User behavior response method applied to identification guide system, server and medium

Abstract

This invention discloses a user behavior response method, server, and medium applied to a signage and wayfinding system. The method includes: first, receiving the behavioral trajectory data of a target user in a three-dimensional passage space and generating a forward-looking behavioral trace data stream; second, performing spatiotemporal mapping processing on the forward-looking behavioral trace data stream based on the layout of user field-of-view edge nodes in the digital wayfinding interface to generate a spatiotemporal evolution trend map of cognitive attention; then, determining a pre-wake-up timing indicator and a pre-wake-up intensity indicator based on the spatiotemporal evolution trend map of cognitive attention; further, creating an optical flow particle motion field with the corresponding position of the edge node identifier as an anchor point, embedding the wayfinding identifier template to generate a dynamic wayfinding response interface frame sequence; finally, distributing the dynamic wayfinding response interface frame sequence to the corresponding user field-of-view edge nodes to trigger the wayfinding pre-wake-up display operation. This invention improves the timing accuracy of wayfinding information push and the visual guidance effect.

Description

Technical Field

[0001] This invention relates to the field of computer vision wayfinding technology, specifically to a user behavior response method, server, and medium applied to signage and wayfinding systems. Background Technology

[0002] Signage and wayfinding technology is applied in three-dimensional spaces such as large commercial complexes, transportation hubs, and exhibition venues to provide users with directional guidance, location markings, and information displays. With the development of digital display technology, traditional static wayfinding signage is gradually being replaced by digital wayfinding interfaces composed of distributed digital display nodes. Digital wayfinding interfaces can present wayfinding information according to preset scheduling strategies or manual triggering methods. Currently, the methods for pushing wayfinding information to users in three-dimensional spaces mainly rely on timed rotation, sensor triggering, or user-initiated interaction. However, these technologies lack the ability to predict the continuous changes in user behavior trajectories when pushing wayfinding information, and lack the ability to match and adjust the visual intensity of the wayfinding information to the dynamic changes in user attention. This results in wayfinding information often lagging behind the user's actual information needs or becoming disconnected from the user's current state of attention. Summary of the Invention

[0003] The purpose of this invention is to provide a user behavior response method, server, and medium for use in signage and wayfinding systems, in order to solve the problems mentioned in the background art.

[0004] This invention provides a user behavior response method applied to a signage and wayfinding system, comprising: Receives the target user's behavior trajectory data continuously collected by a non-contact sensing device within a three-dimensional passage space, and generates a forward-looking behavior trace data stream based on the behavior trajectory data; Based on the preset layout of user field edge nodes in the digital navigation interface, the forward-looking behavior trace data stream is subjected to spatiotemporal mapping processing to generate a spatiotemporal evolution trend map of cognitive attention corresponding to each user field edge node. The spatiotemporal evolution trend map of cognitive attention represents the change law of attention of the forward-looking behavior trace data stream in each edge area of ​​the digital navigation interface. The triggering attention mutation peaks and their corresponding timestamps and edge node identifiers are extracted from the spatiotemporal evolution trend map of cognitive attention, serving as pre-wake-up timing indicators for guidance information needs, and the slope of the attention change trend of the mutation peaks is extracted as a pre-wake-up intensity indicator for adjusting the brightness of the guidance interface. Using the pre-wake-up timing indicator and the pre-wake-up intensity indicator, an optical flow particle motion field with directional gradient properties is created in the digital wayfinding interface with the position corresponding to the edge node identifier as the anchor point. The preset wayfinding identifier template is embedded into the optical flow particle motion field in a semantic enhancement manner to generate dynamic wayfinding response interface frames arranged in time sequence. The dynamic wayfinding response interface frame sequence is distributed to the corresponding user field edge nodes according to the edge node identifier, triggering the user field edge nodes to perform wayfinding pre-wake display operations.

[0005] This invention provides a user behavior response server, comprising: A processor; a storage device having a computer program stored thereon; a network interface for providing network communication functions; when the computer program is executed by the processor, the processor implements the above-described user behavior response method applied to a signage and wayfinding system.

[0006] The present invention provides a readable storage medium on which a program or instruction is stored, and when the program or instruction is executed by a processor, it implements the above-described user behavior response method applied to a signage and wayfinding system.

[0007] Compared with existing technologies, the beneficial effects of this invention are as follows: By generating a forward-looking behavior trace data stream from the three-dimensional behavioral trajectory data of the target user, and performing spatiotemporal mapping processing on this data stream based on the layout of edge nodes in the user's field of vision of the digital wayfinding interface, a spatiotemporal evolution trend map of cognitive attention, representing the changing pattern of attention, is generated, enabling the dynamic correlation between the behavioral trajectory and the edge area of ​​the wayfinding interface to be quantitatively captured. Based on this, triggering attention abrupt change peaks are analyzed from the spatiotemporal evolution trend map of cognitive attention. The timestamp and edge node identifier of the abrupt change peak are used as indicators of pre-awakening timing, and the slope of the attention change trend of the abrupt change peak is used as an indicator of pre-awakening intensity, thus achieving accurate prediction of the timing and intensity of the need for wayfinding information. Furthermore, by utilizing pre-wake-up timing and pre-wake-up intensity indicators, an optical flow particle motion field with directional gradient attributes is created in the digital wayfinding interface using the corresponding position of the edge node identifier as the anchor point. The wayfinding identifier template is then embedded into this field in a semantically enhanced manner to generate dynamic wayfinding response interface frames. Finally, the dynamic wayfinding response interface frame sequence is distributed to the corresponding edge nodes of the user's field of vision to trigger the wayfinding pre-wake-up display operation. Overall, the spatiotemporal mapping of user behavior trajectory, the analysis of cognitive attention abrupt change peaks, and the wayfinding expression of optical flow particle motion field are organically connected, realizing end-to-end closed-loop processing from user behavior perception to wayfinding pre-wake-up response, thereby improving the timing accuracy of wayfinding information push and the visual guidance effect.

[0008] In this way, it is possible to predict the continuous changes in user behavior patterns when pushing wayfinding information, and to improve the ability to match and adjust the visual intensity of wayfinding information with the dynamic changes in user attention. This improves the technical problem of wayfinding information being out of sync with users' actual information needs or disconnected from their current attention. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart of a user behavior response method applied to a signage and wayfinding system, provided as an embodiment of the present invention.

[0011] Figure 2 This is a schematic diagram of the basic structure of a user behavior response server provided in an embodiment of the present invention.

[0012] Figure 3 This is a functional block diagram of a user behavior response device provided in an embodiment of the present invention. Detailed Implementation

[0013] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] Please see Figure 1 , Figure 1 This is a flowchart of a user behavior response method for a signage and wayfinding system provided by an embodiment of the present invention. The method can be executed by a user behavior response server or jointly by a user behavior response server and a server. The method includes steps 110-150.

[0015] The technical solution provided by this invention can be applied to user behavior response scenarios in signage and wayfinding systems. In this scenario, a non-contact sensing device and a digital wayfinding interface are deployed in a three-dimensional passage space. The digital wayfinding interface consists of multiple user field-of-view edge nodes distributed at different spatial locations, each with an independent display area and display characteristics. The target user moves freely within the three-dimensional passage space, and their behavioral trajectory is continuously collected by the non-contact sensing device and transmitted to a user behavior response method server executing the method of this invention for processing. The user behavior response method server generates a forward-looking behavioral trace data stream based on the behavioral trajectory data, performs spatiotemporal mapping of the behavioral data according to the preset user field-of-view edge node layout of the digital wayfinding interface, parses out the pre-wake-up timing indication and pre-wake-up intensity indication of the wayfinding information requirement, and then creates an optical flow particle motion field with directional gradient attributes in the digital wayfinding interface, embedding the wayfinding sign template in a semantically enhanced manner, generating a dynamic wayfinding response interface frame sequence and distributing it to the corresponding user field-of-view edge nodes to trigger the wayfinding pre-wake-up display operation.

[0016] Step 110: Receive the target user's behavior trajectory data continuously collected by the non-contact sensing device in the three-dimensional passage space, and generate a forward-looking behavior trace data stream based on the behavior trajectory data.

[0017] In this embodiment of the invention, the core of the behavioral trajectory data is the head orientation trajectory reflecting the target user's attention. To generate this data, head and neck joints are specifically selected as key tracking points from the complete human skeletal joint data collected by a non-contact sensing device. Based on multi-view point cloud registration and fusion processing of skeletal joint spatial coordinates, the three-dimensional coordinates of these two joints in the three-dimensional travel space reference coordinate system are calculated in real time, and the spatial vector from the neck joint to the head joint is defined as the user's attention pointing vector. By performing an intersection operation between the attention pointing vector and the plane where the digital wayfinding interface is located, the behavioral trajectory sampling point representing the user's visual attention projection point on the wayfinding interface at a single acquisition moment can be obtained. The sampling point data structure includes the acquisition timestamp and the three-dimensional spatial coordinates calculated by the intersection. The reason for using the head orientation trajectory instead of the body movement trajectory is that, in a wayfinding scenario, the user's attention exploration behavior is mainly driven by head rotation, and there is a stable statistical coupling relationship between the head orientation and the visual focus point. Body displacement mainly reflects the spatial travel intention and cannot accurately characterize the cognitive tendency change of the edge wayfinding information.

[0018] Multiple behavioral trajectory sampling points generated at various acquisition times are arranged in ascending order of acquisition timestamps to construct a temporally continuous sampling point sequence. The continuity of motion state is verified in this sequence by calculating the ratio of the change in the angle of the attention pointing vector between adjacent sampling points to the acquisition time interval. When this ratio exceeds a preset effective rate of attention transfer, the adjacent sampling point is marked as a trajectory breakpoint. The sampling point sequence is segmented based on these breakpoints, and intermediate connecting sampling points are generated between adjacent segments using spherical linear interpolation. The interpolation density matches the acquisition time interval of the preceding and following sampling points to ensure a continuous and smooth change in the angle of the attention pointing vector. The sampling points from all segments and connecting segments are merged to obtain a continuous and consistent proactive behavioral trace data stream that accurately reflects changes in the user's attention direction.

[0019] Motion state description information is added to each behavior trajectory sampling point in the behavior trajectory data stream. The motion state description information is generated as follows: based on the spatial displacement vector formed by the behavior trajectory sampling point and its preceding adjacent sampling points, the acquisition time interval, and the velocity change, the instantaneous movement direction angle, instantaneous movement speed, and movement acceleration magnitude at the behavior trajectory sampling point are determined. The instantaneous movement direction angle, instantaneous movement speed, and movement acceleration magnitude are combined into a motion state triplet. The motion state triplet is associated and encapsulated with the acquisition timestamp and three-dimensional spatial position coordinate vector of the behavior trajectory sampling point to generate the behavior trajectory data record corresponding to the behavior trajectory sampling point.

[0020] All behavioral trajectory data records are sorted in ascending order according to the acquisition timestamp and written into the forward behavioral trace data stream. Each data record in the forward behavioral trace data stream includes the acquisition timestamp, sampling point identifier, three-dimensional spatial position coordinate vector, instantaneous movement direction angle, instantaneous movement speed, and movement acceleration magnitude.

[0021] Step 120: Based on the preset layout of user field of vision edge nodes in the digital navigation interface, perform spatiotemporal mapping processing on the forward-looking behavior trace data stream to generate a spatiotemporal evolution trend map of cognitive attention corresponding to each user field of vision edge node.

[0022] The spatiotemporal evolution trend map of cognitive attention represents the change pattern of attention in each edge area of ​​the proactive behavior trace data stream on the digital wayfinding interface. The generation of the spatiotemporal evolution trend map of cognitive attention includes the following sub-steps: Step 121: Extract the node identifier of each user field of view edge node in the user field of view edge node layout and the boundary coordinate sequence of the edge node region of the node in the digital navigation interface coordinate system, and obtain the display characteristic information of each user field of view edge node.

