Cargo stacking management method, system, electronic device and storage medium

Abstract

The present application relates to the field of information technology, and provides a cargo stacking management method, system, electronic device and storage medium. The method comprises: monitoring the deformation of the cargo to determine the geometric shape change of the cargo; updating a stacking geometric model for representing the spatial relationship between the cargos according to the geometric shape change; and adjusting the stacking scheme of the cargos based on the updated stacking geometric model. The present application realizes the leap from static planning to dynamic adaptive management by monitoring the deformation trajectory of the cargos caused by environmental factors in real time and dynamically updating the spatial relationship matrix and the stacking geometric model. The contact surface coordinates are corrected in real time by using the deformation vector field, and the embedded depth is optimized in the mechanical level by combining the particle swarm optimization algorithm, so that when the physical dimension of the cargo shrinks or expands, the re-mapping of the layout grid can be automatically triggered, and the long-term geometric stability and safety of the cargo stacking in a complex environment are improved.

Description

Technical Field

[0001] This invention relates to the field of information technology, and specifically to a cargo stacking management method, system, electronic device, and storage medium. Background Technology

[0002] In modern logistics warehousing systems, stable stacking of goods is a core element in ensuring operational safety and space utilization. Traditional goods management solutions typically involve a one-time dimensional measurement upon goods arrival and stacking planning based on this static data. However, during actual storage, goods are often affected by factors such as changes in environmental temperature and humidity, their own gravitational creep, and stacking pressure, causing dynamic deformation in their physical dimensions. For example, cardboard boxes may develop side dents, while flexible packaging may develop localized bulges. Because current technology lacks a real-time monitoring and feedback mechanism for goods deformation, the initially planned stacking scheme cannot adapt to the continuous evolution of goods geometry, easily leading to mismatches between stack layers. This can result in safety hazards such as stack tilting and collapse, reducing the reliability of warehouse management. Summary of the Invention

[0003] To address the problem that static stacking planning in existing technologies cannot adapt to the dynamic deformation of goods, resulting in poor stacking stability, this application proposes a goods stacking management method that monitors goods deformation in real time and dynamically updates the stacking geometry model, thereby achieving adaptive adjustment of the stacking scheme and ensuring the long-term stability of the goods stacking structure.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A cargo stacking management method includes: monitoring cargo deformation to determine changes in the cargo's geometric shape; updating a stacking geometric model characterizing the spatial relationships between the cargo based on the geometric shape changes; and adjusting the cargo stacking scheme based on the updated stacking geometric model. This method, by introducing a deformation monitoring mechanism, transforms a static geometric description into a dynamic evolutionary model, enabling stacking decisions to be based on the real-time physical state of the cargo, effectively solving the geometric mismatch problem caused by environmental factors.

[0006] As one implementation method, monitoring the deformation of goods to determine the geometric shape changes of the goods includes: acquiring real-time surface monitoring data of the goods within a preset time interval, and mapping the real-time surface monitoring data to the coordinate system of the stacked geometric model using a rigid body transformation matrix to obtain time-series monitoring point cloud data; identifying feature points in the time-series monitoring point cloud data, and tracking the displacement of the feature points to construct the deformation trajectory of the goods, thereby obtaining a set of trajectory vectors; calculating the length contraction value of the protruding part on the goods and the depth change value of the bottom of the recessed space based on the set of trajectory vectors to obtain a geometric deformation dataset; and determining the amplitude and direction of the geometric shape changes through vector synthesis calculation based on the geometric deformation dataset.

[0007] The above scheme achieves quantitative capture of minute local deformations of goods through coordinate system mapping and feature point tracking, providing high-dimensional trajectory data support for accurate updates of the subsequent model.

[0008] As one implementation, updating the stacking geometry model characterizing the spatial relationship between the goods based on the geometric shape change includes: constructing a deformation vector field based on the magnitude and direction of the geometric shape change; mapping the initial contact surface coordinates in the stacking geometry model using the deformation vector field to calculate a coordinate offset sequence; correcting the coordinate offset sequence by combining a preset embedding depth adjustment parameter to generate corrected contact point displacement data; and updating the spatial relationship matrix contained in the stacking geometry model based on the corrected contact point displacement data to obtain an updated stacking geometry model characterizing the adjusted contact surface distribution state.

[0009] The above scheme can correct the contact mechanics between goods in real time by updating the deformation vector field and the spatial relationship matrix in a linked manner, so as to ensure that the model can accurately reflect the distribution of the contact surface after deformation.

[0010] As one implementation, adjusting the stacking scheme of the goods based on the updated stacking geometry model includes: constructing a surface texture feature model based on the contact surface distribution state represented by the updated stacking geometry model, and calculating the offset angle relative to the reference plane; comparing the deviation metric value corresponding to the offset angle with a preset matching deviation threshold; triggering a layout mesh mapping update mechanism to correct the original mapping relationship table if the deviation metric value exceeds the matching deviation threshold; and reallocating the position allocation matrix according to the corrected mapping relationship table to generate a new stacking position allocation scheme as the adjusted stacking scheme.

[0011] The above scheme triggers mesh remapping by determining the threshold of the offset angle, thereby achieving automatic rearrangement of stacked positions and avoiding overall structural failure caused by the accumulation of local deviations.

[0012] As one implementation, the method further includes establishing an initial reference geometric model of the stacked geometric model before monitoring the deformation of the goods. The establishment process includes: periodically collecting initial three-dimensional dimension data of the goods using a scanning device, and extracting contour features for the length of the protruding part and the depth of the recessed space to obtain a contour feature dataset; aggregating the contour feature dataset using a hierarchical clustering algorithm to generate multiple feature clusters; calculating the relative position vector between the centroids of the feature clusters, and constructing an initial spatial relationship matrix based on the relative position vector; and using the initial spatial relationship matrix as the reference geometric model for subsequent deformation monitoring and comparison.

[0013] The above scheme extracts the original geometric features of the cargo through hierarchical clustering algorithm, and constructs a high-precision initial reference system, laying the foundation for subsequent deformation difference analysis.

