertial sensors on the augmented reality device; Based on the target archiving location, initial pose, and pre-built 3D map of the archive, an initial navigation path is generated and rendered and superimposed onto the real image currently captured by the camera in the form of a virtual arrow sequence, wherein the virtual arrow sequence extends along the ground in the real image; As the user moves along the initial navigation path, the system continuously tracks the user's current real-time pose and calculates the deviation distance between the current real-time pose and the initial navigation path. When the deviation distance exceeds the preset deviation range, the current real-time pose position is used as the new starting point to replan the navigation path to the target archiving position and update the virtual arrow sequence.

[0006] Optionally, the target archiving location for the archive to be archived includes: The identification code of the archive to be archived is scanned using a camera on an augmented reality device to identify its archive ID; Query the metadata corresponding to the file ID based on the file ID.

[0007] Read the real-time storage space occupancy status table of the archive warehouse to determine all available storage spaces and their physical attributes; Based on the archive ID's metadata, all available storage spaces, and their physical attributes, a pre-defined storage optimization algorithm is used to determine the target archiving location.

[0008] Optionally, metadata includes file type, historical borrowing frequency, security level, and project association identifier; physical attributes include storage space size, security control level of the area, and horizontal distance from the main passageway; using a preset storage optimization algorithm, the target archiving location is determined, including: From all available storage spaces, select those with physical dimensions not less than the preset volume threshold for the file type and whose area security control level is not lower than the security level of the file ID to form an initial selection set; If the historical borrowing frequency is greater than or equal to 5 times / month, then select the positions that are less than or equal to 8 meters away from the main channel in the initial selection set as the hot zone candidate set; otherwise, retain all the initial selection positions as the cold zone candidate set. Check if there are any existing associated files with the same association identifier as the project. If they exist, select the storage space in the same filing cabinet or adjacent cabinet group as the existing associated file from the hot zone candidate set or cold zone candidate set to obtain the candidate storage space. Candidate storage locations are sorted by the following priority: first priority is whether they meet the requirements for associated storage; second priority is the horizontal distance from the main channel from near to far; and third priority is the cabinet layer number from low to high. Select the first location in the sorting results and use its structured address as the target archiving location; the structured address includes the region, filing cabinet number, layer number, and location number.

[0009] Optionally, the user's initial pose and current real-time pose are determined using the following methods: Acquire image information captured by the camera and sensor data collected by the inertial sensor; Based on image information and sensor data, the user's pose is determined using a visual inertial odometry method.

[0010] Optionally, when continuously tracking the user's current real-time pose, the method further includes: The system periodically scans the beacons deployed in the archive to obtain the signal strength value and beacon ID of each beacon. Based on the preset transmit power and received signal strength value of each beacon, the system calculates the distance from the user to each beacon using a logarithmic distance path loss model. Based on the distance from the user to each beacon, the system uses triangulation to determine the user's absolute position in the archive. The absolute position is then used as an external observation input to an extended Kalman filter. The extended Kalman filter fuses the absolute position with the pose calculated by the visual inertial odometry method, and outputs the corrected current real-time pose.

[0011] Optionally, based on the target archiving location, initial pose, and pre-built 3D map of the archive, the initial navigation path is generated, including: Extract the user's current position coordinates from the initial pose, and parse the physical coordinates of the target filing cabinet from the target archiving location; The 3D map of the archive is projected into a 2D raster map of passable areas, where each raster is marked as passable or obstructed. Obstructions include the archive cabinet itself, pillars, and restricted areas. On the grid map, starting from the user's current location and ending at the center point of the bottom of the target filing cabinet, the path planner is invoked. A cost function that takes into account turning penalties and path smoothness is used to generate the shortest feasible path. Spline interpolation is performed on discrete landmark points on the shortest feasible path to generate a continuous and smooth sequence of 3D path points; the Z coordinates of all path points are constrained to a preset distance from the ground to ensure that the virtual arrow sequence is rendered along the real ground. Output a sequence of 3D path points as the initial navigation path.

[0012] Optionally, the target archiving location includes the target filing cabinet and its location number; when the distance between the user's current location and the target archiving location is less than a preset distance, the method further includes: Stop rendering the virtual arrow sequence and render the target filing cabinet as a virtual frame overlaid on the real image of the augmented reality device; scan the preset marks on each compartment of the target filing cabinet based on the camera; bind each mark to the corresponding compartment number; when a mark matching the compartment number of the target filing location is scanned, obtain the compartment size of the corresponding compartment number; based on the compartment size, generate a virtual local highlight frame that matches the compartment size, and render the local highlight frame onto the real image currently captured by the camera.