[0023] The layout of user field of view edge nodes is a structured configuration data pre-configured in the storage space of the user behavior response method server. This structured configuration data is organized in the form of a node list. Each item in the node list corresponds to a user field of view edge node. Each item includes a node identifier field, a node center point coordinate field, a node display area width field, a node display area height field, a node display direction angle field, and a display characteristic data block field.

[0024] The node identifier field is generated by combining the node's installation location number in the digital wayfinding interface and the node type code. The node center point coordinate field and the node display direction angle field together define the spatial position and orientation of the node in the coordinate system of the digital wayfinding interface. The node display area width field and the node display area height field define the effective rectangular display range of the node.

[0025] After reading the node list, for each node at the edge of the user's field of view, the two-dimensional coordinates of the four vertices of the node's display area in the digital navigation interface coordinate system are calculated based on the node center point coordinate field and the node display direction angle field. The two-dimensional coordinates of the four vertices are arranged in clockwise order to form the boundary coordinate sequence of the node's edge node area. The display characteristic data block field stores the node's maximum display brightness value, minimum display brightness value, color gamut descriptor, contrast ratio, and refresh rate value. The above display characteristic information is parsed from the display characteristic data block field for subsequent optical flow particle brightness modulation and display rendering.

[0026] Step 122: Extract behavior trajectory sampling points sequentially from the forward behavior trace data stream according to the increasing timestamp order, obtain the three-dimensional spatial position coordinates of each behavior trajectory sampling point and its acquisition timestamp, and convert the three-dimensional spatial position coordinates into a three-dimensional trajectory point sequence under the three-dimensional travel space reference coordinate system.

[0027] Data records are read sequentially from the forward-looking behavior trace data stream, and the acquisition timestamp field and 3D spatial position coordinate vector field are obtained from each data record. The coordinate values ​​of the 3D spatial position coordinate vector field are currently in the 3D travel space reference coordinate system. When it is confirmed that the coordinate axis definition of this coordinate system is consistent with the coordinate system definition required for subsequent perspective projection mapping, the 3D spatial position coordinate vector is directly written into the 3D trajectory point sequence. When there are differences in coordinate axis direction or scale, coordinate system rotation transformation and scale transformation are performed to transform the 3D spatial position coordinate vector to a coordinate space that meets the input requirements of subsequent perspective projection mapping before writing it into the 3D trajectory point sequence. Each element in the 3D trajectory point sequence is a triplet, containing the acquisition timestamp, behavior trajectory sampling point identifier, and the transformed 3D spatial position coordinate vector.

[0028] Step 123: Based on the spatial pose parameters of the imaging plane of the digital navigation interface, construct a perspective projection mapping relationship from the three-dimensional access space reference coordinate system to the pixel coordinate system of the digital navigation interface. Through the perspective projection mapping relationship, map each behavior trajectory sampling point in the three-dimensional trajectory point sequence to the coordinates of a two-dimensional projection point on the digital navigation interface to obtain a sequence of two-dimensional projection point coordinates with corresponding timestamps.

[0029] The imaging plane spatial pose parameters of the digital navigation interface are obtained. These imaging plane spatial pose parameters include the three-dimensional coordinates of the center point of the imaging plane in the three-dimensional travel space reference coordinate system, the plane normal direction vector, the plane horizontal axis direction vector and the plane vertical axis direction vector, as well as the pixel width value, pixel height value and physical space size value of each pixel.

[0030] A view transformation description from the three-dimensional travel space reference coordinate system to the imaging plane coordinate system is constructed using the three-dimensional coordinates of the center point of the imaging plane, the plane normal direction vector, the plane horizontal axis direction vector, and the plane vertical axis direction vector. The view transformation description is achieved by loading a combination of rotation matrix and translation vector.

[0031] Based on this, a projection scaling transformation description from the imaging plane coordinate system to the digital wayfinding interface pixel coordinate system is constructed using the pixel width, pixel height, and physical space dimensions of each pixel. The view transformation description and the projection scaling transformation description are concatenated to obtain the perspective projection mapping relationship from the 3D access space reference coordinate system to the digital wayfinding interface pixel coordinate system.

[0032] For each behavior trajectory sampling point in the 3D trajectory point sequence, the 3D spatial coordinate vector of the behavior trajectory sampling point is sequentially transformed by view transformation and projection scaling to calculate the corresponding 2D projection point pixel coordinates on the digital navigation interface. The acquisition timestamp of the behavior trajectory sampling point is then associated with the 2D projection point pixel coordinates to generate a 2D projection point coordinate record. After all behavior trajectory sampling points are mapped, the 2D projection point coordinate records are arranged in ascending order according to the acquisition timestamps to form a 2D projection point coordinate sequence.

[0033] Step 124: Perform inter-frame motion pointing analysis on the coordinates of adjacent timestamps of the two-dimensional projection point in the two-dimensional projection point coordinate sequence. Determine the direction of movement of the behavior trajectory projection within the corresponding time interval based on the relative position changes of the coordinates of adjacent two-dimensional projection points on the digital navigation interface. Obtain the movement offset of the behavior trajectory projection within the corresponding time interval based on the spatial interval distance of the coordinates of adjacent two-dimensional projection points.

[0034] Extract pairs of 2D projection point coordinate records with adjacent timestamps sequentially from the 2D projection point coordinate sequence. The earlier time record is designated as the first projection point coordinate, and the later time record as the second projection point coordinate. Calculate the coordinate difference between the second and first projection point coordinates in the pixel coordinate system of the digital navigation interface along the horizontal axis (referred to as the first coordinate axis difference) and the coordinate difference along the vertical axis (referred to as the second coordinate axis difference). Construct the projection point displacement vector within the adjacent time interval using the first and second coordinate axis differences.

[0035] Then, the magnitude of the projection point displacement vector is calculated. This magnitude is obtained by taking the square root of the sum of the squares of the differences between the first and second coordinate axes, and is used as the movement offset within the adjacent time interval. The direction angle of the projection point displacement vector is calculated by using the arctangent function of the ratio of the difference between the second and first coordinate axes, and is used as the movement direction within the adjacent time interval. The movement direction reflects the movement trend of the target user's gaze projection point on the digital navigation interface from the coordinates of the first projection point to the coordinates of the second projection point. The movement direction and movement offset are associated with the later acquisition timestamp in the pair of two-dimensional projection point coordinate records and stored as an inter-frame motion direction description record.

[0036] Step 125: Based on the boundary coordinate sequence of the edge node region and the movement direction of the behavior trajectory projection, determine the user's field of vision edge node that the behavior trajectory projection point is facing at the current moment by considering the projection intersection relationship of the two-dimensional projection point coordinates extending along the movement direction to the edge of the digital navigation interface, and establish an association mapping between the behavior trajectory projection point at the current moment and the node identifier of the user's field of vision edge node.

[0037] For each 2D projection point coordinate record in the 2D projection point coordinate sequence, obtain the corresponding movement direction. Starting from the 2D pixel coordinates of that 2D projection point coordinate record, generate a directional ray that originates from that starting point and passes through the edge of the digital wayfinding interface along the movement direction. The directional ray is represented parametrically, where any point on the directional ray is represented as the starting pixel coordinates plus the product of the unit direction vector of the movement direction and a distance parameter.

[0038] The system sequentially reads the boundary coordinate sequence of each user's field of view edge node. For each node, it determines whether the pointing ray intersects geometrically with the rectangular region enclosed by the node's boundary coordinate sequence. This is done by performing line segment intersection checks on the parameterized expression of the pointing ray and the four boundary segments of the rectangular region one by one. If an intersection exists, the pixel coordinates of the intersection and the distance along the pointing ray between the intersection and the starting point are recorded. For all user's field of view edge nodes with intersections, the distances between each intersection and the starting point are compared, and the user's field of view edge node with the shortest distance is selected as the user's field of view edge node towards which the current behavior trajectory projection point is directed.

[0039] When a unique nearest intersection point exists, a triplet association mapping is established between the node identifier of the user's field of view edge node and the current 2D projection point coordinate record and acquisition timestamp. When the pointing ray does not intersect with the rectangular area of ​​any user's field of view edge node, the 2D projection point coordinate record is marked as a node without orientation.

[0040] Step 126: Set a sliding time window for each user's visual edge node, and collect behavioral trajectory projection points with associated mappings based on the sliding time window, and generate the spatiotemporal evolution trend map of cognitive attention through the behavioral trajectory projection points.

[0041] Step 1261: Set a sliding time window with the current time as the end boundary for each user's field of vision edge node, collect all behavior trajectory projection points that establish an association mapping with the user's field of vision edge node within the sliding time window, and generate a set of associated projection points of the user's field of vision edge node within the corresponding time window.

[0042] A sliding time window is defined for each user's field of view edge node. The length of the sliding time window is determined by two preset parameters: the window time span and the sliding time step. The window time span is a fixed time interval, and the sliding time step represents the interval between two adjacent time windows advancing on the time axis. The current processing time is used as the end boundary timestamp of the sliding time window. The start boundary timestamp of the sliding time window is obtained by subtracting the window time span from the end boundary timestamp. The start boundary timestamp and the end boundary timestamp together define the time coverage of the current time window. Within this time coverage, all association mapping records whose node identifiers match the node identifier of the user's field of view edge node and whose acquisition timestamps fall within the time coverage are retrieved from the set of triples used to establish association mappings. The pixel coordinates and acquisition timestamps of the two-dimensional projection point coordinate records in the above association mapping records are extracted to form the set of association projection points of the user's field of view edge node within the current sliding time window. Each element in the set of association projection points contains a two-dimensional projection point coordinate and an acquisition timestamp.

[0043] Step 1262: Generate cognitive attention tendency status information of the user's visual field edge node within the current time window based on the spatial distribution dispersion and the consistency of the movement direction of the behavioral trajectory projection points in the associated projection point set. The cognitive attention tendency status information includes an aggregation tendency description reflecting the spatial distribution concentration trend and a directional convergence description reflecting the degree of convergence of directions.

[0044] For the two-dimensional projection coordinates of all behavioral trajectory projection points in the associated projection point set, calculate the mean of the coordinates along the horizontal axis as the first axis coordinate of the spatial distribution center point, and the mean of the coordinates along the vertical axis as the second axis coordinate of the spatial distribution center point, thus obtaining the coordinates of the spatial distribution center point. Calculate the Euclidean distance between the two-dimensional projection coordinates of each behavioral trajectory projection point and the coordinates of the spatial distribution center point. Take the arithmetic mean of all calculated Euclidean distance values, and use this arithmetic mean as the spatial distribution dispersion value. The smaller the spatial distribution dispersion value, the more concentrated the distribution of the behavioral trajectory projection points.

[0045] The aggregation tendency description is generated by transforming the spatial distribution dispersion value through a pre-defined monotonically decreasing mapping from dispersion to aggregation tendency. This mapping ensures that a high aggregation tendency description value corresponds to a low dispersion value, and a low aggregation tendency description value corresponds to a high dispersion value. Simultaneously, the movement direction corresponding to each behavioral trajectory projection point in the associated projection point set is obtained. All movement directions are expressed as unit direction vectors, and the composite vector of all unit direction vectors is calculated. The magnitude of the composite vector is divided by the total number of direction vectors involved in the composite to obtain the directional convergence description. The closer the directional convergence description value is to the upper limit, the more consistent the movement directions of the behavioral trajectory projection points are. The aggregation tendency description and the directional convergence description are combined to form cognitive attention tendency state information.

[0046] Step 1263: Slide the sliding time window along the time axis with a set time step, and iteratively generate cognitive attention tendency state information for each time window for each user's visual field edge node, so as to obtain the cognitive attention tendency state change sequence of each user's visual field edge node as the sliding time window moves.

[0047] After generating the cognitive attention tendency state information for the current time window, the current processing time is increased by a sliding time step to obtain a new current processing time. The new current processing time is then used as the end boundary timestamp of the new sliding time window. Steps 1261 and 1262 are repeated to generate the cognitive attention tendency state information for the next time window. During this iteration, adjacent sliding time windows partially overlap on the time axis. The length of the overlapping area is the window time span minus the sliding time step.