[0014] As one implementation, after generating the new stacking position allocation scheme, the method further includes: acquiring geometric feature data of the protruding portion and three-dimensional contour data of the recessed space according to the new stacking position allocation scheme, and calculating the volume overlap rate of the protruding portion cutting into the recessed space to generate matching state data; inputting the matching state data into the particle swarm optimization algorithm model, and optimizing the embedding depth adjustment parameters by iteratively updating the velocity and position of the particle swarm; outputting the optimal embedding depth adjustment value after the particle swarm optimization algorithm model converges; and performing virtual reconstruction of the stacking model using the optimal embedding depth adjustment value, and determining the optimized stacking structure integrity index based on the force balance parameters after the virtual reconstruction.

[0015] The above scheme uses particle swarm optimization algorithm to nonlinearly optimize the embedding depth and verifies the force balance through virtual reconstruction, which significantly improves the mechanical stability index of the stacked structure.

[0016] As one implementation, the method further includes: continuously monitoring the matching status of real-time deformation trajectory data and the spatial relationship matrix based on the optimized stacking structure integrity index; triggering a closed-loop update process when a preset deviation is detected between the matching status and the stacking structure integrity index; in the closed-loop update process, cyclically executing the correction of the contact surface coordinates and the update of the layout mesh mapping; and determining that the stacking geometric model has reached a stable state until the update change of the spatial relationship matrix is ​​less than a preset convergence threshold, and outputting the final stable stacking geometric configuration.

[0017] The above solution achieves continuous self-healing of the stacking state through a closed-loop feedback mechanism, ensuring the ultimate stability of cargo stacking configuration in complex dynamic environments.

[0018] As one implementation, this application also provides a cargo stacking management device, comprising: a monitoring module for monitoring the deformation of cargo to determine the geometric shape change of the cargo; an updating module for updating a stacking geometric model representing the spatial relationship between the cargo according to the geometric shape change; and an adjustment module for adjusting the stacking scheme of the cargo based on the updated stacking geometric model.

[0019] As one implementation, this application also provides an electronic device and a computer-readable storage medium that implement the methods described in any of the above-mentioned embodiments by executing a computer program.

[0020] Beneficial effects:

[0021] This invention provides a cargo stacking management method that, by real-time monitoring of cargo deformation trajectories caused by environmental factors and dynamically updating the spatial relationship matrix and stacking geometry model, achieves a leap from static planning to dynamic adaptive management. Its principle lies in using a deformation vector field to correct the contact surface coordinates in real time, and combining this with a particle swarm optimization algorithm to optimize the embedding depth at a mechanical level. This allows for automatic triggering of a remapping of the layout mesh when cargo undergoes physical dimensional contraction or expansion. Compared to existing technologies, this solution significantly reduces the risk of stack tilting caused by cargo deformation, improving the safety and space utilization efficiency of warehousing operations. Attached Figure Description

[0022] Figure 1 A flowchart of a cargo stacking management method provided in an embodiment of the present invention;

[0023] Figure 2 The system schematic diagram provided for an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. In the following description, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Furthermore, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0025] To address the common problem of stacking instability caused by dynamic deformation of goods due to environmental factors in modern warehousing and logistics, embodiments of this application provide a goods stacking management scheme. This scheme establishes a dynamically evolving geometric model, transforming the real-time monitored deformation trajectory into correction parameters for a spatial relationship matrix, thereby achieving adaptive adjustment of the stacking scheme. This decision-making mechanism based on real-time feedback of physical state aims to replace the traditional static planning model to cope with the complex dimensional evolution of goods during the storage cycle, ensuring the mechanical balance and space utilization efficiency of the stacking structure in dynamic environments.

[0026] Example 1:

[0027] Figure 1 This is a flowchart of a cargo stacking management method according to an embodiment of this application. Figure 1 As shown, this embodiment provides a cargo stacking management method, which specifically includes the following steps:

[0028] Step S101: Monitor the deformation of the cargo to determine the changes in the geometric shape of the cargo.

[0029] In real-world warehousing scenarios, goods (such as cardboard boxes, flexible packaging bags, and plastic buckets) undergo dynamic physical evolution during long-term stacking due to fluctuations in ambient temperature and humidity, creep caused by their own gravity, and compressive stress from upper layers of goods. Monitoring the deformation of goods refers to acquiring spatial coordinate data of the goods' surface in real-time or near real-time through external sensing methods. For example, laser scanners, depth cameras (such as RGB-D cameras), or ultrasonic sensor arrays installed above the conveyor belt or on the side of the stack can be used to perform multi-dimensional contour scanning of the target goods. By comparing scan data at different time points, the system can identify subtle changes in the goods' geometry, such as outward bulging or inward depression of the sidewalls of cardboard boxes due to pressure, or bottom expansion of flexible packaging goods due to internal material settling. This deformation monitoring transforms traditional static dimensional measurement into dynamic trajectory tracking, providing a real and real-time physical basis for subsequent stacking stability analysis.

[0030] Step S102: Update the stacking geometry model used to characterize the spatial relationship between the goods according to the geometric shape change.

[0031] The stacking geometry model is a mathematical model used to simulate and record the relative positions, contact states, and embedding depths of goods in a virtual space. After step S101 determines the geometric changes of the goods (such as the length contraction of protruding parts or the increase in the depth of recessed spaces), the system quantifies these changes into deformation vectors and applies them to the existing model parameters. Specifically, the system updates the spatial relationship matrix recorded in the model, which contains key data such as the coordinates of the contact surfaces between goods, the distance between their centroids, and the overlap volume. For example, if the depth of the recessed space of the lower goods increases by 5mm, the corresponding contact surface height parameter in the stacking geometry model will be adjusted downwards to reflect the actual change in the physical support surface. Through this linked update mechanism, the model can always maintain a high degree of consistency with the physical entity, avoiding stacking decision errors caused by model distortion.

[0032] Step S103: Based on the updated stacking geometry model, adjust the stacking scheme of the goods.

[0033] After acquiring a stacking geometry model reflecting the current true state of the goods, the system reassesses the stability of the existing stacking structure. Adjusting the stacking scheme involves replanning the placement, rotation angle, and embedding depth of subsequent goods to be stacked. For example, if the updated model shows that the original contact surfaces can no longer provide sufficient friction or support due to deformation, the system automatically triggers layout adjustment logic, assigning new goods to more stable grid coordinate points or adjusting the embedding depth of the goods to compensate for gaps caused by deformation. Through this adaptive adjustment, it ensures that the stacking of each layer of goods is based on the most reliable geometric matching at present.