[0013] Secondly, a smart archive shelving guidance device is provided, comprising: The acquisition unit is used to acquire the target archiving location of the archive to be archived, as well as the initial pose output by the user through the fusion of the camera and inertial sensor on the augmented reality device; The generation unit is used to generate an initial navigation path based on the target archiving location, initial pose, and pre-built 3D map of the archive, and to render and overlay the initial navigation path as a sequence of virtual arrows onto the real image currently captured by the camera, wherein the sequence of virtual arrows extends along the ground in the real image; The tracking unit is used to continuously track the user's current real-time pose as the user moves along the initial navigation path, and to calculate the deviation distance between the current real-time pose and the initial navigation path. The update unit is used to replan the navigation path to the target archiving position and update the virtual arrow sequence when the deviation distance is greater than the preset deviation range, taking the current real-time pose position as the new starting point.

[0014] Thirdly, an electronic device is provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements any of the methods of the first aspect.

[0015] Fourthly, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, it implements any of the methods of the first aspect.

[0016] This invention provides a method, device, electronic device, and storage medium for intelligent archive shelving guidance. It acquires the target archiving location of the archive to be archived, and the initial pose of the user output via a fusion of a camera and inertial sensor on an augmented reality device. Based on the target archiving location, the initial pose, and a pre-constructed 3D map of the archive, an initial navigation path is generated and rendered as a sequence of virtual arrows overlaid on the real-world image captured by the camera. As the user moves along the initial navigation path, the system continuously tracks the user's current real-time pose and calculates the deviation distance between the current real-time pose and the initial navigation path. When the deviation distance exceeds a preset range, the system uses the current real-time pose as a new starting point to replan the navigation path to the target archiving location and updates the virtual arrow sequence. This invention effectively overcomes the limitations of existing technologies that lack path navigation, enabling staff to quickly traverse complex warehouse environments and reach the target area without relying on memory or paper indexes. It not only significantly shortens the wayfinding time, but also greatly reduces operational errors caused by going to the wrong channel or accidentally entering adjacent areas through real-time pose tracking and dynamic path correction mechanisms. This improves the efficiency of product listing while enhancing the accuracy of the entire guidance process and the user experience.

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The flowchart illustrates a method for guiding intelligent file shelving provided by an embodiment of the present invention. Figure 2 This diagram illustrates the structure of an intelligent archive shelving guidance device provided in an embodiment of the present invention. Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present invention is shown. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the 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.

[0021] This invention provides a method for guiding intelligent shelving of archives, applied to an intelligent archive shelving system. The system includes a backend server and an augmented reality device carried by the user, wherein the augmented reality device is equipped with an AR client application for performing archive shelving guidance tasks.

[0022] Augmented reality devices can be smartphones, tablets, or AR glasses that support ARKit or ARCore. They have built-in cameras to capture environmental images and inertial measurement units (IMUs) that include three-axis accelerometers and gyroscopes to acquire environmental images and motion data in real time and interact with the backend server.

[0023] The following is a detailed description through examples, such as... Figure 1 As shown, it includes the following steps: Step S101: Obtain the target archiving location of the archive to be archived, and the initial pose output by the user through the fusion of the camera and inertial sensor on the augmented reality device.

[0024] In this embodiment of the invention, the initial pose refers to the user's six-degree-of-freedom pose in the global coordinate system of the warehouse, including three-dimensional position. And the user's orientation.

[0025] User gestures are represented by unit quaternions. This is an efficient and singularity-free way to describe a user's orientation in three-dimensional space.

[0026] in, The real part, the rest , , The imaginary part, together, encodes the equivalent rotation about a unit axis by a specific angle. Specifically, if the axis of rotation is... User's orientation angle ,but ,and , , .

[0027] This quaternion satisfies the normalization condition. .

[0028] Specifically, the pose is obtained using the visual-inertial odometry (VIO) method. This includes the following steps: Step A: Acquire image information captured by the camera and sensor data collected by the inertial sensor.

[0029] Step B: Based on image information and sensor data, determine the user's pose using the visual inertial odometry method.