[0048] The cognitive attention tendency state information generated sequentially for each time window targeting the same user's visual field edge node is arranged in ascending order according to the end boundary timestamp of the time window, forming a cognitive attention tendency state change sequence for that user's visual field edge node. Each time point entry in the cognitive attention tendency state change sequence includes the end boundary timestamp of the time window, an aggregation tendency description value, and a direction convergence description value.

[0049] Step 1264: The cognitive attention tendency state change sequence of all user vision edge nodes is time-synchronized and aligned according to a common time axis to generate a multi-dimensional state evolution description with the time axis as the first dimension and the node identifier as the second dimension, and the multi-dimensional state evolution description is used as the spatiotemporal evolution trend diagram of cognitive attention.

[0050] Obtain the cognitive attention tendency state change sequences of all user visual field edge nodes, and construct a common time axis covering the timestamp range of each cognitive attention tendency state change sequence. The time resolution of the common time axis is set to the minimum value of each sliding time step. For each user visual field edge node's cognitive attention tendency state change sequence, at each time scale position of the common time axis, use linear interpolation to calculate the aggregation tendency description value and the direction convergence description value at that time scale. The linear interpolation is calculated based on the values ​​of the two adjacent known time window entries before and after that time scale according to the time interval ratio.

[0051] The aggregation tendency description value and the directional convergence description value at each time scale after interpolation are filled into a two-dimensional state description matrix with the time axis as the first dimension index and the node identifier as the second dimension index. The value of each matrix element is a binary tuple containing the aggregation tendency description value and the directional convergence description value. This two-dimensional state description matrix is ​​the multidimensional state evolution description. This multidimensional state evolution description is used as the matrix representation of the spatiotemporal evolution trend diagram of cognitive attention.

[0052] Step 1265: Perform state propagation continuity analysis between nodes on the cognitive attention tendency state change sequence corresponding to spatially adjacent user field edge nodes in the cognitive attention spatiotemporal evolution trend map, and enhance the state transition of adjacent nodes with synchronous state change trends to obtain a cognitive attention spatiotemporal evolution trend map for triggering attention mutation peak analysis.

[0053] The spatial adjacency relationships of all user-visual-edge nodes on the digital navigation interface are extracted from the user-visual-edge node layout. The spatial adjacency is determined by calculating the spatial distance between the center points of any two user-visual-edge nodes. If this spatial distance is less than a preset node proximity threshold, the two user-visual-edge nodes are considered spatially adjacent. For each pair of spatially adjacent user-visual-edge nodes, their respective cognitive attention tendency state change sequences are extracted. Pearson correlation coefficients are calculated for the aggregation tendency description value sequence and the direction convergence description value sequence at the same time scale. When the Pearson correlation coefficient of the aggregation tendency description value sequence is greater than a preset aggregation correlation coefficient threshold and the Pearson correlation coefficient of the direction convergence description value sequence is greater than a preset direction correlation coefficient threshold, the adjacent node pair is considered to have a synchronous state change trend. For adjacent node pairs with synchronous state change trends, the aggregation tendency description value and the direction convergence description value at the corresponding time scale of the cognitive attention spatiotemporal evolution trend map are subjected to weighted smoothing.

[0054] The weighted smoothing process involves multiplying the aggregation tendency description value of the first node at that time scale in an adjacent node pair by a first smoothing weight coefficient, and then adding the aggregation tendency description value of the second node at that time scale multiplied by a second smoothing weight coefficient. This multiplier is used to obtain the enhanced aggregation tendency description value of the first node. Similarly, the directional convergence description value is processed. The sum of the first and second smoothing weight coefficients is a fixed total value. The enhanced aggregation tendency description value and directional convergence description value replace the original values ​​and are written into the spatiotemporal evolution trend map of cognitive attention, forming an enhanced spatiotemporal evolution trend map of cognitive attention for subsequent analysis of abrupt change peaks.

[0055] Step 130: Extract the triggering attention mutation peak and its corresponding timestamp and edge node identifier from the spatiotemporal evolution trend map of cognitive attention, as an indication of the pre-wake-up timing for the guidance information demand, and extract the slope of the attention change trend of the mutation peak as an indication of the pre-wake-up intensity for the brightness adjustment of the guidance interface.

[0056] In step 130, the triggering attention spike refers to the turning point in the user's cognitive attention tendency state change sequence where the state change step size suddenly jumps from a stable stage to a rapid growth stage. It marks the moment when the user's demand for guidance information changes from potential to explicit, serving as the basis for pre-awakening timing indication.

[0057] Step 131: Perform temporal difference analysis on the cognitive attention tendency state change sequence of each user's field of vision edge node in the cognitive attention spatiotemporal evolution trend map, calculate the state change step size of the cognitive attention tendency state change sequence between adjacent time windows, and mark the time position when the state change step size enters the rapid growth stage from the steady change stage as the candidate mutation time position.

[0058] For the cognitive attention tendency state change sequence of each user's visual edge node in the spatiotemporal evolution trend map of cognitive attention, the absolute values ​​of the aggregation tendency description difference and the absolute values ​​of the directional convergence description difference between adjacent time window entries are calculated from front to back along the common time axis. These absolute values ​​are then divided by the time interval between the two adjacent time windows to obtain the aggregation tendency description change rate and the directional convergence description change rate. The aggregation tendency description change rate and the directional convergence description change rate are then weighted and merged, with the weighting coefficients preset according to the normalized weights of the aggregation tendency change rate and the directional convergence change rate. The weighted merged result serves as the state change step size between adjacent time windows.

[0059] Traverse all adjacent time window pairs of the cognitive attention tendency state change sequence to generate a state change step length sequence. Perform stage segmentation detection on this state change step length sequence, which is performed by calculating the mean and variance of the state change step length sequence within the sliding observation sub-window. In the steady change stage, the mean of the state change step length is less than a preset first step length threshold and the variance is less than a preset first variance threshold. In the rapid growth stage, the mean of the state change step length is greater than a preset second step length threshold or the variance is greater than a preset second variance threshold. When the state change step length sequence segment before a certain time position meets the characteristics of the steady change stage, and the immediately following state change step length sequence segment meets the characteristics of the rapid growth stage, that time position is marked as a candidate mutation time position.

[0060] Step 132: Extract the sequence fragments of cognitive attention tendency state change within adjacent time windows before and after the candidate mutation time position as background state fragments and mutation state fragments. Compare the deviation degree of the state change step size distribution pattern of the mutation state fragments with that of the background state fragments. Select the confirmed mutation peak positions from the candidate mutation time positions whose deviation degree meets the preset cognitive baseline mutation conditions.

[0061] For each candidate mutation time position, extract the cognitive attention tendency state change sequence segment extending forward by a preset first segment length of time window with the candidate mutation time position as the dividing point as the background state segment, and extend the cognitive attention tendency state change sequence segment backward by a preset second segment length of time window as the mutation state segment.

[0062] Calculate the statistical distribution characteristics of the state change step lengths in the background state segments, including the mean and standard deviation of the state change step lengths in the background state segments, and the statistical distribution characteristics of the state change step lengths in the abrupt state segments, including the mean and standard deviation of the state change step lengths in the abrupt state segments. Calculate the deviation value, which is the difference between the mean of the state change step lengths in the abrupt state segments and the mean of the state change step lengths in the background state segments, divided by the root mean square of the standard deviations of the state change step lengths in the background state segments and the abrupt state segments.

[0063] The calculated deviation value is compared with the preset cognitive baseline mutation condition threshold. When the deviation value is greater than or equal to the threshold, the candidate mutation time position is confirmed as the confirmed mutation peak position.

[0064] Step 133: Obtain the timestamp of the time window corresponding to the confirmed mutation peak position on the common time axis as the trigger time, and extract the node identifier of the user's field of vision edge node to which the confirmed mutation peak position belongs as the node to be woken up, and establish a pairwise mapping relationship between the trigger time and the node to be woken up.

[0065] Iterate through all confirmed mutation peak positions. For each confirmed mutation peak position, based on its time axis index and node identifier index in the spatiotemporal evolution trend chart of cognitive attention, read the end boundary timestamp of the corresponding time window on the common time axis as the trigger time, and read the node identifier string corresponding to the node identifier index as the node to be awakened. Create a paired mapping record containing a trigger time field and a node to be awakened field, and write the trigger time and the node to be awakened into the corresponding fields respectively.

[0066] Step 134: Combine the pairwise mapping relationship between the trigger time and the node to be woken up corresponding to each confirmed mutation peak position into a pre-wake-up timing indication for the guidance information requirement. The pre-wake-up timing indication includes the correspondence between the edge node that needs to perform the guidance pre-wake-up operation and the corresponding trigger time point.

[0067] All paired mapping records generated in step 133 are aggregated into a pre-wake-up timing indication list. Each record in the pre-wake-up timing indication list contains a node identifier to be woken up and a trigger time. The pre-wake-up timing indication list is arranged in ascending order of trigger time, representing the time scheduling arrangement for the guidance pre-wake-up operation to be performed on different edge nodes at different times.

[0068] Step 135: Based on the mutation state segment corresponding to the mutation peak position in the cognitive attention tendency state change sequence, generate the attention change trend slope of the confirmed mutation peak position, and convert the attention change trend slope into the pre-awakening intensity indicator.

[0069] Step 1351: Extract the mutation state segment corresponding to the confirmed mutation peak position from the cognitive attention tendency state change sequence, extract the aggregation tendency description change amount and the direction convergence description change amount experienced by the cognitive attention tendency state information in the mutation state segment from the mutation initiation state to the mutation peak state, and synthesize the aggregation tendency description change amount and the direction convergence description change amount into the cumulative amplitude of state rise.

[0070] For each confirmed mutation peak location, the aggregation tendency description value and directional convergence description value of the first time window entry arranged chronologically within the mutation state segment extracted in step 132 are obtained as the aggregation tendency description value and directional convergence description value of the mutation initiation state. The aggregation tendency description value and directional convergence description value of the last time window entry within the segment are obtained as the aggregation tendency description value and directional convergence description value of the mutation peak state. The difference between the aggregation tendency description value of the mutation peak state and the aggregation tendency description value of the mutation initiation state is calculated to obtain the aggregation tendency description change. The difference between the directional convergence description value of the mutation peak state and the directional convergence description value of the mutation initiation state is calculated to obtain the directional convergence description change. The cumulative amplitude of state rise is calculated as the square root of the sum of the square of the aggregation tendency description change and the square of the directional convergence description change, multiplied by a preset state synthesis scaling factor.

[0071] Step 1352: Obtain the timestamps corresponding to the start time window and the end time window of the mutation state segment, and calculate the time span between the start time window timestamp and the end time window timestamp as the rising duration span.

[0072] Obtain the end boundary timestamp corresponding to the first time window entry in the mutation state segment, and use it as the start time window timestamp. Obtain the end boundary timestamp corresponding to the last time window entry in the mutation state segment, and use it as the end time window timestamp. Calculate the difference between the end time window timestamp and the start time window timestamp to obtain the rising duration span.

[0073] Step 1353: Generate the slope of the attention change trend at the confirmed mutation peak position based on the cumulative magnitude of the state increase and the span of the increase duration. The slope of the attention change trend represents the rate at which the cognitive attention tendency state jumps from the background level to the peak level.

[0074] Dividing the cumulative increase in state by the duration of the increase yields the numerical value of the slope of the attention change trend. The physical meaning of the slope of the attention change trend trend is the rate at which the cognitive attention tendency state transitions from the background state to the peak state per unit time.

[0075] Step 1354: The slope of the attention change trend is converted into brightness adjustment trend direction description information that matches the brightness adjustment range of the digital wayfinding interface, as a pre-wake intensity indicator for the brightness adjustment of the wayfinding interface.