[0034] The method provided in this embodiment solves the lag problem exhibited by static stacking planning when faced with dynamic deformation of goods through a closed-loop logic of "monitoring-updating-adjustment". Its technical principle lies in mapping the physical deformation of goods to a digital spatial relationship matrix in real time, and using a dynamically updated model to guide stacking decisions. Experimental data shows that adopting this solution can significantly enhance the operational safety and space utilization efficiency of automated warehousing systems.

[0035] It should be noted that the above steps and examples are merely explanations of the core solution of this invention and are not intended to limit the scope of protection. In practical applications, the monitoring frequency, model update algorithm, and adjustment strategy can be flexibly configured according to the material properties of the goods and the specific needs of the storage environment.

[0036] Example 2:

[0037] This embodiment provides a refined scheme for monitoring cargo deformation to determine geometric shape changes. Monitoring cargo deformation to determine geometric shape changes specifically includes the following steps:

[0038] Step S201: Within a preset time interval, acquire real-time surface monitoring data of the cargo, and use a rigid body transformation matrix to map the real-time surface monitoring data to the coordinate system of the stacked geometric model to obtain time-series monitoring point cloud data.

[0039] In practical applications, the preset time interval can be flexibly set according to the material characteristics of the goods and the frequency of environmental fluctuations. For example, for molded pulp packaging that is easily affected by moisture and deformation, the time interval can be set to 10-30 minutes, while for relatively stable plastic pallet goods, the time interval can be set to 2-4 hours. The original 3D point cloud of the goods surface is acquired in real time using laser scanners or depth cameras placed around the stack. Since the scanning equipment may experience slight vibrations or positional shifts, this embodiment uses a rigid body transformation matrix (including rotation matrix R and translation vector T) and an Iterative Closest Point (ICP) algorithm to spatially align the real-time acquired point cloud data with the initially established stacked geometric model. This coordinate system mapping eliminates displacement noise caused by non-deformation factors, ensuring that each set of point cloud data in subsequent analysis is in a unified reference coordinate system, thereby forming time-series monitoring point cloud data that reflects the deformation evolution process.

[0040] Step S202: Identify feature points in the time series monitoring point cloud data and track the displacement of the feature points to construct the cargo deformation trajectory, thereby obtaining a set of trajectory vectors.

[0041] After acquiring time-series monitoring point cloud data, the system uses feature extraction operators (such as Harris3D or ISS feature point detection algorithms) to identify key geometric feature points on the cargo surface, such as corners, midpoints of edges, or vertices of protrusions. By matching these feature points across consecutive time frames, the movement path of each feature point in three-dimensional space can be tracked. For example, tracking the displacement trajectory of the center point on the side of a cardboard box over 24 hours; if the point continuously moves along the negative direction of the coordinate axis, it indicates that the area is collapsing. Vectorizing the displacement paths of all tracked feature points yields a set of trajectory vectors representing the overall or local deformation trend of the cargo.

[0042] Step S203: Based on the trajectory vector set, calculate the length contraction value of the protruding part on the cargo and the depth change value of the bottom of the recessed space to obtain a geometric deformation dataset.

[0043] Based on the aforementioned set of trajectory vectors, the system further quantifies local geometric dimensional changes. For protruding parts on the cargo surface (such as reinforcing ribs or fastening points on packaging), the length contraction value is determined by calculating the projection of its vertex trajectory vector onto the normal direction. For example, if the length of a protrusion is detected to shrink from an initial 15.0 mm to 13.2 mm, the contraction value is 1.8 mm. Similarly, for recessed spaces on the cargo surface (such as reserved stacking slots or pits formed by pressure), the depth change value is calculated by analyzing the displacement of feature points at the bottom of the recess. These quantified values ​​together constitute a geometric deformation dataset, providing a data foundation for evaluating the physical matching degree between stacked layers.

[0044] Step S204: Based on the geometric deformation dataset, determine the magnitude and direction of the geometric shape change through vector synthesis calculation.

[0045] After obtaining the local shrinkage and depth change values, the system uses a vector synthesis algorithm to process these discrete deformation data globally. By performing weighted summation or tensor analysis on each vector in the geometric deformation dataset, the resultant vector of the overall geometric shape change of the cargo is calculated. The magnitude of this resultant vector represents the magnitude of the deformation, and its direction represents the main direction of the deformation (such as inward shrinkage, tilting to the left, etc.).

[0046] This embodiment achieves precise capture of dynamic deformation of goods at the micrometer to millimeter level through the synergistic effect of coordinate system mapping, feature tracking, and vector synthesis. Compared with traditional static measurement, this dynamic monitoring mechanism can quantify the real-time impact of environmental factors (such as humidity causing cardboard box softening) on ​​the geometric dimensions of goods. Experimental data shows that, using the deformation monitoring method described in this embodiment, the system can achieve an accuracy of less than 0.5 mm in recognizing changes in the geometric dimensions of goods, providing a high-confidence input for the real-time correction of the subsequent stacking geometry model and effectively preventing the risk of stacking instability caused by cumulative deformation.

[0047] Example 3:

[0048] This embodiment provides a refinement scheme for updating a stacked geometry model, used to accurately correct the spatial topological relationships between goods after detecting deformation. The updating of the stacked geometry model characterizing the spatial relationships between goods based on the geometric shape changes described in this embodiment specifically includes the following steps:

[0049] Step S301: Construct a deformation vector field based on the magnitude and direction of the geometric shape change.

[0050] Specifically, the system acquires the geometric deformation dataset determined by the aforementioned steps. This dataset contains the displacement vectors of each feature point on the cargo surface in three-dimensional space. The deformation vector field is a continuous vector field generated by spatial interpolation (such as radial basis function interpolation or bilinear interpolation) of these discrete displacement vectors. It can characterize the deformation trend of any coordinate point on the cargo surface and even in the near-surface region. For example, if it is detected that the center of a cardboard box shrinks inward by 3.5 mm, while the edge shrinks by 1.2 mm, the constructed deformation vector field will show a vector distribution that increases from the edge to the center and points inward towards the cargo, thus providing a more refined description of the non-uniform deformation characteristics.