[0030] The specific steps of the VIO method are as follows: based on image feature points and IMU pre-integration constraints, a sliding window nonlinear optimization framework is used to jointly minimize the visual reprojection error and the IMU motion consistency error, and output a high-precision pose.

[0031] Among them, IMU pre-integration compresses the high-frequency IMU measurement values ​​(200 Hz) between two keyframes into relative rotation ΔR, velocity Δv, and position Δp increments, avoiding repeated integration; A sliding window maintains 10 keyframes, and through Ceres Solver optimization, 60 Hz real-time pose output is achieved.

[0032] For example, when a user launches the application and aligns it with the A-zone passage, ORB feature points are matched from the pre-built archive map to calculate the initial pose as: position (5.21 m, 3.78 m, 1.62 m) and orientation yaw angle 22.5°.

[0033] Meanwhile, the target archiving location is dynamically assigned by the backend server based on the archive ID, in the form of a structured address. The structured address includes the area, archive cabinet number, layer number, and storage location number, such as {"area": ​​"A", "cabinet": "C035", "level": 4, "slot": 2}.

[0034] This step provides the starting and ending points for subsequent path planning and AR rendering, ensuring that the guidance task has a clear objective. By integrating VIO with pre-built maps, a reliable initial pose can be quickly obtained even in indoor environments without GPS, significantly improving the startup efficiency and spatial consistency of AR guidance.

[0035] Step S102: Based on the target archiving location, initial pose, and pre-built 3D map of the archive, generate an initial navigation path and render and overlay the initial navigation path as a sequence of virtual arrows onto the real image currently captured by the camera.

[0036] The sequence of virtual arrows extends along the ground in the real scene.

[0037] Following the previous example, the pre-built 3D map of the archive is a point cloud model generated by LiDAR scanning during the deployment phase and converted into a semantically labeled 3D mesh, containing all filing cabinets, pillars, and passageway areas. The AR client loads this map locally for path planning and spatial anchoring.

[0038] For example, the virtual arrow sequence consists of semi-transparent blue 3D arrows placed every 1.2 meters along the ground, pointing towards the next path point.

[0039] In this embodiment of the invention, the AR application calls an AR engine (such as ARCore), which uses plane detection and world tracking technology to anchor the arrow to real ground coordinates, so that the arrow remains stable and close to the ground even if the user shakes the device. For example, when the user is at (5.2, 3.8), the first arrow appears at (5.3, 3.85) with a height Z=0.1 m, ensuring that it appears to grow on the ground.

[0040] This design significantly improves the intuitiveness of directional guidance, avoiding misjudgments caused by traditional 2D arrows floating on the screen.

[0041] Step S103: During the user's movement along the initial navigation path, continuously track the user's current real-time pose and calculate the deviation distance between the current real-time pose and the initial navigation path.

[0042] In this embodiment of the invention, the current real-time pose is updated at a frequency of no less than 30 Hz, and feature points are continuously tracked by the VIO method and fused with IMU data for output. The deviation distance refers to the shortest Euclidean distance from the user's current position to the initial navigation path.

[0043] In the specific implementation, the path is discretized into a sequence of landmark points. For the current pose position ,calculate As the deviation distance.

[0044] For example, if the waypoint is (6.0, 3.5) and the user's current location is (5.7, 4.1), then the deviation distance is approximately 0.67 meters.

[0045] This monitoring mechanism can promptly detect whether users are taking detours, going down the wrong path, or being blocked by obstacles, providing triggering conditions for dynamic replanning.

[0046] Step S104: When the deviation distance is greater than the preset deviation range, take the current real-time pose position as the new starting point, replan the navigation path to the target archiving position, and update the virtual arrow sequence.

[0047] In this embodiment of the invention, the typical value of the preset deviation range can be set to 0.8 meters. This threshold is set according to the width of the warehouse passage (usually 1.2–1.5 meters) to ensure that replanning is triggered once the user crosses the center line.

[0048] Following the previous example, when the deviation exceeds 0.8 meters, the A* algorithm is immediately invoked, taking the current real-time pose position as the new starting point, with the target still being the center of the bottom of cabinet C035, and the path is regenerated.

[0049] After spline interpolation, the AR engine seamlessly switches the arrow sequence, with old arrows fading out and new arrows appearing along the corrected route. This mechanism effectively addresses non-ideal behavior during actual walking, preventing AR guidance failure and ensuring robust guidance. The entire closed-loop process continues until the user reaches the vicinity of the target location.