[0076] Obtain the maximum and minimum display brightness values ​​from the display characteristic information corresponding to the node to be woken up at the confirmed mutation peak position. Calculate the brightness adjustment range span value as the maximum display brightness value minus the minimum display brightness value. Input the slope of the attention change trend into a pre-calibrated slope mapping relationship. This mapping relationship is implemented through a piecewise linear mapping function. The input of the piecewise linear mapping function is the slope of the attention change trend, and the output is the allocation ratio of the brightness adjustment range span value. The value of this allocation ratio ranges from zero to one. Multiply the brightness adjustment range span value by the allocation ratio to obtain the absolute value of the brightness change to be adjusted. Combine the absolute value of the brightness change to be adjusted with the sign of the brightness adjustment trend direction to form the brightness adjustment trend direction description information. Use this brightness adjustment trend direction description information as the pre-wake-up intensity indicator for the confirmed mutation peak position and store it in association with the record of the node to be woken up in the corresponding pre-wake-up timing indicator.

[0077] Step 140: Using the pre-wake-up timing indicator and the pre-wake-up intensity indicator, create an optical flow particle motion field with directional gradient attributes in the digital wayfinding interface with the position corresponding to the edge node identifier as the anchor point, and embed the preset wayfinding identifier template into the optical flow particle motion field in a semantic enhancement manner to generate dynamic wayfinding response interface frames arranged in time sequence.

[0078] The optical flow particle motion field is a dynamic particle swarm generated from edge nodes as emission origins, following historical behavior trajectories and evolving with time. Particle brightness is modulated according to pre-wake intensity, creating a flowing light effect with visual pull through directional gradient attributes, guiding user attention to the target node. The dynamic wayfinding response interface frame is a sequence of image frames synthesized by embedding the wayfinding sign template frame-by-frame with the optical flow particle motion field animation after contour halo enhancement and hue shift processing. The dynamic flow of optical particles is superimposed on semantically enhanced wayfinding signs, triggering pre-wake display at corresponding edge nodes.

[0079] Step 141: Extract the identifier of the node to be woken up and the corresponding trigger time contained in the pre-wake-up timing indication; obtain the anchor point coordinates of the node to be woken up in the digital navigation interface from the user's field of view edge node layout based on the identifier of the node to be woken up; and use the trigger time as the rendering start time point of the dynamic navigation response interface frame sequence.

[0080] Iterate through each record in the pre-wake-up timing indicator list, extracting the node identifier field and trigger time field. Using the node identifier as the query key, retrieve the matching node record from the structured configuration data of the user's field of view edge node layout, and read the coordinate value of the node center point coordinate field from that node record as the anchor point position coordinate. Set the trigger time as the rendering start time point of the dynamic navigation response interface frame sequence corresponding to the node to be woken up, and the timeline after the rendering start time point will generate dynamic navigation response interface frames frame by frame.

[0081] Step 142: Extract behavior trajectory segments from the forward-looking behavior trace data stream that are within a preset retrospective observation interval before the rendering start time point, extract the two-dimensional projection point coordinates and movement direction of each behavior trajectory sampling point on the digital navigation interface, and generate a historical behavior trajectory direction reference sequence to guide the movement direction of optical flow particles.

[0082] A fixed time length is set for the retrospective observation interval. The retrospective starting time is obtained by retrospectively observing this fixed time length from the rendering start time. Behavior trajectory sampling points whose acquisition timestamps are located between the retrospective starting time and the rendering start time are selected from the forward-looking behavior trace data stream to form behavior trajectory segments. For each behavior trajectory sampling point in a behavior trajectory segment, the three-dimensional spatial coordinate vector of the behavior trajectory sampling point is mapped to two-dimensional projection point coordinates on the digital navigation interface using the perspective projection mapping relationship established in step 123. The movement direction corresponding to the behavior trajectory sampling point is obtained using the inter-frame motion pointing analysis method in step 124. The two-dimensional projection point coordinates and movement direction are combined into a direction reference entry. The direction reference entries corresponding to all behavior trajectory sampling points in the behavior trajectory segment are arranged in ascending order by acquisition timestamp to obtain the historical behavior trajectory direction reference sequence.

[0083] Step 143: Convert the movement pointing direction in the historical behavior trajectory direction reference sequence into the initial emission direction vector of the optical flow particle in the digital navigation interface, and construct a description of the deflection evolution of the optical flow particle emission direction over time based on the direction change angle of the movement pointing direction within the preset retrospective observation interval.

[0084] The movement direction of the first direction reference entry in the historical behavior trajectory direction reference sequence is extracted as the initial launch direction reference direction. This movement direction is then normalized to obtain the initial launch direction vector, which is represented as a two-dimensional unit vector in the pixel coordinate system of the digital navigation interface. The angle change between the movement directions of two adjacent direction reference entries in the historical behavior trajectory direction reference sequence is calculated sequentially. This angle change is obtained by calculating the inverse cosine of the dot product of the unit vectors of the two movement directions. All angle change values ​​are arranged in chronological order to form a direction change angle sequence.

[0085] A deflection evolution description is constructed based on the direction change angle sequence. The deflection evolution description specifies that in each unit time step after emission, the current motion direction vector of the optical flow particle rotates around the current time direction by a specified angle. The rotation angle value is taken from the angle change value of the corresponding time step position in the direction change angle sequence. When the angle values ​​in the direction change angle sequence are exhausted, the rotation angle of the subsequent time step adopts the decayed value of the last angle value.

[0086] Step 144: Using the anchor point coordinates as the particle emission origin, set the motion direction trajectory for each optical flow particle according to the initial emission direction vector and the deflection evolution description, so that the optical flow particles move in the digital navigation interface along the motion direction trajectory after starting from the anchor point coordinates.

[0087] The anchor point coordinates are set as the origin coordinates of the particle emission in the optical flow particle motion field. At the origin coordinates, optical flow particles are generated in batches according to the preset total number of particle emissions and the particle emission time interval. Each optical flow particle is assigned a unique particle identifier and emission time upon generation. The motion direction of the optical flow particle at the emission time is set as the initial emission direction vector. In each subsequent unit time step, the motion direction of the optical flow particle is updated according to the deflection evolution description. The update method is to rotate the current motion direction vector around itself by the corresponding deflection angle, which is obtained from the deflection evolution description according to the index of the current time step.

[0088] Simultaneously, the instantaneous position coordinates of each optical flow particle are calculated for each unit time step. The instantaneous position coordinates are calculated by adding the particle's instantaneous position coordinates from the previous unit time step to the motion direction vector of the current unit time step, multiplied by a preset particle motion step size. Starting from the particle emission origin coordinates, each optical flow particle forms its own motion direction trajectory according to the instantaneous position coordinates updated unit by unit time step. To enhance the naturalness of the motion direction trajectory, a random angle perturbation is applied to the motion direction vector each time the motion direction is updated. The amplitude of the random angle perturbation follows a zero-mean Gaussian distribution, and the variance is preset as the direction perturbation variance parameter.

[0089] Step 145: Determine the brightness modulation mode of the optical flow particles during their movement based on the brightness adjustment trend direction description information in the pre-wake intensity indication.

[0090] The brightness adjustment trend direction description information is read from the associated storage pre-wake intensity indication. This description includes the absolute value of the proposed brightness change and a trend direction sign. A positive trend direction sign indicates that the brightness has gradually increased since the particle emission time, while a negative sign indicates that the brightness has gradually decreased. The brightness modulation function type is determined based on the trend direction sign; it can be either an increasing or decreasing type. The brightness modulation function uses the elapsed time of the optical flow particle as its independent variable, and its output is the particle brightness value at the current moment.

[0091] When the trend direction is positive, the particle brightness value monotonically increases from the base brightness value with increasing movement time. This increase is achieved through a linear or S-shaped growth curve, and the upper limit of the particle brightness value is constrained by the maximum display brightness value of the node to be woken up. When the trend direction is negative, the particle brightness value monotonically decreases from an initial higher brightness value with increasing movement time. This decrease is achieved through a linear or exponential decay curve, and the lower limit of the particle brightness value is constrained by the minimum display brightness value of the node to be woken up. The total brightness change involved in the brightness modulation function is driven by the absolute value of the brightness change to be adjusted. The duration of the brightness change process is determined by the product of the total number of frames in the dynamic guidance response interface frame sequence and the display duration of each frame, ensuring that the brightness of the optical flow particles completes a full modulation amplitude change throughout the entire dynamic guidance response period.

[0092] Step 146: Based on the particle emission origin, the motion direction trajectory, and the brightness modulation method, create a parameterized motion description of optical flow particles in the coordinate system of the digital navigation interface. The parameterized motion description of optical flow particles defines the instantaneous position coordinates and instantaneous brightness value of each optical flow particle at any moment of motion.

[0093] For each optical flow particle, the particle emission origin coordinates, the emission time assigned to the particle, and the particle identifier are recorded as the initialization parameters of the optical flow particle. Based on the trajectory calculation method for the optical flow particle's motion direction in step 144, a sequence of instantaneous position coordinates for all unit time steps from the particle emission time to the preset end time of the particle's lifetime is generated. Based on the brightness modulation method determined in step 145, a sequence of instantaneous brightness values ​​for the optical flow particle at each unit time step is calculated. The instantaneous position coordinate sequence and the instantaneous brightness value sequence are mapped in pairs according to the unit time step index to form the particle time state sequence of the optical flow particle. The set of particle time state sequences of all optical flow particles constitutes the parameterized motion description of the optical flow particle.

[0094] Step 147: The parameterized motion description of the optical flow particles is sampled at a preset particle motion frame rate to generate a particle state sequence consisting of the particle position and particle brightness of each optical flow particle at each discrete motion moment. The particle state sequences of all optical flow particles are superimposed and synthesized to obtain the optical flow particle motion field animation frame sequence.

[0095] Set the particle motion frame rate, which should be consistent with the rendering frame rate of the dynamic navigation response interface. Starting from the rendering start time, generate a discrete sequence of sampling time points on the time axis, using the reciprocal of the duration of each frame as the sampling time interval. For each sampling time point, traverse all optical flow particles in the parameterized motion description of the optical flow particles. For each optical flow particle, based on the distribution of its particle time state sequence on the time axis, obtain the instantaneous position coordinates and instantaneous brightness value of the optical flow particle at that sampling time point through linear interpolation in the time dimension. The interpolation calculation of the instantaneous position coordinates is based on the instantaneous position coordinates of two adjacent unit time steps before and after the sampling time point, proportional to the time interval. The interpolation calculation of the instantaneous brightness value is similar.

[0096] The instantaneous position coordinates and instantaneous brightness values ​​of all optical flow particles at the same sampling time point are written into the single-frame particle state map corresponding to that sampling time point. The canvas size of the particle state map is consistent with the pixel size of the digital navigation interface, and the pixel values ​​of non-particle pixel positions in the particle state map are set to transparent. The single-frame particle state maps corresponding to all sampling time points are arranged in chronological order of the sampling time points to form a sequence of animation frames for the optical flow particle motion field.

[0097] Step 148: Select the wayfinding sign template associated with each node to be woken up from the preset wayfinding sign template library and perform semantic enhancement embedding to obtain the dynamic wayfinding response interface frame.

[0098] Step 1481: Access the preset wayfinding sign template library, select the wayfinding sign template associated with the node to be woken up according to the node identifier to be woken up, and separate the wayfinding graphic elements and wayfinding text elements in the wayfinding sign template into independent renderable layers.

[0099] The pre-defined wayfinding sign template library is stored in the persistent storage space of the user behavior response method server. The template data in the library is organized using the identifier of the node to be woken up as the index key. The wayfinding sign template library is queried based on the identifier of the node to be woken up to retrieve the corresponding wayfinding sign template object. The wayfinding sign template object internally contains a wayfinding graphic element data block and a wayfinding text element data block. The wayfinding graphic element data block stores the vector path data and fill color information of the wayfinding sign, while the wayfinding text element data block stores the string content and font style attributes of the wayfinding text.