[0051] Step S302: The deformation vector field is used to map the initial contact surface coordinates in the stacked geometric model to calculate the coordinate offset sequence.

[0052] After acquiring the deformation vector field, the system applies it to the initial contact surface mesh defined in the reference geometric model. By inputting the coordinates of each mesh node of the initial contact surface into the deformation vector field, the new position of that node after deformation is calculated, thus obtaining a set of coordinate offset sequences describing the evolution of the contact surface position. For example, for a horizontal contact surface between two goods, if the top surface of the lower goods is recessed, the corresponding contact surface coordinate offset sequence will record the displacement components of each contact point in the vertical direction, such as a 2.8mm subsidence in the central area and a 0.5mm subsidence in the edge area.

[0053] Step S303: Combine the preset embedding depth adjustment parameters to correct the coordinate offset sequence in order to generate corrected contact point displacement data.

[0054] Considering that deformation of the goods may cause the originally tightly fitted protruding parts and recessed spaces to loosen or become excessively compressed, this embodiment introduces an embedding depth adjustment parameter for compensation. This embedding depth adjustment parameter can be preset according to the material properties of the goods (such as the elastic modulus of corrugated cardboard, the toughness of plastic film, etc.). The system calculates the current theoretical embedding depth based on the coordinate offset sequence and determines whether it is within a safe range. If deformation results in insufficient embedding depth, the adjustment parameter is used to correct the displacement data. For example, when the calculated embedding depth decreases by 2mm, combined with a correction coefficient of 1.15, the generated corrected contact point displacement data will guide the contact surface to extend further downwards to compensate for the gap caused by shrinkage.

[0055] Step S304: Based on the corrected contact point displacement data, update the spatial relationship matrix contained in the stacked geometric model to obtain the updated stacked geometric model that characterizes the adjusted contact surface distribution state.

[0056] The spatial relationship matrix is ​​the core mathematical expression of the stacked geometric model, used to record parameters such as the relative pose, overlap volume, and contact pressure distribution between goods. The system substitutes the corrected contact point displacement data into the matrix operation to update the translation vector and rotation operator in the matrix. By analyzing the updated spatial relationship matrix, the adjusted contact surface distribution can be intuitively obtained, such as the effective proportion of the contact area and the offset of the pressure center point.

[0057] This embodiment constructs a deformation vector field and combines it with embedding depth parameters for coordinate correction. The principle behind this is to transform microscopic physical deformation into macroscopic geometric topology updates, overcoming the limitation of traditional static models in failing to reflect the dynamic evolution of the contact surface. Experimental data shows that the matrix update mechanism described in this embodiment can improve the goodness of fit of the stacking model to the actual physical state to over 95%, providing a reliable data foundation for the precise adjustment of subsequent stacking schemes and effectively avoiding mechanical instability caused by model distortion.

[0058] Example 4:

[0059] This embodiment provides a specific implementation method for adjusting the cargo stacking scheme, aiming to achieve dynamic rearrangement of stacking positions by performing refined modeling and threshold determination of the contact state after deformation. The adjustment of the cargo stacking scheme based on the updated stacking geometry model may specifically include the following steps:

[0060] Step S401: Based on the contact surface distribution state represented by the updated stacked geometry model, construct a surface texture feature model and calculate the offset angle relative to the reference plane.

[0061] Specifically, the system acquires point cloud data of the contact surface of the target object and uses this data to construct a surface texture feature model reflecting the current contact state. This surface texture feature model not only includes the geometric contour of the contact surface but also integrates surface roughness distribution and local normal vector information. For example, when the goods are corrugated cardboard boxes, side indentations caused by moisture will change the normal vector field of the contact surface. The system calculates the angle between the normal vector of each sampling point on the contact surface and the ideal horizontal reference plane, and uses a weighted average method to calculate the overall offset angle. If a displacement is detected at the center pressure point of the contact surface of a batch of goods, for example, a 15mm shift to the left from the geometric center, the calculated offset angle may increase from the initial 0.2 degrees to 3.5 degrees.

[0062] Step S402: Compare the deviation metric value corresponding to the offset angle with a preset matching deviation threshold.

[0063] In actual operation, the system pre-sets a matching deviation threshold range, which is usually dynamically set according to the material stiffness and stacking height of the goods. For example, for high-level stacking (more than 5 layers), the preset matching deviation threshold can be set to 2.5 degrees; for low-level stacking, the threshold can be relaxed to 4.0 degrees. The system compares the offset angle (e.g., 3.5 degrees) calculated in real time in step S401 with this threshold to calculate its deviation metric.

[0064] Step S403: If the deviation metric exceeds the matching deviation threshold, trigger the layout grid mapping update mechanism to correct the original mapping relationship table.

[0065] If the deviation metric indicates that the current offset has exceeded the safe steady-state range, the system will automatically output a layout mesh mapping update signal. Upon receiving this signal, the system will correct the original mapping table. The mapping table records the correspondence between feature points on the bottom surface of the goods and mesh nodes on the top surface of the pallet or lower goods. The correction process includes translating or rotating the mesh nodes according to the offset direction to compensate for geometric mismatch caused by deformation. For example, if the offset angle is too large, the system will correct the original linear mapping relationship to a non-linear compensation mapping to ensure that the protrusions of the upper goods can be realigned with the center of the depressions in the lower goods that have shifted due to deformation.

[0066] Step S404: Based on the revised mapping table, the position allocation matrix is ​​reallocated to generate a new stacking position allocation scheme as the adjusted stacking scheme.

[0067] The system recalculates the position allocation matrix based on the revised mapping table. This matrix defines the precise coordinates and attitude angles of each item in the three-dimensional stacking space. Through this reallocation, a new stacking position allocation scheme is generated. For example, an item originally located at coordinates (500, 500) is moved to (512, 495), and its yaw angle is adjusted by 1.2 degrees.