[0050] This invention effectively overcomes the limitations of existing technologies that lack path navigation, enabling staff to quickly traverse complex warehouse environments and reach their target areas without relying on memory or paper indexes. It not only significantly shortens wayfinding time but also greatly reduces operational errors caused by going to the wrong passage or entering adjacent areas through real-time pose tracking and dynamic path correction mechanisms. This improves shelving efficiency while enhancing the accuracy and user experience of the entire guidance process.

[0051] Based on the above embodiments, obtaining the target archiving location of the archive to be archived includes: Step S101A: Scan the identification code of the archive to be archived using the camera on the augmented reality device to identify its archive ID.

[0052] For example, the identification code is a one-dimensional Code128 barcode or QR code, printed on the document cover. The AR client calls the device's camera to scan in real time, decodes it through the ZXing library, and obtains a unique document ID, such as "ARCH_2025_METRO5_CONTRACT".

[0053] Step S101B: Query the metadata corresponding to the file ID based on the file ID.

[0054] In this embodiment of the invention, the AR client requests the archive metadata database of the backend server via HTTPS, returning metadata in JSON format, including: archive type (e.g., construction contract), historical borrowing frequency (e.g., 12 times / year), security level (e.g., general), and project association identifier (e.g., METRO_LINE5_2025). This metadata is the core input for intelligent storage allocation.

[0055] Step S101C: Read the real-time storage space occupancy status table of the archive warehouse to determine all available storage spaces and their physical attributes.

[0056] Continuing from the previous example, the occupancy status table is maintained by the backend and updated every 5 seconds. Each occupancy location records its structured address and physical attributes: occupancy dimensions (length × width × height, in cm), security control level of the area (e.g., Area A: general; Area S: confidential), and horizontal distance from the main aisle (e.g., 6.2 meters for container C035).

[0057] A vacant storage space refers to a storage space that is currently unassigned and has no RFID occupancy signal.

[0058] Step S101D: Based on the metadata of the archive ID, all available storage spaces and their physical attributes, determine the target archive location using a preset storage optimization algorithm.

[0059] In one feasible implementation, determining the target archiving location using a preset warehousing optimization algorithm includes: Step S101D1: Among all available storage spaces, select those with physical dimensions not less than the preset volume threshold for the file type and whose area security control level is not lower than the security level of the file ID to form a preliminary selection set.

[0060] For example, the construction contract has a preset volume threshold of 30×25×5 cm³. If a certain storage space is only 28 cm wide, it will be filtered out; if the file security level is confidential, only the S area storage space will be retained.

[0061] Step S101D2: If the historical borrowing frequency is greater than or equal to 5 times / month, select the positions in the initial selection set that are less than or equal to 8 meters away from the main channel as the hot zone candidate set; otherwise, retain all the initial selection positions as the cold zone candidate set.

[0062] In this embodiment of the invention, 5 times / month is the high-frequency threshold, and 8 meters is the hot zone boundary, ensuring that frequently used files are easy to retrieve and store. For example, files with a borrowing frequency of 12 times / year (i.e., 1 time / month) are considered for the cold zone.

[0063] Step S101D3: Check if there are any existing associated files with the same association identifier as the project. If so, select the storage space in the hot zone candidate set or cold zone candidate set that is located in the same filing cabinet or adjacent cabinet group as the existing associated file to obtain the candidate storage space.

[0064] Continuing from the previous example, if "METRO_LINE5_2025" already has a file in cabinet C035, then vacant slots in C035, C034, or C036 will be allocated first. Cabinet groups are defined as those with a difference of cabinet number ≤ 2.

[0065] Step S101D4: Sort the candidate storage locations according to the following priorities: the first priority is whether they meet the requirements for associated storage, the second priority is the horizontal distance from the main channel from near to far, and the third priority is the cabinet layer number from low to high.

[0066] This multi-level sorting prioritizes relevance, followed by convenience, and finally ergonomics (lower levels are easier to operate).

[0067] Step S101D5: Select the first location in the sorting results and use its structured address as the target archive location.

[0068] The final output of this step is as follows: {"area": ​​"A", "cabinet": "C035", "level": 4, "slot":2}. This address is directly used for subsequent AR highlighting and marker matching.