[0100] The guide view graphic element data block is parsed into a first renderable layer, which records the outline path control point sequence and closed path fill rules of the guide view graphic elements. The guide view text element data block is parsed into a second renderable layer, which records the text rendering baseline position, font size, letter spacing, and font color attributes. The first and second renderable layers are spatially offset relative to each other, using the anchor point coordinates as the layer center reference point.

[0101] Step 1482: Perform edge detection processing on the guide graphic element to extract the contour boundary path, expand outward along the contour boundary path to generate a halo region, and determine the color saturation parameter and transparency parameter of the halo region based on the pre-wake intensity indication. Superimpose the halo region onto the guide graphic element to form a contour halo enhanced guide graphic element, and perform hue shift processing on the guide text element to obtain a color enhanced guide text element.

[0102] Canney edge detection is performed on the guide view graphic elements in the first renderable layer. This process includes four sub-steps: Gaussian smoothing filtering, gradient magnitude and direction calculation, non-maximum suppression, and double-threshold hysteresis connection. The final result is the extraction of the contour boundary paths of the guide view graphic elements, represented by a sequence of closed polygon points. A halo region is generated by extending outwards from each vertex along the contour boundary path. The extension method involves calculating the normal direction of the contour at each vertex, and then shifting outwards along the normal direction by a predetermined halo width distance to obtain the extended vertices. All extended vertices are connected sequentially to form the outer boundary path of the halo. The inner boundary path of the halo is the contour boundary path itself, and the area enclosed by the inner and outer boundary paths constitutes the halo region.

[0103] The absolute value of the proposed brightness adjustment change is read from the brightness adjustment trend direction description information in the pre-wake intensity indicator. This absolute value is then mapped to the color saturation and transparency parameter ranges. The mapping method involves dividing the absolute value of the proposed brightness adjustment change by the maximum display brightness value to obtain a normalized intensity ratio. The color saturation parameter is set to the normalized intensity ratio multiplied by the maximum color saturation value, and the transparency parameter is set to one minus the normalized intensity ratio multiplied by the maximum transparency value. The color saturation of each pixel in the halo area is set to the color saturation parameter, and the transparency is set to the transparency parameter, forming a semi-transparent gradient halo layer.

[0104] A semi-transparent gradient halo layer is overlaid onto the first renderable layer using alpha blending. The overlay calculation is as follows: the color value of the overlaid pixel is the color value of the first renderable layer pixel multiplied by a first blending coefficient, plus the color value of the semi-transparent gradient halo layer pixel multiplied by a second blending coefficient. The sum of the first and second blending coefficients is one, resulting in an outline-enhanced guide visual element. A hue shift is performed on the guide text element in the second renderable layer. This hue shift is achieved by rotating the font color attribute on the color wheel by a preset hue shift angle to obtain a color-enhanced font color. This color-enhanced font color replaces the font color attribute in the second renderable layer, while the rendering attributes of the remaining text in this layer remain unchanged, resulting in a color-enhanced guide text element.

[0105] Step 1483: The contour halo enhanced guide graphic element and the color enhanced guide text element are embedded frame by frame into the screen area corresponding to the anchor point position coordinates in the optical flow particle motion field animation frame sequence. Through frame-by-frame synthesis processing, a dynamic guide response interface frame is generated in which the optical flow particles at the anchor point position dynamically flow along the historical behavior trajectory direction and the guide sign has contour halo and color enhanced visual effects.

[0106] Using the anchor point coordinates as the embedding reference point, the contour halo-enhanced guide visual elements and color-enhanced guide text elements are composited into layers. During compositing, the relative offset relationship and layer stacking order of the guide visual elements and guide text elements within the pre-defined guide sign template are maintained, forming an enhanced guide sign layer. Each frame of the optical flow particle motion field animation frame sequence is traversed, using the single-frame particle state image as the bottom layer and the enhanced guide sign layer as the top layer. The enhanced guide sign layer is tiled onto the single-frame particle state image, centered on the anchor point coordinates. The two layers are then composited using alpha blending. During compositing, the transparency attribute of pixels in the enhanced guide sign layer determines its blending weight with the corresponding pixels in the bottom layer.

[0107] After compositing, the visual effects of optical flow particles and enhanced wayfinding signage coexist in the region corresponding to the anchor point coordinates in the single-frame particle state diagram. While the optical flow particles dynamically flow along their historical trajectory, the wayfinding signage exhibits a visual state of outline halo and enhanced color. Arranging all frames after frame-by-frame compositing according to the original chronological order of the optical flow particle motion field animation frame sequence yields the dynamic wayfinding response interface frame sequence.

[0108] Step 150: Distribute the dynamic wayfinding response interface frame sequence to the corresponding user field edge nodes according to the edge node identifier, and trigger the user field edge nodes to perform wayfinding pre-wake-up display operation.

[0109] Step 151: Read the edge node identifier associated with each wayfinding response interface frame from the dynamic wayfinding response interface frame sequence, and use the edge node identifier as the grouping basis to split the dynamic wayfinding response interface frame sequence into an independent frame sequence that corresponds one-to-one with each edge node identifier.

[0110] When generating the dynamic navigation response interface frame sequence, each frame contains an embedded metadata field for the associated node identifier to be woken up. The metadata field for the node identifier to be woken up is extracted from each frame, and the edge node identifier is read. Using the edge node identifier as the grouping key, all frame data is hashed and grouped. Frame data with the same edge node identifier are grouped into the same frame queue. The frame data in each queue are arranged in ascending order by timestamp, forming an independent frame sequence that corresponds one-to-one with each edge node identifier.

[0111] Step 152: Obtain the origin coordinates, width, and height of the display area for each edge node, establish the rectangular boundary description of the display area for each edge node, and obtain the display color configuration file for each edge node.

[0112] Using the node identifier of each edge node as the query key, the node center point coordinate field, node display area width field, and node display area height field are read from the structured configuration data of the edge node layout in the user's field of view.

[0113] Based on the node's center point coordinates, the width and height of the display area, the coordinates of the top-left origin and bottom-right endpoint of the node's display area are calculated. The top-left origin coordinate is equal to the node's center point x-coordinate minus half the node's display area width and the node's center point y-coordinate minus half the node's display area height. The bottom-right endpoint coordinate is equal to the node's center point x-coordinate plus half the node's display area width and the node's center point y-coordinate plus half the node's display area height. The top-left origin and bottom-right endpoint coordinates are combined and recorded as the rectangular boundary description of the node's display area. Simultaneously, the color gamut descriptor, contrast ratio, and preset International Color Consortium color characteristic file identifier are extracted from the display characteristic data block field of the edge node. The corresponding display color configuration file is loaded, containing a color lookup table from the digital navigation interface standard color space to the edge node's physical display panel color space and a set of gamma correction parameters.

[0114] Step 153: For each edge node, perform spatial coordinate transformation processing on the global canvas coordinate system of each frame image in the independent frame sequence assigned to the edge node and the node display area rectangle boundary description of the edge node to determine the corresponding sub-region coordinate range of the node display area rectangle boundary in each frame image.

[0115] In the independent frame sequence, the default canvas coordinate system of each frame is the same as the pixel coordinate system of the digital navigation interface, and the canvas size is consistent with the pixel size of the digital navigation interface. For each edge node, the coordinates of the upper left origin and the lower right end point of the rectangular boundary of the node display area are directly applied to each frame in the independent frame sequence. A rectangular sub-region is delineated in the canvas coordinate system with the upper left origin coordinates and the lower right end point coordinates as its boundaries. The coordinate range of this rectangular sub-region is the coordinate range of the corresponding sub-region.

[0116] Step 154: Perform pixel-level cropping operation on each frame image according to the coordinate range of the corresponding sub-region, and extract sub-image blocks that completely match the rectangular boundary of the node display area. Combine all the extracted sub-image blocks in frame order to generate a local guide response interface sub-sequence frame that matches the size of the edge node display area.

[0117] For each frame of image data in an independent frame sequence, using the coordinate range of its corresponding sub-region as the cropping boundary, all pixel values ​​within this cropping boundary are read to generate a new sub-image block with a size equal to the width multiplied by the height of the node display area. The pixel arrangement order of the sub-image block remains consistent with the pixel arrangement order of the corresponding region in the original frame image. The sub-image blocks obtained from cropping multiple consecutive frames are arranged in the original frame order to form a local navigation response interface sub-sequence frame.

[0118] Step 155: Perform continuity detection on the motion trajectory of optical flow particles between adjacent sub-image blocks in the local guide response interface sub-sequence frame to restore the continuity of the motion direction of optical flow particles when they cross the clipping boundary.

[0119] For every two temporally adjacent sub-image blocks in the local navigation response interface sub-sequence, the pixel brightness values ​​of optical flow particles are extracted at the same pixel coordinate positions in the two sub-image blocks. The gradient vectors of the optical flow particle pixel positions in the previous sub-image block are calculated in the horizontal and vertical directions. Based on the gradient vectors, the possible location region of the optical flow particles in the next sub-image block is predicted. Within the predicted location region of the next sub-image block, the corresponding optical flow particle pixels that match the brightness features and local texture features of the optical flow particles in the previous sub-image block are searched.

[0120] When a corresponding optical flow particle pixel is matched in the next frame sub-image block, the angle between the motion direction of the optical flow particle pixel in the previous frame sub-image block and the motion direction in the next frame sub-image block is checked. If the angle exceeds the preset continuity angle threshold, the motion direction of the current frame is weighted and corrected according to the motion directions of the previous and next frames. The correction direction is the normalized weighted vector sum and direction of the previous and next motion directions, thereby restoring the continuity of the motion direction of the optical flow particle at the cropping boundary of the sub-image block.

[0121] Step 156: Use the display color profile corresponding to the edge node to perform color space mapping and gamma correction on each frame of the repaired local guide response interface sub-sequence frame, and encode and package it according to the data transmission protocol supported by the edge node to generate a node push data packet.

[0122] The color lookup table in the display color profile of the edge node is loaded, and color space mapping is performed on each pixel of each sub-image block in each frame of the repaired local wayfinding response interface sub-sequence. The color space mapping process uses the pixel's color representation value in the standard color space of the digital wayfinding interface as input to look up the color lookup table, and outputs the corresponding color representation value of the pixel in the color space of the physical display panel of the edge node. Then, gamma correction is performed on each color channel value of each pixel using the gamma correction parameters in the display color profile. The gamma correction is calculated by normalizing each color channel value, raising its gamma correction parameter to a power, and then inversely normalizing it back to the original value range.

[0123] After completing color space mapping and gamma correction processing one by one, a color-corrected sub-image block frame sequence is obtained. Based on the data transmission protocol type supported by the edge node, the color-corrected sub-image block frame sequence is encoded and packaged. If the protocol is an image frame stream transmission protocol, each frame is encoded as an independent network abstraction layer unit, and a metadata header containing a frame sequence number, timestamp, and timestamp alignment information is inserted before each frame. If the protocol is a video stream transmission protocol, inter-frame predictive coding and entropy coding are performed on the sub-image block frame sequence to generate a compressed video elementary stream, and an inter-frame decoding dependency description is embedded in the structure of the compressed video elementary stream. The encoded and packaged transmission data is then encapsulated into node push data packets.

[0124] Step 157: Transmit the node push data packet of each edge node to the corresponding edge node. After receiving the corresponding node push data packet, control each edge node to parse the frame sequence timestamp alignment information and inter-frame decoding dependency description in the node push data packet. According to the inter-frame decoding dependency description, decode each frame of local guide response interface sub-sequence frame in sequence, and present each frame in the display area of ​​the edge node in the order of timestamps.

[0125] Through the underlying communication network of the digital wayfinding interface, each edge node sends node push data packets via unicast to its corresponding network address. Upon receiving the push data packets, the edge node unpacks them, extracting frame sequence timestamp alignment information, inter-frame decoding dependency descriptions, and encoded frame data. The built-in decoder unit of the edge node constructs an inter-frame reference frame list and decoding order based on the inter-frame decoding dependency description. It then performs decoding operations frame by frame on the encoded frame data. If inter-frame prediction coding frames exist during decoding, the decoded reference frame data is retrieved according to the reference frame list to perform inter-frame prediction compensation and reconstruction on the current frame. For each decoded local wayfinding response interface sub-sequence frame image, the image is written to the edge node's frame buffer. The edge node's display controller reads the frame image from the frame buffer at the correct presentation time based on the frame sequence timestamp alignment information and drives the display panel to present the frame within the edge node's display area. When all frames in the node push data packets have been decoded and presented sequentially, the wayfinding pre-wake-up display operation is complete.