[0068] The solution provided in this embodiment enables deep coupling between stacking decisions and the real-time physical state of goods. Its principle lies in predicting stacking instability risks through quantitative monitoring of offset angles and utilizing mesh remapping technology to eliminate geometric deviations caused by accumulated deformation. Experimental data shows that after adopting the dynamic adjustment mechanism described in this embodiment, local stress concentration in the stacking structure can be reduced by approximately 15% to 22%, effectively avoiding secondary stacking tilting accidents caused by goods becoming "out of round" or "collapsed," and significantly improving the operational safety of automated warehousing systems in complex environments.

[0069] Example 5:

[0070] This embodiment provides a scheme for establishing an initial reference geometric model, used to construct a high-precision reference datum before monitoring cargo deformation. The process of establishing the initial reference geometric model of the stacked geometric model may specifically include:

[0071] Step S501: The initial three-dimensional dimension data of the goods are collected periodically by scanning equipment, and contour features are extracted for the length of the protruding part and the depth of the recessed space to obtain a contour feature dataset.

[0072] In practical applications, the scanning device can be a multi-line LiDAR or a depth camera (such as an RGB-D camera) fixedly mounted above the conveyor belt. The initial three-dimensional dimensional data is usually raw point cloud data, and its point cloud density can be set according to the accuracy requirements of the goods, for example, 10-20 sampling points per square centimeter. After acquiring the point cloud, the edges and surface undulations of the goods are identified through a preset geometric topology algorithm, and the physical length of the protruding parts (such as the corner protrusions of a carton or the sealing allowance of a packaging bag) and the vertical depth of the recessed space (such as pallet holes or stacking gaps between goods) are extracted. For example, for a batch of carton goods, the system identifies the protruding lengths of its four apex corners as 12mm, 15mm, 11mm and 14mm, respectively, and the depth of the indentation formed by the pressure on the side is identified as 8mm. These quantified geometric parameters together constitute the contour feature dataset.

[0073] Step S502: The contour feature dataset is aggregated using a hierarchical clustering algorithm to generate multiple feature clusters.

[0074] The hierarchical clustering algorithm described can employ agglomerative clustering, using Euclidean distance as the similarity criterion. The system treats multiple extracted contour feature points as initial independent clusters, and continuously merges geometric features with high similarity by calculating inter-cluster distances (e.g., using Ward's minimum variance method). For example, when the protrusion lengths of a batch of goods are distributed in the 10-15mm range and are spatially close, the algorithm aggregates them into a cluster representing "standard corner protection features"; while for concave areas with a depth between 5-10mm, they are aggregated into another cluster representing "lateral deformation risk areas." This clustering method can abstract representative geometric structure patterns from messy raw data, effectively filtering scanning noise.

[0075] Step S503: Calculate the relative position vector between the centroids of the feature clusters, and construct an initial spatial relationship matrix based on the relative position vector.

[0076] After obtaining multiple feature clusters, the system calculates the geometric center, i.e., the centroid coordinates, of all feature points within each cluster. By calculating the three-dimensional spatial vectors between the centroids of different clusters, the relative topological relationships between different parts of the cargo can be accurately described. The initial spatial relationship matrix is ​​a multi-dimensional matrix whose elements record the distance, azimuth angle, and overlap weight between each feature cluster. For example, element A in the matrix... ij This can be represented as a center distance of 0.45m between the i-th convex cluster and the j-th concave cluster. This matrix not only records the static dimensions of the goods, but also the logical relationships between the various geometric features in three-dimensional space.

[0077] Step S504: The initial spatial relationship matrix is ​​used as the reference geometric model for subsequent deformation monitoring and comparison.

[0078] The baseline geometric model serves as a "digital fingerprint" for goods upon entry into the warehouse, providing a comparison base map for subsequent dynamic management. When environmental factors cause deformation of the goods, the system only needs to convert the real-time collected data back into a spatial relationship matrix and perform a difference operation with the baseline matrix to quickly pinpoint the specific location of the deformation.

[0079] This embodiment utilizes a hierarchical clustering algorithm to deeply mine contour features, transforming scattered dimensional data into a spatial relationship matrix with topological associations. The principle lies in leveraging the noise resistance of clustering analysis and the structured characteristics of matrix representation to construct a benchmark model that reflects the essential geometric features of the goods. Compared to traditional single-dimensional recording methods, the benchmark model established in this scheme can carry richer spatial constraint information, improving the sensitivity of subsequent deformation monitoring by more than 15%, and laying a reliable data foundation for achieving high-precision dynamic stacking management.

[0080] Example 6:

[0081] This embodiment provides a scheme for optimizing embedding depth and determining integrity metrics based on particle swarm optimization (PSO) algorithm. The aim is to further refine the mechanical stability of the stacked structure after generating a new stacking position allocation scheme. This process specifically includes the following steps:

[0082] Step S601: According to the new stacking position allocation scheme, obtain the geometric feature data of the protruding part and the three-dimensional contour data of the recessed space, and calculate the volume overlap rate of the protruding part cutting into the recessed space to generate matching state data.

[0083] In actual operation scenarios, the system first retrieves the reassigned position allocation matrix to determine the relative coordinates between the goods to be stacked. Using a high-precision 3D scanner or depth camera, it acquires geometric feature data of the protruding portion at the bottom of the upper goods (such as the height, cross-sectional shape, and surface roughness of the protrusion) and 3D contour data of the recessed space at the top of the lower goods (such as the depth, opening width, and internal slope of the recess). Then, using Boolean operations or voxel modeling methods, it simulates the physical process of the protruding portion of the upper goods cutting into the recessed space of the lower goods at the current allocation position, calculating the volume overlap rate between the two in space. For example, if the volume of the protrusion is 100 cm³... 3 The overlapping portion of its embedded recessed space is 85cm. 3 The volume overlap rate is 0.85. This overlap rate, along with parameters such as the normal vector of the contact surface and the friction coefficient, constitutes the matching state data, which is used to quantitatively evaluate the tightness of the current geometric matching.

[0084] Step S602: Input the matching state data into the particle swarm optimization algorithm model, and optimize the embedding depth adjustment parameters by iteratively updating the velocity and position of the particle population.