[0069] This embodiment achieves intelligent mapping from file attributes to physical storage locations. This solution not only meets hard constraints such as security isolation and size adaptation, but also proactively optimizes storage layout to improve the efficiency of accessing high-frequency files and the aggregation of related files, effectively reducing later management costs and the risk of misplacement.

[0070] Based on the above embodiments, when continuously tracking the user's current real-time pose, the method further includes: Step C: Periodically scan the beacons deployed in the archive to obtain the signal strength value and beacon ID of each beacon.

[0071] In this embodiment of the invention, the beacon type is, for example, a Bluetooth beacon, a UWB beacon, etc. For example, a Bluetooth 5.0 iBeacon (such as an Estimote Beacon) is deployed every 6 meters on the warehouse ceiling, with a transmission power of... 59 dBm. The AR client initiates a BLE scan every second to obtain the RSSI (Received Signal Strength Indicator) value and UUID of surrounding beacons. Step D: Based on the preset transmit power and received signal strength value of each beacon, the distance from the user to each beacon is calculated using a logarithmic distance path loss model.

[0072] In this embodiment of the invention, a standard model is adopted:

[0073] in, This is a 1-meter reference value; For example, the warehouse environment attenuation factor; , , Then the calculation yields Step E: Based on the user's distance to each beacon, use triangulation to determine the user's absolute location in the archive.

[0074] Continuing from the previous example, when ≥3 beacons are detected, a trilateration model is adopted based on the distance from the user to each beacon, and the 2D position of the user in the archive is solved by nonlinear least squares method (z is fixed at 1.6 m).

[0075] If there are only two beacons, the position is estimated by combining the VIO heading. Typical positioning accuracy: Bluetooth ±1.2 m, which can be upgraded to ±0.1 m with UWB. Step F: The absolute position is used as the external observation input to the extended Kalman filter. The extended Kalman filter fuses the absolute position with the pose calculated by the visual inertial odometry method, and outputs the corrected current real-time pose.

[0076] The Kalman filtering process mainly includes a prediction stage and an update stage; During the prediction phase, VIO provides prior state estimates and their covariance. Because VIO relies on visual feature tracking and IMU integration, drift accumulates during long walks or in areas with weak texture. This leads to a gradual increase in covariance, reflecting an increase in uncertainty. The update phase introduces the two-dimensional absolute position calculated by the beacon system as an external observation. This observation has global consistency but a low update frequency (approximately 1 Hz), and its measurement noise covariance is denoted as... . Finally, the fused corrected pose is updated according to the following formula:

[0077] in, The observation matrix; The pose calculated using VIO technology; The absolute position calculated using beacons; This represents the Kalman filter gain.

[0078]

[0079] in, This refers to the sub-blocks corresponding to the positional components in the state covariance matrix; this gain essentially achieves intelligent weighting based on uncertainty: when the VIO drift is small, When VIO values ​​approach zero, the system trusts VIO more and avoids high-frequency jitter; when VIO drift accumulates, As the value approaches 1, the system significantly adopts beacon observations to bring the global position back to the correct level.

[0080] Through the above embodiments, a multi-source fusion positioning architecture combining high-frequency VIO and low-frequency beacons was constructed. This mechanism effectively suppresses long-term cumulative drift of VIO while maintaining a smooth 60 Hz output, enabling the AR virtual arrow to stably conform to the real ground along paths exceeding 50 meters, thus providing a reliable pose foundation for high-precision guidance.

[0081] Based on the above embodiments, the initial navigation path is generated based on the target archiving location, initial pose, and pre-built 3D map of the archive, including: Step S102A: Extract the user's current position coordinates from the initial pose, and parse the physical coordinates of the target filing cabinet from the target archiving location.

[0082] In this embodiment of the invention, the physical coordinates of the target filing cabinet are provided by the cabinet-coordinate mapping table maintained by the backend server, such as the bottom center of cabinet C035 being (12.35, 8.92, 0.0).

[0083] Step S102B: Project the 3D map of the archive into a 2D raster map of accessible areas, where each raster is marked as accessible or an obstacle.

[0084] Obstacles include the filing cabinet itself, the uprights, and restricted areas.

[0085] For example, the raster resolution is 0.1 m × 0.1 m, and the filing cabinet occupies multiple consecutive grid cells and is marked as an obstacle. This map is used for efficient pathfinding.