[0126] As an optional embodiment, after step 150, the method further includes: Step 161: Receive the subsequent behavioral trajectory data of the target user after the pre-wake-up display operation of the wayfinding system collected by the non-contact sensing device, and generate a post-wake-up behavioral trace data stream; convert the post-wake-up behavioral trace data stream into a post-wake-up two-dimensional projection point coordinate sequence on the digital wayfinding interface through perspective projection mapping relationship.

[0127] After the pre-wake-up display operation begins, the non-contact sensing device continues to collect the target user's behavioral trajectory data within the three-dimensional passageway, acquiring newly collected skeletal joint spatial position data and somatosensory contour point cloud data, and generating a post-wake-up behavioral trace data stream in the same manner as in step 110. The data recording format of the post-wake-up behavioral trace data stream is consistent with that of the forward-facing behavioral trace data stream, including the acquisition timestamp, sampling point identifier, three-dimensional spatial position coordinate vector, instantaneous movement direction angle, instantaneous movement speed, and movement acceleration magnitude. The perspective projection mapping relationship constructed in step 123 is invoked to map each behavioral trajectory sampling point in the post-wake-up behavioral trace data stream to two-dimensional projection point coordinates on the digital navigation interface, resulting in a post-wake-up two-dimensional projection point coordinate sequence.

[0128] Step 162: Perform continuous sequential comparison of the movement pointing directions corresponding to adjacent timestamps in the post-wake-up two-dimensional projection point coordinate sequence. When it is identified that the movement pointing direction presents an alternating change pattern of facing and moving away from the edge node of the user's field of vision in multiple consecutive time intervals, generate cognitive hesitation state indication information.

[0129] Referring to step 124, the movement direction between adjacent timestamps is calculated pairwise for the coordinate sequence of the post-wake-up 2D projection points to obtain a movement direction sequence. The movement direction of multiple consecutive time intervals in the movement direction sequence is read sequentially, and it is determined whether the movement direction of each time interval is towards or away from the user's visual field edge node. The determination of orientation or deviation is made by calculating the intersection of the reverse extension of the movement direction with the display area boundary of the nearest user's visual field edge node. If an intersection exists, it is an orientation type; otherwise, it is a deviation type. The number of times the orientation and deviation types alternate in consecutive time intervals is recorded. When the number of alternations exceeds a preset threshold within a preset observation time window, the target user is determined to be in a state of cognitive hesitation, and cognitive hesitation state indication information is generated. This cognitive hesitation state indication information includes the start timetamp of the cognitive hesitation state, the end timetamp of the cognitive hesitation state, and the identifier of the user's visual field edge node involved.

[0130] Step 163: Based on the cognitive hesitation state indication information, select a simplified wayfinding sign template with a higher visual abstraction level than the currently used wayfinding sign template from the preset wayfinding sign template library, and call the edge detection processing and hue shift processing corresponding to the semantic enhancement embedding to transform the simplified wayfinding sign template into a simplified enhanced wayfinding element.

[0131] In the preset wayfinding sign template library, each wayfinding sign template, in addition to the associated node identifier, also stores a visual abstraction level attribute. The visual abstraction level attribute is a set of preset discrete level values. The higher the level value, the simpler the wayfinding graphic elements and the less detailed information. Based on the user's visual field edge node identifier involved in the cognitive hesitation state indication information, multiple wayfinding sign templates with visual abstraction levels associated with that node are obtained from the wayfinding sign template library. The simplified wayfinding sign template with a visual abstraction level higher than the template used in the current dynamic wayfinding response interface frame and the smallest difference in visual abstraction levels between the two is selected. The edge detection processing described in step 1482 is performed on the wayfinding graphic elements in the simplified wayfinding sign template to generate contour boundary paths. A halo area is generated by expanding outward along the contour boundary path, and a halo layer is generated and superimposed according to the determination method of the color saturation parameters and transparency parameters in step 1482 to obtain simplified contour halo enhanced wayfinding graphic elements. Hue shift processing is performed on its wayfinding text elements to obtain simplified color enhanced wayfinding text elements. Combine the simplified outline halo enhanced guide graphic element and the simplified color enhanced guide text element into a simplified enhanced guide element.

[0132] Step 164: Replace the original guide graphic elements and original guide text elements located at the anchor point coordinates in the optical flow particle motion field animation frame sequence with the simplified and enhanced guide elements frame by frame, synthesize and generate an optimized dynamic guide response interface frame sequence adapted to the state of cognitive hesitation, and distribute the optimized dynamic guide response interface frame sequence to the user's field of vision edge node.

[0133] In the optical flow particle motion field animation frame sequence, locate the original enhanced wayfinding label layer at the anchor point coordinates within each frame. Replace the outline halo enhanced wayfinding graphic elements and color enhanced wayfinding text elements in the original enhanced wayfinding label layer with simplified outline halo enhanced wayfinding graphic elements and simplified color enhanced wayfinding text elements from the simplified enhanced wayfinding elements, respectively, while maintaining the spatial position, scaling ratio, and layer stacking relationship of the wayfinding elements before and after the replacement. The replaced frames form an optimized dynamic wayfinding response interface frame sequence adapted to cognitive hesitation states. Following the distribution process in step 150, the optimized dynamic wayfinding response interface frame sequence is packaged into a node push data package and transmitted to the corresponding user's field of vision edge node, triggering the edge node to update and present the wayfinding pre-wake display content.

[0134] As an optional embodiment, after step 150, the method further includes: Step 210: Continuously receive behavioral trajectory data of associated users within the three-dimensional passage space, generate associated forward behavioral trace data stream, and generate a spatiotemporal evolution trend map of associated cognitive attention based on the layout of user field of vision edge nodes.

[0135] While the target user is present in the three-dimensional passage space, the non-contact sensing device simultaneously collects behavioral trajectory data of other associated users within the same three-dimensional passage space. Following the method in step 110, a corresponding associated forward behavioral trace data stream is generated for each associated user. For each associated user, the spatiotemporal mapping processing flow in step 120 is executed to generate a spatiotemporal evolution trend map of associated cognitive attention corresponding to each user's visual field edge nodes.

[0136] Step 220: Extract the associated pre-awakening timing indicator and the associated pre-awakening intensity indicator from the spatiotemporal evolution trend chart of the associated cognitive attention.

[0137] For each associated user's spatiotemporal evolution trend map of associated cognitive attention, the mutation peak analysis process in step 130 is executed to obtain the associated pre-awakening timing indicator and associated pre-awakening intensity indicator for each associated user. The data structure of the associated pre-awakening timing indicator and associated pre-awakening intensity indicator is consistent with the corresponding indicator data structure of the target user.

[0138] Step 230: Compare the edge node identifier in the associated pre-wake-up timing indication with the edge node identifier occupied by the target user. When the identifier matches and the associated trigger time falls within the presentation period of the target user's dynamic navigation response interface frame sequence, a multi-user display conflict event is generated.

[0139] The program retrieves the presentation time of the current dynamic navigation response interface frame sequence for the target user. This presentation time is defined by the rendering start time and the end time calculated by multiplying the number of frames in the frame sequence by the display duration of each frame. It then iterates through all records in the associated pre-wake-up timing indicator list, extracting the wake-up node identifier and trigger time for each record. Finally, it performs an exact string match between the extracted wake-up node identifier and the wake-up node identifier corresponding to the target user's dynamic navigation response interface frame sequence.

[0140] When a record with a matching identifier is found, it is further determined whether the trigger time of the record falls within the presentation period of the target user's dynamic navigation response interface frame sequence. The determination method is whether the trigger time is greater than or equal to the rendering start time and less than the end time. If it meets the requirements, a multi-user display conflict event record is generated. The multi-user display conflict event record includes the conflict edge node identifier, the target user's occupation start time, the target user's occupation end time, the associated user's trigger time, and the associated user's identifier.

[0141] Step 240: In response to the multi-user display conflict event, based on the intensity difference tendency between the pre-wake intensity indication of the target user and the associated pre-wake intensity indication, the particle emission start time of the optical flow particle motion field associated with the conflict edge node is delayed until after the dynamic navigation response interface frame sequence of the target user is presented.

[0142] For each multi-user display conflict event record, obtain the brightness adjustment trend direction description information from the pre-wake intensity indicator of the target user at the conflict edge node, and extract the absolute value of the proposed brightness adjustment change as the target intensity value. Obtain the absolute value of the proposed brightness adjustment change from the associated pre-wake intensity indicator of the conflict-related users at the conflict edge node, and use it as the association intensity value. Calculate the difference between the association intensity value and the target intensity value. When the difference is greater than zero, the intensity difference tendency is determined to be association intensity priority; when the difference is less than zero, the intensity difference tendency is determined to be target intensity priority; when the difference is equal to zero, it is determined to be equal intensity.

[0143] When the intensity difference tendency prioritizes the target intensity, the optical flow particle motion field of the associated user at the conflict edge node maintains its original rendering time schedule. When the intensity difference tendency prioritizes the associated intensity, the particle emission start time of the optical flow particle motion field of the associated user at the conflict edge node is set to the first available frame time point after the end time point of the target user's dynamic navigation response interface frame sequence. The delay in particle emission start time is the time difference between this end time point and the original trigger time of the associated pre-wake-up timing indication. When the intensity difference tendency is equal intensity, the particle emission start time is similarly delayed until after the end of the target user's dynamic navigation response interface frame sequence presentation.

[0144] Based on the adjusted particle emission start time, the optical flow particle motion field animation frame sequence and dynamic guidance response interface frame sequence corresponding to the conflict edge node are regenerated for the associated users, and the regenerated dynamic guidance response interface frame sequence is distributed to the conflict edge node according to the delayed time node.

[0145] In the above technical solution, the thresholds on which the state change step size is divided into stages are adaptively determined based on the statistical distribution characteristics of the cognitive attention tendency state change sequence within the historical operating cycle.

[0146] Specifically, before processing the behavioral trajectory data of each target user, the user behavior response method server acquires multiple historical user cognitive attention tendency state change sequence samples that have completed a full passage within the current 3D passage space. The state change step length data of these historical samples are aggregated, and a probability density distribution curve of the state change step length data is plotted. This probability density distribution curve is then smoothed using a kernel density estimation method. Two main peak modes are identified on the smoothed probability density distribution curve: the first peak mode corresponds to the step length concentration region during the stable change phase, and the second peak mode corresponds to the step length concentration region during the rapid growth phase. The step length value corresponding to the probability density minimum point between the two peak modes is used as the benchmark reference value for the first and second step length thresholds. The benchmark reference values ​​are offset and adjusted according to the probability density decrease slope on both sides of the probability density minimum point to obtain the first and second step length thresholds applicable to the current 3D passage space environment.

[0147] Meanwhile, the variance of all state change steps in the stable change phase in the historical samples is calculated, and the percentile value of the above variance is taken as the first variance threshold. The variance of all state change steps in the rapid growth phase is calculated, and the percentile value of the above variance is taken as the second variance threshold. This adaptive determination process ensures that the phase division threshold always maintains statistical consistency with the actual usage characteristics of the three-dimensional passage space and the behavioral patterns of the user group, without relying on fixed empirical values.

[0148] When calculating the cumulative magnitude of state rise, the aggregation tendency description and the directional convergence description are derived from observation dimensions with different dimensions. In order to unify the dimensions of the two for meaningful composite calculation, before each confirmation of abrupt change peak, the aggregation tendency description value and the directional convergence description value of all user vision edge nodes in the spatiotemporal evolution trend map are globally standardized within all time windows.