[0085] To find the optimal mechanical equilibrium point under dynamic deformation conditions, this embodiment introduces a particle swarm optimization (PSO) algorithm. The embedding depth adjustment parameters, horizontal compensation increment, and rotational fine-tuning angle are used as the dimensional space of the particles. A particle swarm (e.g., 30-50 particles) is initialized, with each particle representing a set of potential embedding depth adjustment schemes. The matching state data obtained in step S601 is used as the input constraint of the algorithm, driving the particles to search within the solution space using a preset fitness function (e.g., maximizing the contact area and minimizing the center of gravity offset). During iteration, each particle continuously updates its velocity and position based on its own historical best position and the group's historical best position. Through this collaborative random search mechanism, local optima can be effectively avoided, and the embedding depth adjustment parameters that maximize the stability of the stacked structure can be quickly locked in a complex nonlinear deformation space.

[0086] Step S603: After the particle swarm optimization algorithm model converges, output the optimal embedding depth adjustment value.

[0087] When the fitness function of the particle swarm stabilizes or reaches a preset number of iterations (e.g., 100-200), the model is considered converged. At this point, the system extracts the parameter values ​​corresponding to the optimal particle in the swarm and outputs the optimal embedding depth adjustment value. This adjustment value is a dynamic correction that can compensate for the initial embedding depth failure caused by cargo shrinkage or expansion. For example, if the initial design embedding depth is 15mm, and the optimized adjustment value is +2.2mm, then the final embedded depth is 17.2mm, ensuring that the protruding part can form an interference fit or a tight fit with the deformed recessed space.

[0088] Step S604: The stacked model is virtually reconstructed using the optimal embedding depth adjustment value, and the optimized stacked structure integrity index is determined based on the force balance parameters after the virtual reconstruction.

[0089] Before physically executing the stacking action, the system virtually reconstructs the stacking model in a digital twin environment using the output optimal embedding depth adjustment value. Through finite element analysis (FEA) or multibody dynamics simulation, the effects of gravity, lateral vibration, and interlayer pressure on the reconstructed model are simulated to obtain force balance parameters, including but not limited to contact stress distribution contour maps, resultant moment deviations, and critical values ​​for structural instability. Based on these parameters, the optimized stacking structural integrity index is calculated. This index is a comprehensive score (e.g., a value between 0 and 1) used to quantitatively evaluate the overturning resistance and structural strength of the current stacking scheme.

[0090] This embodiment employs a particle swarm optimization algorithm to nonlinearly optimize the embedding depth and combines this with virtual reconstruction to verify force balance. The principle behind this approach is to utilize heuristic search to address the mechanical instability caused by deformation uncertainty. Compared to traditional fixed-depth stacking, this solution increases the effective contact area between stacked layers by 15%-25%, significantly enhancing the self-locking effect of goods during transportation or long-term storage, thereby reducing the risk of stacking collapse due to geometric mismatch at the source.

[0091] Example 7:

[0092] This embodiment provides a scheme for triggering closed-loop updates based on integrity indicators to output a stable configuration. Building upon the previous embodiment, this scheme introduces a closed-loop feedback mechanism to achieve continuous self-healing and dynamic stability of the cargo stacking state. The specific steps are as follows:

[0093] Step S701: Based on the optimized stacking structure integrity index, continuously monitor the matching of real-time deformation trajectory data and spatial relationship matrix.

[0094] In actual operation scenarios, the system acquires real-time deformation trajectory data and an initial spatial relationship matrix of stacked objects. Specifically, point cloud data of the goods is collected at a preset frequency (e.g., every 2 to 5 seconds) using LiDAR or a depth camera, and the centroid coordinate sequence of each layer of goods is extracted to form a trajectory vector that evolves over time. Simultaneously, the system calculates the geometric deviation between adjacent layers in the current spatial relationship matrix. Based on the real-time deformation trajectory data and the initial spatial relationship matrix, the current stacking structure integrity index is calculated. For example, if the centroid offset of a certain layer of goods exceeds a preset stability threshold (e.g., 3cm-5cm), or the gap change rate between adjacent goods exceeds 10%, the current integrity index is determined to have fluctuated.

[0095] Step S702: When a preset deviation is detected between the matching situation and the stacking structure integrity index, a closed-loop update process is triggered.

[0096] The system maps the stacked structure integrity index calculated in real time to the initial spatial relationship matrix to obtain a matching degree value reflecting the consistency between the current physical state and the model's predicted state. If the matching degree value shows a deviation, for example, if the matching degree is lower than a preset reliability threshold (e.g., 0.85), the system automatically generates corrected contact surface coordinates. This correction process aims to find a new mechanical equilibrium point by compensating for the displacement caused by deformation.

[0097] In step S703, the closed-loop update process is performed cyclically to correct the contact surface coordinates and update the layout mesh mapping.

[0098] After receiving the corrected contact surface coordinates, the system generates an updated layout mesh mapping. During this process, the system uses a nonlinear optimization algorithm (such as the Levenberg-Marquardt algorithm) to iteratively correct the contact surface control points and simultaneously adjusts the topology of the layout mesh. The system aggregates the updated layout mesh mapping and the corrected contact surface coordinates, continuously reducing the residual between the measured trajectory and the model space relationship through multiple iterative cycles.

[0099] Step S704: Until the update change of the spatial relationship matrix is ​​less than the preset convergence threshold, determine that the stacked geometry model has reached a stable state, and output the final stable stacked geometry configuration.

[0100] The system continuously monitors the change in the spatial relationship matrix between adjacent iterations (e.g., using the Frobenius norm for measurement). When the change is less than a preset convergence threshold (e.g., 0.005-0.01), it indicates that the stacked structure has reached a new geometric equilibrium under the current deformation state. At this point, the final stable stacked geometric configuration is output, including the coordinates of the centroid of each layer of goods, the rotation angle, and the embedding depth parameter.

[0101] This embodiment solves the structural instability problem caused by cumulative deformation of goods during long-term storage by constructing a closed-loop control logic of "monitoring-evaluation-correction-re-evaluation". Through continuous matching monitoring and cyclic updates, the geometric stability of the stacked structure can be improved by more than 25%, ensuring the final reliability of the goods stacking configuration in complex dynamic environments (such as drastic fluctuations in temperature and humidity or micro-vibrations of the foundation), and significantly reducing the frequency of human intervention and the risk of goods damage.