[0086] Step S102C: On the grid map, starting from the user's current location and ending at the center point of the bottom of the target filing cabinet, call the path planner and generate the shortest feasible path by combining a cost function that considers turning penalties and path smoothness.

[0087] In this step, the path planner can use the A* algorithm for path planning, such as A-Star.

[0088] Specifically, the cost function of the A* algorithm is:

[0089] in, The distance already traveled; It is a Euclidean heuristic function. Penalty for the number of turns (λ=0.5); This represents a candidate node during the search process. This design reduces jagged paths and improves walking comfort.

[0090] Step S102D: Perform spline interpolation on the discrete landmarks on the shortest feasible path to generate a continuous and smooth three-dimensional path point sequence.

[0091] In this process, the Z coordinates of all path points are constrained to a preset distance from the ground to ensure that the virtual arrow sequence is rendered along the real ground.

[0092] Following the previous example, the Z-axis of the path points after spline interpolation is uniformly set to 0.1 m (10 cm above the ground) to prevent arrows from hovering or penetrating the ground. This constraint ensures that AR rendering conforms to physical intuition.

[0093] Step S102E: Output the three-dimensional path point sequence as the initial navigation path.

[0094] This sequence directly drives the AR engine to render arrows, creating intuitive guidance.

[0095] The navigation path generated in this embodiment not only avoids obstacles such as filing cabinets, but also optimizes the human experience and AR display effect through turning penalties and Z-coordinate constraints. The generated path combines geometric optimization and visual rationality, providing users with smooth and natural walking guidance.

[0096] Based on the above embodiments, the target archiving location includes the target filing cabinet and its location number; when the distance between the user's current location and the target archiving location is less than a preset distance, the method further includes: Step S105: Stop rendering the virtual arrow sequence and render the target filing cabinet as a virtual frame overlaid on the real screen of the augmented reality device.

[0097] For example, the preset distance is 1.0 meter. When the user approaches cabinet C035, the arrow on the ground fades out, and a semi-transparent cyan cube-shaped highlight frame covers the entire cabinet (2.0 m × 0.4 m × 1.8 m), indicating that the target is here. Step S106: Scan the preset marks on each compartment of the target filing cabinet using the camera; each mark is bound to the corresponding compartment number.

[0098] For example, a 4 cm × 4 cm marker is pasted on the edge of each compartment (e.g., ID=1024 corresponds to level=4, slot=2).

[0099] Step S107: When a marker matching the warehouse number of the target archiving location is scanned, obtain the warehouse size corresponding to the warehouse number of the marker.

[0100] Continuing from the previous example, based on the mapping table of Marker ID-structured address-physical size, after identifying ID=1024, the size is found to be 35 cm (width) × 28 cm (height).

[0101] Step S108: Based on the warehouse size, generate a virtual local highlight frame that matches the warehouse size, and render the local highlight frame onto the real image currently captured by the camera.

[0102] Specifically, the highlighted area is a red semi-transparent 3D rectangle, slightly larger than the actual storage space (+5 cm margin), and is anchored after the 3D pose of the marker is calculated using the PnP algorithm. Users will see that this frame is precisely placed on the 2nd cell of the 4th layer, allowing for direct placement of files.

[0103] Through the above embodiments, two levels of precise guidance are achieved, from "cabinet level" to "grid level". When a user approaches the target area, the guidance mode is automatically switched. First, the entire cabinet is highlighted to establish a macro-position, and then the unique correct storage location is locked through visual marker identification. This completely solves the industry pain point of finding the right cabinet but placing the wrong item in the wrong grid, and improves the accuracy of shelf placement to over 99.9%.

[0104] Based on the above embodiments, after the user arrives at the target storage location and completes the file placement, the method also includes confirmation of the shelving operation and task closure processing: Staff place the files to be archived into the target storage location precisely indicated by a virtual highlighted box. Once placement is complete, confirmation of shelving is triggered using any of the following methods: Manual confirmation mode: Users click the "Listing Completed" button on the AR application interface of the augmented reality device to actively submit an operation completion signal to the system; Automatic confirmation mode: If the target warehouse has an embedded RFID reader, when it detects that the RFID tag ID carried by the file book entering the warehouse matches the file ID of the current task, it will automatically send a successful shelving signal to the AR client without manual intervention.