[0149] The specific implementation of global standardization is as follows: calculate the mean and standard deviation of the aggregation tendency description value over all samples, subtract the mean from each aggregation tendency description value and divide by the standard deviation to obtain the standardized aggregation tendency description value; at the same time, calculate the mean and standard deviation of the direction convergence description value over all samples, subtract the mean from each direction convergence description value and divide by the standard deviation to obtain the standardized direction convergence description value.

[0150] After standardization, the change in aggregation tendency description is calculated based on the standardized aggregation tendency description value, and the change in directional convergence description is calculated based on the standardized directional convergence description value. The cumulative magnitude of state ascent is then calculated based on this, with a preset state synthesis scaling factor set to one. This square root of the sum of squares synthesis method ensures that the two equally weighted dimensions contribute on a unified scale after standardization. Through this standardization preprocessing, this technical solution eliminates the imbalance in synthesis weights caused by the difference in dimensions between the aggregation tendency description and the directional convergence description.

[0151] To address the issue of particle motion direction guidance after the exhaustion of the direction change angle sequence values ​​in the description of optical flow particle deflection evolution, the rotation angle decay process in subsequent time steps is driven by a rate decay function after the last angle value obtained from the direction change angle sequence is applied to the corresponding time step of the optical flow particle. Specifically, the rate decay function is expressed as follows: a decay time constant is set, which is related to the preset backtracking observation interval length and is equal to the backtracking observation interval length multiplied by a preset decay scaling factor. Using the last angle value as the initial decay value, in each subsequent time step, the current remaining angle value is multiplied by a decay factor calculated based on the decay time constant and the time step length. The decay factor is calculated as the ratio of the negative time step length of the natural constant e to the decay time constant raised to the power of its value.

[0152] After multiple decay iterations, when the remaining angle value is lower than the preset angle truncation threshold, the rotation angle of subsequent time steps is directly set to zero. The optical flow particles then maintain linear motion until the end of their lifetime. This decay mechanism allows the deflection amplitude of the optical flow particles to naturally decrease and stabilize over time after losing the historical behavior trajectory direction reference. After the historical reference ends, its motion trajectory transitions from gradually weakening directional change motion to uniform linear motion, ensuring that the motion behavior of the optical flow particles has a consistent and predictable physical performance throughout its lifetime.

[0153] The system identifies the cognitive hesitation state indication information generated after recognizing the alternating change pattern of the direction of movement towards and away from the edge node of the user's field of vision within a continuous time interval. It establishes the correspondence between this information and the selection of a simplified wayfinding sign template with higher visual abstraction, and establishes a wayfinding design strategy based on cognitive load adjustment.

[0154] When a target user experiences difficulty processing the signage presented in a digital wayfinding interface, their gaze will repeatedly and rapidly shift between the signage area and other interface areas. This gaze shift is reflected in the trajectory projection as an alternating pattern of the direction of movement towards and away from edge nodes. After capturing this gaze shift pattern by setting a preset threshold for the number of alternations, it is determined that the user is experiencing a high level of visual cognitive load.

[0155] In this state, continuing to present wayfinding signage elements with rich details and complex visual hierarchies would further increase the cognitive processing pressure on users. Therefore, a strategy of enhancing visual abstraction by reducing redundant decorative details and improving the generalization of graphic outlines is adopted. The simplified wayfinding signage template with higher visual abstraction retains the core information structure and semantic content of the original wayfinding signage while removing secondary visual features that might interfere with rapid recognition, thereby reducing the difficulty of information decoding and visual search time for users. The connection between this wayfinding design strategy and eye movement behavior allows the technical solution to invoke targeted wayfinding signage adjustment measures to improve the effectiveness of wayfinding communication after recognizing the user's cognitive state.

[0156] The reliability of the technical approach that maps three-dimensional behavioral trajectories to two-dimensional projection point coordinates through perspective projection and infers the user's attention direction based on the inter-frame motion direction of the projection point in the spatial layout environment of digital navigation interfaces is based on the assumption that there is a statistical coupling relationship between the target user's head orientation and the torso movement direction.

[0157] Specifically, during movement within the three-dimensional passageway, the angle deviation between the target user's facial orientation and the displacement direction of the skeletal centroid remains less than a preset dispersion angle in most consecutive time windows. Based on this statistical regularity, after mapping the spatial position data of the three-dimensional skeletal joints to the pixel coordinate system of the digital navigation interface, the direction of movement of the two-dimensional projection points in adjacent time windows can serve as an approximate description of the user's attention projection direction. During the initial deployment phase, the user behavior response method server establishes a distribution model of the angle deviation between the head orientation vector and the torso displacement vector by collecting sample users' head posture sensing data and synchronized behavioral trajectory data at preset observation points. When the central tendency index of the angle deviation distribution meets the preset confidence condition, the validity of the attention estimation mapping within the three-dimensional passageway is confirmed, and the navigation pre-wake logic based on the movement direction of the two-dimensional projection points is activated.

[0158] Furthermore, all preset parameters of the optical flow particle motion field, including the total number of particles emitted, particle motion step size, directional perturbation variance parameters, and specific parameters of the brightness modulation function, are calibrated through a multi-round iterative parameter optimization process during system initialization. In this iterative optimization process, the user behavior response method server generates multiple sets of optical flow particle motion field animation sample sequences with different parameter combinations. Testers evaluate the guidance pre-wake response effect corresponding to each parameter combination under the same pre-wake timing conditions, using response time, gaze duration, and guidance task completion accuracy as evaluation indicators. A parameter search strategy combining random search and Bayesian optimization is employed. After a specified number of iterations of evaluation, the parameter combination with the highest comprehensive score is selected as the parameter configuration for the officially running optical flow particle motion field. This calibration process ensures that the visual effect of the optical flow particle motion field can stably achieve the expected technical effect of guiding user attention in practical applications.

[0159] This invention generates a forward-looking behavior trace data stream from the three-dimensional behavioral trajectory data of the target user. Based on the layout of edge nodes in the user's field of vision within the digital wayfinding interface, this data stream is spatiotemporally mapped to generate a cognitive attention spatiotemporal evolution trend map representing the changing patterns of attention. This allows for the quantification and capture of the dynamic correlation between the behavioral trajectory and the edge areas of the wayfinding interface. Based on this, triggering attention abrupt change peaks are extracted from the cognitive attention spatiotemporal evolution trend map. The timestamps and edge node identifiers of these peaks serve as indicators of pre-awakening timing, while the slope of the attention change trend of these peaks indicates pre-awakening intensity. This enables accurate prediction of the timing and intensity of wayfinding information demand. Furthermore, by utilizing pre-wake-up timing and pre-wake-up intensity indicators, an optical flow particle motion field with directional gradient attributes is created in the digital wayfinding interface using the corresponding position of the edge node identifier as the anchor point. The wayfinding identifier template is then embedded into this field in a semantically enhanced manner to generate dynamic wayfinding response interface frames. Finally, the dynamic wayfinding response interface frame sequence is distributed to the corresponding edge nodes of the user's field of vision to trigger the wayfinding pre-wake-up display operation. Overall, the spatiotemporal mapping of user behavior trajectory, the analysis of cognitive attention abrupt change peaks, and the wayfinding expression of optical flow particle motion field are organically connected, realizing end-to-end closed-loop processing from user behavior perception to wayfinding pre-wake-up response, thereby improving the timing accuracy of wayfinding information push and the visual guidance effect.

[0160] In this way, it is possible to predict the continuous changes in user behavior patterns when pushing wayfinding information, and to improve the ability to match and adjust the visual intensity of wayfinding information with the dynamic changes in user attention. This improves the technical problem of wayfinding information being out of sync with users' actual information needs or disconnected from their current attention.

[0161] Please see Figure 2 The figure is a schematic diagram of the basic structure of a user behavior response server 200 provided in an embodiment of the present invention. The user behavior response server 200 includes: a processor 201; a storage device 202 on which a computer program 2020 is stored; and a network interface 203 for providing network communication functions. When the computer program 2020 is executed by the processor 201, the processor 201 implements any of the user behavior response methods applied to the signage and wayfinding system described above.

[0162] Please see Figure 3 The present invention provides a functional block diagram of a user behavior response device, which includes: The behavior data generation module is used to receive the behavior trajectory data of the target user in the three-dimensional passage space continuously collected by the non-contact sensing device, and generate a forward-looking behavior trace data stream based on the behavior trajectory data; The spatiotemporal mapping processing module is used to perform spatiotemporal mapping processing on the forward behavior trace data stream according to the preset layout of user field edge nodes in the digital navigation interface, and generate a spatiotemporal evolution trend map of cognitive attention corresponding to each user field edge node. The spatiotemporal evolution trend map of cognitive attention represents the change law of attention of the forward behavior trace data stream in each edge area of ​​the digital navigation interface. The wake-up indication determination module is used to parse the triggering attention change peak and its corresponding timestamp and edge node identifier from the spatiotemporal evolution trend map of cognitive attention, as a pre-wake-up timing indication for the guidance information demand, and to extract the slope of the attention change trend of the change peak as a pre-wake-up intensity indication for the brightness adjustment of the guidance interface. The wayfinding interface generation module is used to create an optical flow particle motion field with directional gradient attributes in the digital wayfinding interface by using the pre-wake-up timing indicator and the pre-wake-up intensity indicator as anchor points, and embed the preset wayfinding identifier template into the optical flow particle motion field in a semantic enhancement manner to generate dynamic wayfinding response interface frames arranged in time sequence. The wayfinding wake-up display module is used to distribute the dynamic wayfinding response interface frame sequence to the corresponding user field edge node according to the edge node identifier, and trigger the user field edge node to perform wayfinding pre-wake-up display operation.

[0163] Based on the above, a readable storage medium is provided, on which a program or instructions are stored, and when the program or instructions are executed by a processor, the steps of the above method are implemented.

[0164] Furthermore, it should be noted that this embodiment of the invention also provides a computer program product, which may include a computer program that can be stored in a computer-readable storage medium. The processor of the user behavior response server reads the computer program from the computer-readable storage medium, and the processor can execute the computer program, causing the user behavior response server to perform the aforementioned... Figure 1 The methods described in the corresponding embodiments are already known, and therefore will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the computer program product embodiments related to this invention, please refer to the description of the method embodiments of this invention.

[0165] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

Claims

1. A user behavior response method applied to a signage and wayfinding system, characterized in that, include: Receives the target user's behavior trajectory data continuously collected by a non-contact sensing device within a three-dimensional passage space, and generates a forward-looking behavior trace data stream based on the behavior trajectory data; Based on the preset layout of user field edge nodes in the digital navigation interface, the forward-looking behavior trace data stream is subjected to spatiotemporal mapping processing to generate a spatiotemporal evolution trend map of cognitive attention corresponding to each user field edge node. The spatiotemporal evolution trend map of cognitive attention represents the change law of attention of the forward-looking behavior trace data stream in each edge area of ​​the digital navigation interface. The triggering attention mutation peaks and their corresponding timestamps and edge node identifiers are extracted from the spatiotemporal evolution trend map of cognitive attention, serving as pre-wake-up timing indicators for guidance information needs, and the slope of the attention change trend of the mutation peaks is extracted as a pre-wake-up intensity indicator for adjusting the brightness of the guidance interface. Using the pre-wake-up timing indicator and the pre-wake-up intensity indicator, an optical flow particle motion field with directional gradient properties is created in the digital wayfinding interface with the position corresponding to the edge node identifier as the anchor point. The preset wayfinding identifier template is embedded into the optical flow particle motion field in a semantic enhancement manner to generate dynamic wayfinding response interface frames arranged in time sequence. The dynamic wayfinding response interface frame sequence is distributed to the corresponding user field edge nodes according to the edge node identifier, triggering the user field edge nodes to perform wayfinding pre-wake display operations.