[0102] Example 8:

[0103] This embodiment provides a cargo stacking management device for executing the cargo stacking management method described in the above embodiment. The cargo stacking management device may include:

[0104] The monitoring module is used to monitor the deformation of goods to determine changes in their geometric shape. In practical applications, the monitoring module can communicate with sensing devices such as laser scanners and depth cameras deployed on warehouse assembly lines or shelving areas. By acquiring surface point cloud data of the goods at different time points, it can identify differences in physical dimensions caused by fluctuations in environmental temperature and humidity or gravitational compression. For example, the monitoring module can quantify the depth change of the inward indentation of the sidewalls of cardboard boxes, or the extent of outward expansion of the bottom of flexible packaging goods under pressure, thereby providing real-time geometric deviation data for subsequent model updates.

[0105] An update module is used to update the stacking geometry model representing the spatial relationship between the goods based on the geometric shape changes. The update module dynamically corrects the parameters in the pre-stored stacking geometry model by receiving deformation vector data output by the monitoring module. Specifically, the update module can adjust the spatial coordinates of each contact surface of the goods in the model and recalculate the element values ​​in the spatial relationship matrix, so that the model can accurately reflect the physical contours of the goods after deformation. For example, when shrinkage of the top surface of the lower layer of goods is detected, the update module synchronously reduces the effective bearing area parameter of the corresponding contact area in the model, ensuring that the model remains highly synchronized with the actual physical state of the goods.

[0106] An adjustment module is used to adjust the stacking scheme of the goods based on the updated stacking geometry model. The adjustment module re-plans the placement or embedding depth of subsequent goods according to the latest spatial constraints reflected in the updated model. For example, if the updated model shows that the original embedding position has a matching deviation exceeding a safety threshold due to deformation, the adjustment module will trigger a remapping of the layout mesh to generate a new stacking position allocation scheme, such as shifting the center of gravity of the upper-layer goods to the left by 5mm-10mm to avoid the severely deformed area of ​​the lower-layer goods.

[0107] Through the synergistic effect of the aforementioned modules, the device described in this embodiment achieves closed-loop management from deformation perception to model evolution and solution decision-making. Its technical principle lies in transforming static warehouse management logic into a dynamic optimization process based on real-time geometric feedback. By capturing microscopic deformations through a monitoring module, correcting spatial relationships through an update module, and optimizing the macroscopic layout through an adjustment module, it effectively solves the stacking instability problem caused by the evolution of the physical dimensions of goods. Experimental data shows that using this device for stacking management can improve the overall force balance of stacks in complex environments by 15%-25%, significantly reducing the collapse accident rate caused by geometric mismatch and ensuring the operational safety of the automated warehousing system.

[0108] Example 9:

[0109] This embodiment provides an electronic device for implementing the above-described cargo stacking management method. The electronic device can be a server, an industrial control computer, or a cloud processing platform, etc. The electronic device may include a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements the cargo stacking management method as described in the foregoing embodiment.

[0110] Specifically, the processor can be a general-purpose central processing unit (CPU), microcontroller (MCU), application-specific integrated circuit (ASIC), or field-programmable gate array (FPGA) or other logic devices with computing capabilities. The memory can include volatile memory (such as random access memory RAM) and non-volatile memory (such as read-only memory ROM, flash memory, hard disk, etc.) for persistently storing the computer program and the stacked geometric model, spatial relationship matrix, deformation trajectory data, etc. generated during the execution of the method.

[0111] In practical applications, the processor is coupled to the memory via a bus architecture. When the computer program is invoked and executed, the processor first controls the monitoring module to acquire deformation data of the goods. For example, the processor receives raw point cloud data from a laser scanner or depth camera via a communication interface and uses its built-in geometric operation unit to determine the geometric shape changes of the goods. Subsequently, based on the geometric shape changes, the processor updates the stacking geometric model in memory that characterizes the spatial relationships between the goods. During this process, the processor can call a matrix operation library to correct the elements in the spatial relationship matrix in real time to reflect changes in the shrinkage of protrusions or the depth of indentations caused by environmental factors. Finally, based on the updated stacking geometric model, the processor adjusts the stacking scheme of the goods by executing a path planning algorithm or a layout optimization algorithm and outputs new stacking position allocation instructions to the automated stacking robot.

[0112] By embedding the cargo stacking management method into a computer program within an electronic device, this embodiment achieves a deep integration of hardware resources and software algorithms. Compared to manual intervention or simple embedded processing, this electronic device provides stronger computing power to handle high-complexity operations such as particle swarm optimization. Experimental data shows that using the electronic device described in this embodiment for stacking management can control the data processing latency to within 200ms, enabling real-time response to minute deformations of goods and improving the overall stability of the stacking structure by approximately 25%, significantly reducing safety risks in the warehousing environment. Furthermore, the electronic device can integrate multiple communication protocols, supporting seamless integration with host computer systems or WMS (Warehouse Management System), enhancing system compatibility and practicality.

[0113] Example 10:

[0114] This embodiment provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the cargo stacking management method as described in the above embodiment.

[0115] Specifically, the computer-readable storage medium may include, but is not limited to, read-only memory (ROM), random access memory (RAM), magnetic disk, optical disk, flash memory, or non-volatile solid-state storage devices. When the computer program is called and executed by the processor, it can drive the hardware device to execute a series of logical instructions such as monitoring cargo deformation, updating the stacking geometry model, and adjusting the stacking scheme.

[0116] In practical applications, the computer-readable storage medium can be integrated into the central control server or edge computing gateway of a warehouse management system. When the computer program is executed, the system can interact with on-site hardware such as laser scanning equipment and environmental sensors to capture in real time the minute geometric deformations of goods caused by temperature and humidity fluctuations or gravitational creep. For example, when processing a batch of cardboard boxes with dents on the sides due to moisture, the program instructions can guide the processor to calculate the change vector of the dent depth and automatically correct the contact surface coordinates in the spatial relationship matrix. Through this automated processing based on storage instructions, the system can maintain the integrity index of the stacked structure above the preset safety threshold in real time. Compared with manual intervention or static management solutions, its response speed to the risk of stack tilting is improved by more than 60%, and it effectively avoids physical mismatch problems caused by data lag.