[0105] Following the previous example, upon receiving any confirmation signal, the AR client immediately generates a task completion record containing the operator's identity (such as login account or device ID), timestamp (accurate to the second), and final physical location (structured address: area, cabinet number, floor number, storage location number), and then transmits the record back to the backend server via the network.

[0106] Upon receiving this information, the backend server executes a database transaction update: changing the status of the corresponding file from "Pending Shelving" to "In Stock," and persistently storing its precise physical location information to ensure accurate location during subsequent retrieval. Simultaneously, the AR client displays a "Successfully Shelved" message on the interface and automatically resets the scanning status, preparing to receive the identification code of the next file to be archived, and enters the next guided loop.

[0107] Based on the same inventive concept, a smart file shelving guidance device is provided, such as... Figure 2 As shown, it includes: The acquisition unit 201 is used to acquire the target archiving location of the archive to be archived, as well as the initial pose output by the user through the fusion of the camera and inertial sensor on the augmented reality device; The generation unit 202 is used to generate an initial navigation path based on the target archiving location, initial pose and pre-built 3D map of the archive, and render and overlay the initial navigation path as a sequence of virtual arrows onto the real image currently captured by the camera, wherein the sequence of virtual arrows extends along the ground in the real image. The tracking unit 203 is used to continuously track the user's current real-time pose while the user is traveling along the initial navigation path, and to calculate the deviation distance between the current real-time pose and the initial navigation path. The update unit 204 is used to replan the navigation path to the target archiving position and update the virtual arrow sequence when the deviation distance is greater than the preset deviation range, taking the current real-time pose position as the new starting point.

[0108] Based on the same technical concept, embodiments of the present invention also provide an electronic device, such as... Figure 3 As shown, it includes a processor 301, a communication interface 302, a memory 303, and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304.

[0109] Memory 303 is used to store computer programs; The processor 301 is used to implement the steps of the intelligent file shelving guidance method when executing the program stored in the memory 303.

[0110] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0111] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0112] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0113] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0114] In another embodiment of the present invention, a computer-readable storage medium is also provided, which stores a computer program. When executed by a processor, the computer program implements the steps of any of the above-described intelligent file shelving guidance methods. Specific implementation details can be found in the method embodiments, and will not be repeated here.

[0115] The intelligent file uploading guidance device provided in this embodiment of the invention can be specific hardware on a device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in this embodiment of the invention are the same as those in the foregoing method embodiments. For the sake of brevity, any parts not mentioned in the device embodiments can be referred to the corresponding content in the foregoing method embodiments. 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 all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0116] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0117] 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.

[0118] In addition, the functional units in the embodiments provided by the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0119] 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 invention, essentially, 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 invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0120] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0121] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for guiding intelligent shelving of archives, characterized in that, include: Obtain the target archiving location of the archive to be archived, as well as the initial pose output by the user through the fusion of the camera and inertial sensors on the augmented reality device; Based on the target archiving location, initial pose, and pre-constructed 3D map of the archive, an initial navigation path is generated, and the initial navigation path is rendered and superimposed onto the real image currently captured by the camera in the form of a virtual arrow sequence, wherein the virtual arrow sequence extends along the ground in the real image; As the user travels along the initial navigation path, the system continuously tracks the user's current real-time pose and calculates the deviation distance between the current real-time pose and the initial navigation path. When the deviation distance is greater than the preset deviation range, the navigation path to the target archiving location is replanned, taking the current real-time pose position as the new starting point, and the virtual arrow sequence is updated.

2. The method according to claim 1, characterized in that, The target archiving location for the archive to be archived includes: The camera on the augmented reality device scans the identification code of the archive to be archived to identify its archive ID; Query the metadata corresponding to the file ID based on the file ID. Read the real-time storage space occupancy status table of the archive warehouse to determine all available storage spaces and their physical attributes; Based on the metadata of the archive ID, all available storage spaces and their physical attributes, a preset storage optimization algorithm is used to determine the target archiving location.