[0117] Furthermore, the program in the computer-readable storage medium also includes optimization algorithm instructions, such as particle swarm optimization algorithm logic. When cargo deformation is detected, causing the offset angle to exceed the matching deviation threshold (e.g., offset angle greater than 5 degrees), the program will automatically trigger a layout mesh mapping update mechanism to reallocate the position allocation matrix. Through this digital storage and execution mechanism, it is ensured that the stacking decision for each layer of cargo is based on the latest physical dimension data, thereby achieving long-term geometric stability and mechanical balance of cargo stacking configuration in complex dynamic warehousing environments. This method of using software instructions to drive hardware adaptive adjustment significantly reduces cargo damage rates in warehousing operations and improves space utilization efficiency.

[0118] The embodiments described above are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications 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 this application. The cargo stacking management solution provided by this application, through deep perception of cargo deformation trajectories and closed-loop iteration of stacking models, provides highly reliable technical support for the field of intelligent warehousing, and has broad industry application prospects. In practical applications, this solution can significantly improve the geometric stability of stacking, reduce safety risks, and can be flexibly configured and expanded according to different business scenarios. All aspects not detailed in this application are conventional technologies well known to those skilled in the art. Finally, it should be noted that the flowcharts and block diagrams in the specification and drawings of this application illustrate the possible architecture, functions, and operations of systems, methods, and computer program products according to various embodiments of this application, and the order of these operations can be adjusted according to actual needs without violating logic.

[0119] The above-disclosed embodiments are merely a few specific examples of the present invention. However, the embodiments of the present invention are not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A method for managing cargo stacking, characterized in that, include: Monitor the deformation of the cargo to determine changes in its geometric shape; Based on the geometric shape changes, update the stacking geometry model used to characterize the spatial relationship between the goods; Based on the updated stacking geometry model, the stacking scheme of the goods is adjusted; The monitoring of cargo deformation to determine changes in the cargo's geometric shape includes: Within a preset time interval, real-time surface monitoring data of the goods is acquired, and a rigid body transformation matrix is ​​used to map the real-time surface monitoring data to the coordinate system of the stacked geometric model to obtain time-series monitoring point cloud data. Feature points in the time-series monitoring point cloud data are identified, and the displacement of the feature points is tracked to construct the cargo deformation trajectory, thereby obtaining a set of trajectory vectors; Based on the set of trajectory vectors, calculate the length contraction value of the protruding part on the cargo and the depth change value of the bottom of the recessed space to obtain the geometric deformation dataset. And based on the geometric deformation dataset, the magnitude and direction of the geometric shape change are determined through vector synthesis calculation.

2. The method according to claim 1, characterized in that, The step of updating the stacking geometry model, which characterizes the spatial relationship between the goods, based on the geometric shape changes includes: Based on the magnitude and direction of the geometric shape change, a deformation vector field is constructed; The deformation vector field is used to map the initial contact surface coordinates in the stacked geometric model to calculate the coordinate offset sequence; The coordinate offset sequence is corrected by combining the preset embedding depth adjustment parameters to generate corrected contact point displacement data. And based on the corrected contact point displacement data, update the spatial relationship matrix contained in the stacked geometry model to obtain the updated stacked geometry model that characterizes the adjusted contact surface distribution state.

3. The method according to claim 1, characterized in that, The adjustment of the cargo stacking scheme based on the updated stacking geometry model includes: Based on the contact surface distribution state represented by the updated stacked geometry model, a surface texture feature model is constructed, and the offset angle relative to the reference plane is calculated. The deviation metric value corresponding to the offset angle is compared with a preset matching deviation threshold. If the deviation metric exceeds the matching deviation threshold, the layout mesh mapping update mechanism is triggered to correct the original mapping table. And based on the revised mapping table, the position allocation matrix is ​​reallocated to generate a new stacking position allocation scheme as the adjusted stacking scheme.

4. The method according to claim 1, characterized in that, The method further includes establishing an initial reference geometric model of the stacked geometry model before monitoring the deformation of the cargo, the establishment process including: The initial three-dimensional dimension data of the goods are collected periodically by scanning equipment, and contour features are extracted for the length of the protruding part and the depth of the recessed space to obtain a contour feature dataset. The contour feature dataset is aggregated using a hierarchical clustering algorithm to generate multiple feature clusters; Calculate the relative position vectors between the centroids of the feature clusters, and construct an initial spatial relationship matrix based on the relative position vectors; The initial spatial relationship matrix is ​​used as the baseline geometric model for subsequent deformation monitoring and comparison.

5. The method according to claim 3, characterized in that, After generating the new stacking location allocation scheme, the method further includes: According to the new stacking position allocation scheme, the geometric feature data of the protruding part and the three-dimensional contour data of the recessed space are obtained, and the volume overlap rate of the protruding part cutting into the recessed space is calculated to generate matching state data. The matching state data is input into the particle swarm optimization algorithm model, and the embedding depth adjustment parameters are optimized by iteratively updating the velocity and position of the particle swarm. After the particle swarm optimization algorithm model converges, the optimal embedding depth adjustment value is output. The stacking model is virtually reconstructed using the optimal embedding depth adjustment value, and the optimized stacking structure integrity index is determined based on the force balance parameters after the virtual reconstruction.

6. The method according to claim 5, characterized in that, The method further includes: Based on the optimized stacking structure integrity index, the matching status between real-time deformation trajectory data and the spatial relationship matrix is ​​continuously monitored. When a preset deviation is detected between the matching condition and the stacking structure integrity index, a closed-loop update process is triggered. In the closed-loop update process, the correction of the contact surface coordinates and the update of the layout mesh mapping are performed cyclically. And until the update change of the spatial relationship matrix is ​​less than a preset convergence threshold, the stacked geometry model is determined to have reached a stable state, and the final stable stacked geometry configuration is output.

7. A cargo stacking management device, applied to the method as described in any one of claims 1 to 6, characterized in that, include: A monitoring module is used to monitor the deformation of the goods in order to determine the changes in the geometric shape of the goods; An update module is used to update the stacking geometry model that characterizes the spatial relationship between the goods based on the geometric shape changes. An adjustment module is used to adjust the stacking scheme of the goods based on the updated stacking geometry model.

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the computer program is executed by the processor, it implements the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

Images