3. The method according to claim 2, characterized in that, The metadata includes file type, historical borrowing frequency, security level, and project association identifier; the physical attributes include storage space size, security control level of the area, and horizontal distance from the main passage; using a preset storage optimization algorithm, the target archiving location is determined as follows: From all available storage spaces, select those with physical dimensions not less than the preset volume threshold of the file type and with a security control level of the area not lower than the security level of the file ID to form an initial selection set; If the historical borrowing frequency is greater than or equal to 5 times / month, then select the warehouses with a horizontal distance of less than or equal to 8 meters from the main channel in the initial selection set as the hot zone candidate set; otherwise, retain all the initial selection warehouses as the cold zone candidate set. Check if there are any existing associated files with the same association identifier as the project. If they exist, select the storage space in the same filing cabinet or adjacent cabinet group as the existing associated file from the hot zone candidate set or cold zone candidate set to obtain the candidate storage space. The candidate storage locations are sorted by the following priority: the first priority is whether they meet the requirements for associated storage; the second priority is the horizontal distance from the main channel from near to far; and the third priority is the cabinet layer number from low to high. Select the first storage location in the sorting results and use its structured address as the target archiving location; the structured address includes the region, filing cabinet number, layer number, and storage location number.

4. The method according to claim 1, characterized in that, The user's initial pose and current real-time pose are determined using the following methods: Acquire image information captured by the camera and sensing data collected by the inertial sensor; Based on the image information and the sensor data, the user's pose is determined using a visual inertial odometry method.

5. The method according to claim 4, characterized in that, When continuously tracking the user's current real-time pose, the method further includes: The system periodically scans the beacons deployed in the archive to obtain the signal strength value and beacon ID of each beacon. Based on the preset transmit power and received signal strength value of each beacon, the system calculates the distance from the user to each beacon using a logarithmic distance path loss model. Based on the distance from the user to each beacon, the system uses triangulation to determine the user's absolute position in the archive. The absolute position is then used as an external observation input to an extended Kalman filter. The extended Kalman filter fuses the absolute position with the pose calculated by the visual inertial odometry method, and outputs the corrected current real-time pose.

6. The method according to claim 1, characterized in that, The generation of the initial navigation path based on the target archiving location, initial pose, and pre-constructed 3D map of the archive includes: Extract the user's current position coordinates from the initial pose, and parse the physical coordinates of the target filing cabinet from the target archiving location; The 3D map of the archive is projected into a 2D raster map of passable areas, where each raster is marked as passable or an obstacle, and the obstacles include the archive cabinet itself, the pillars, and the restricted areas; On the grid map, starting from the user's current location and ending at the center point of the bottom of the target filing cabinet, the path planner is invoked, and a shortest feasible path is generated by combining a cost function that takes into account turning penalties and path smoothness. Spline interpolation is performed on the discrete landmark points on the shortest feasible path to generate a continuous and smooth three-dimensional path point sequence; wherein, the Z coordinates of all path points are constrained to a preset distance from the ground height to ensure that the virtual arrow sequence is rendered along the real ground. The three-dimensional path point sequence is output as the initial navigation path.

7. The method according to claim 1, characterized in that, The target archiving location includes the target filing cabinet and its location number; When the distance between the user's current location and the target archive location is less than a preset distance, the method further includes: Stop rendering the virtual arrow sequence and render the target filing cabinet as a virtual frame overlaid on the real image of the augmented reality device; scan the preset marks on each compartment of the target filing cabinet based on the camera; bind each mark to the corresponding compartment number; when a mark matching the compartment number of the target filing location is scanned, obtain the compartment size of the compartment number corresponding to the mark; based on the compartment size, generate a virtual local highlight frame matching the compartment size, and render the local highlight frame onto the real image currently captured by the camera.

8. An intelligent archive shelving guidance device, characterized in that, include: The acquisition unit is used to acquire the target archiving location of the archive to be archived, as well as the initial pose output by the user through the fusion of the camera and inertial sensor on the augmented reality device; The generation unit is used to generate an initial navigation path based on the target archiving location, initial pose, and pre-built 3D map of the archive, and to render and overlay the initial navigation path as a sequence of virtual arrows onto the real image currently captured by the camera, wherein the sequence of virtual arrows extends along the ground in the real image. The tracking unit is used to continuously track the user's current real-time pose while the user is traveling along the initial navigation path, and to calculate the deviation distance between the current real-time pose and the initial navigation path. The update unit is used to replan the navigation path to the target archiving location and update the virtual arrow sequence when the deviation distance is greater than the preset deviation range, taking the current real-time pose position as the new starting point.

9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; The memory is used to store computer programs; When the processor executes the program stored in the memory, it implements the method described in any one of claims 1-7.

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 method described in any one of claims 1-7.

Images