2. The method according to claim 1, characterized in that, The step involves performing spatiotemporal mapping processing on the forward-looking behavior trace data stream based on the preset layout of user field-of-view edge nodes in the digital navigation interface, generating a spatiotemporal evolution trend map of cognitive attention corresponding to each user field-of-view edge node, including: Extract the node identifier of each user field of view edge node in the user field of view edge node layout and the boundary coordinate sequence of the edge node region of the node in the digital navigation interface coordinate system, and obtain the display characteristic information of each user field of view edge node. Behavioral trajectory sampling points are extracted sequentially from the forward-looking behavioral trace data stream in ascending order of timestamps. The three-dimensional spatial position coordinates of each behavioral trajectory sampling point and its acquisition timestamp are obtained, and the three-dimensional spatial position coordinates are converted into a three-dimensional trajectory point sequence under the three-dimensional travel space reference coordinate system. Based on the spatial pose parameters of the imaging plane of the digital navigation interface, a perspective projection mapping relationship is constructed from the three-dimensional access space reference coordinate system to the pixel coordinate system of the digital navigation interface. Through the perspective projection mapping relationship, each behavior trajectory sampling point in the three-dimensional trajectory point sequence is mapped to the coordinates of a two-dimensional projection point on the digital navigation interface, resulting in a sequence of two-dimensional projection point coordinates with corresponding timestamps. Inter-frame motion pointing analysis is performed on the coordinates of two-dimensional projection points with adjacent timestamps in the two-dimensional projection point coordinate sequence. The direction of movement of the behavior trajectory projection within the corresponding time interval is determined based on the relative changes in the positions of adjacent two-dimensional projection point coordinates on the digital navigation interface. The movement offset of the behavior trajectory projection within the corresponding time interval is obtained based on the spatial interval distance between adjacent two-dimensional projection point coordinates. Based on the boundary coordinate sequence of the edge node region and the movement direction of the behavior trajectory projection, the projection intersection relationship of the current two-dimensional projection point coordinates extending along the movement direction to the edge of the digital guidance interface is determined. The user's field of vision edge node that the behavior trajectory projection point is facing at the current moment is then determined, and an association mapping is established between the current behavior trajectory projection point and the node identifier of the user's field of vision edge node. A sliding time window is set for each user's visual edge node, and behavioral trajectory projection points with associated mappings are collected based on the sliding time window. The spatiotemporal evolution trend map of cognitive attention is generated through the behavioral trajectory projection points.

3. The method according to claim 2, characterized in that, The step of setting a sliding time window for each user's visual field edge node, collecting behavioral trajectory projection points with associated mappings based on the sliding time window, and generating the spatiotemporal evolution trend map of cognitive attention through the behavioral trajectory projection points includes: For each user's visual field edge node, a sliding time window with the current time as the end boundary is set. All behavioral trajectory projection points that establish an association mapping with the user's visual field edge node within the sliding time window are collected, and a set of associated projection points of the user's visual field edge node within the corresponding time window is generated. Based on the spatial distribution dispersion and the consistency of the movement direction of the behavioral trajectory projection points in the associated projection point set, the cognitive attention tendency state information of the user's visual field edge node in the current time window is generated. The cognitive attention tendency state information includes an aggregation tendency description reflecting the spatial distribution concentration trend and a directional convergence description reflecting the degree of convergence of directions. The sliding time window is slid along the time axis with a set time step, and the cognitive attention tendency state information of each time window is generated iteratively for each user's visual field edge node, so as to obtain the sequence of changes in the cognitive attention tendency state of each user's visual field edge node as the sliding time window moves. The sequence of cognitive attention tendency changes of all user visual edge nodes is synchronized and aligned according to a common time axis to generate a multi-dimensional state evolution description with the time axis as the first dimension and the node identifier as the second dimension, and the multi-dimensional state evolution description is used as the spatiotemporal evolution trend diagram of cognitive attention. A continuity analysis of state propagation between nodes is performed on the sequence of cognitive attention tendency state changes corresponding to spatially adjacent user field edge nodes in the cognitive attention spatiotemporal evolution trend map. State transition enhancement is performed on adjacent nodes with synchronous state change trends to obtain a cognitive attention spatiotemporal evolution trend map for analyzing triggering attention mutation peaks.

4. The method according to claim 1, characterized in that, The step of parsing the triggering attention abrupt change peak and its corresponding timestamp and edge node identifier from the spatiotemporal evolution trend map of cognitive attention, as an indication of the pre-wake-up timing for wayfinding information needs, and extracting the slope of the attention change trend of the abrupt change peak as an indication of the pre-wake-up intensity for adjusting the brightness of the wayfinding interface, includes: Temporal difference analysis is performed on the cognitive attention tendency state change sequence of each user's field of vision edge node in the spatiotemporal evolution trend map of cognitive attention, and the state change step size of the cognitive attention tendency state change sequence between adjacent time windows is calculated. The time position when the state change step size enters the rapid growth stage from the steady change stage is marked as the candidate mutation time position. Extract the sequence fragments of cognitive attention tendency state change within adjacent time windows before and after the candidate mutation time position as background state fragments and mutation state fragments. Compare the deviation degree of the state change step size distribution pattern of the mutation state fragments with the state change step size distribution pattern of the background state fragments. Select the confirmed mutation peak positions from the candidate mutation time positions whose deviation degree meets the preset cognitive baseline mutation conditions. The timestamp of the time window corresponding to the confirmed mutation peak position on the common time axis is obtained as the trigger time, and the node identifier of the user's field of vision edge node to which the confirmed mutation peak position belongs is extracted as the node to be woken up. The trigger time and the node to be woken up are established as a pairwise mapping relationship. The pairwise mapping relationship between the trigger time and the node to be woken up corresponding to each confirmed mutation peak position is combined into a pre-wake-up timing indication for the guidance information requirement. The pre-wake-up timing indication includes the correspondence between the edge node that needs to perform the guidance pre-wake-up operation and the corresponding trigger time point. Based on the mutation state fragment corresponding to the mutation peak position in the cognitive attention tendency state change sequence, the slope of the attention change trend at the confirmed mutation peak position is generated, and the slope of the attention change trend is converted into the pre-awakening intensity indicator.

5. The method according to claim 4, characterized in that, The step of identifying the mutation state segment corresponding to the mutation peak position in the cognitive attention tendency state change sequence, generating the attention change trend slope at the identified mutation peak position, and converting the attention change trend slope into the pre-arousal intensity indicator includes: From the sequence of changes in cognitive attention tendency state, extract the mutation state segment corresponding to the confirmed mutation peak position, extract the aggregation tendency description change amount and the direction convergence description change amount experienced by the cognitive attention tendency state information in the mutation state segment from the mutation initiation state to the mutation peak state, and synthesize the aggregation tendency description change amount and the direction convergence description change amount into the cumulative state rise amplitude. Obtain the timestamps corresponding to the start time window and the end time window of the mutation state segment, and calculate the time span between the start time window timestamp and the end time window timestamp as the rising duration span; The slope of the attention change trend at the confirmed mutation peak position is generated based on the cumulative magnitude of the state increase and the span of the increase duration. The slope of the attention change trend represents the rate at which the cognitive attention tendency state jumps from the background level to the peak level. The slope of the attention change trend is converted into brightness adjustment trend direction description information that matches the brightness adjustment range of the digital wayfinding interface, and used as a pre-wake intensity indicator for the brightness adjustment of the wayfinding interface.

6. The method according to claim 1, characterized in that, The method involves using the pre-wake-up timing indicator and the pre-wake-up intensity indicator to create an optical flow particle motion field with directional gradient attributes in the digital wayfinding interface, using the position corresponding to the edge node identifier as the anchor point. A preset wayfinding identifier template is then embedded into the optical flow particle motion field in a semantically enhanced manner to generate dynamic wayfinding response interface frames arranged in chronological order. Extract the identifier of the node to be woken up and the corresponding trigger time contained in the pre-wake-up timing indication. Based on the identifier of the node to be woken up, obtain the anchor point coordinates of the node to be woken up in the digital navigation interface from the user's field of view edge node layout. Use the trigger time as the rendering start time point of the dynamic navigation response interface frame sequence. From the forward-looking behavior trace data stream, extract behavior trajectory segments within a preset retrospective observation interval before the rendering start time point, extract the two-dimensional projection point coordinates and movement direction of each behavior trajectory sampling point on the digital guide interface, and generate a historical behavior trajectory direction reference sequence to guide the movement direction of optical flow particles. The movement direction in the historical behavior trajectory direction reference sequence is converted into the initial emission direction vector of the optical flow particle in the digital navigation interface, and the deflection evolution description of the emission direction of the optical flow particle over time is constructed based on the direction change angle of the movement direction within the preset retrospective observation interval. Using the anchor point coordinates as the particle emission origin, and based on the initial emission direction vector and the deflection evolution description, each optical flow particle is set with a motion direction trajectory, so that the optical flow particle moves in the digital navigation interface along the motion direction trajectory after starting from the anchor point coordinates. The brightness modulation mode of the optical flow particles during their movement is determined based on the brightness adjustment trend direction description information in the pre-wake intensity indicator; Based on the particle emission origin, the motion trajectory, and the brightness modulation method, a parameterized motion description of optical flow particles is created in the coordinate system of the digital navigation interface. The parameterized motion description of optical flow particles defines the instantaneous position coordinates and instantaneous brightness value of each optical flow particle at any moment of motion. The parameterized motion description of the optical flow particles is sampled at a preset particle motion frame rate to generate a particle state sequence consisting of the particle position and particle brightness of each optical flow particle at each discrete motion moment. The particle state sequences of all optical flow particles are superimposed and synthesized to obtain the optical flow particle motion field animation frame sequence. The dynamic wayfinding response interface frame is obtained by selecting the wayfinding sign template associated with each node to be woken up from the preset wayfinding sign template library and embedding it with semantic enhancement.

7. The method according to claim 6, characterized in that, The step of selecting the wayfinding sign template associated with each node to be woken up from the preset wayfinding sign template library and embedding it with semantic enhancement to obtain the dynamic wayfinding response interface frame includes: Access the preset wayfinding sign template library, select the wayfinding sign template associated with the node to be woken up according to the node identifier to be woken up, and separate the wayfinding graphic elements and wayfinding text elements in the wayfinding sign template into independent renderable layers; Edge detection processing is performed on the guide graphic element to extract the contour boundary path. A halo region is generated by expanding outward along the contour boundary path. The color saturation and transparency parameters of the halo region are determined based on the pre-wake intensity indication. The halo region is superimposed on the guide graphic element to form a contour halo enhanced guide graphic element. Hue shift processing is performed on the guide text element to obtain a color enhanced guide text element. The contour halo enhanced guide graphic element and the color enhanced guide text element are embedded frame by frame into the screen area corresponding to the anchor point position coordinates in the optical flow particle motion field animation frame sequence. Through frame-by-frame synthesis processing, a dynamic guide response interface frame is generated in which the optical flow particles at the anchor point position dynamically flow along the historical behavior trajectory direction and the guide sign has contour halo and color enhanced visual effects.

8. The method according to claim 1, 2, 4, or 6, characterized in that, The step of distributing the dynamic wayfinding response interface frame sequence to the corresponding user field-of-view edge nodes according to the edge node identifier, and triggering the user field-of-view edge nodes to perform wayfinding pre-wake-up display operations, includes: Read the edge node identifier associated with each wayfinding response interface frame from the dynamic wayfinding response interface frame sequence, and use the edge node identifier as the grouping basis to split the dynamic wayfinding response interface frame sequence into an independent frame sequence that corresponds one-to-one with each edge node identifier. Obtain the origin coordinates, width, and height of the display area for each edge node; establish a rectangular boundary description of the display area for each edge node; and obtain the display color configuration file for each edge node. For each edge node, the global canvas coordinate system of each frame in the independent frame sequence assigned to that edge node is transformed with the rectangular boundary description of the node display area of ​​that edge node to determine the coordinate range of the corresponding sub-region of the rectangular boundary of the node display area in each frame image. Perform pixel-level cropping operation on each frame image according to the coordinate range of the corresponding sub-region, and extract sub-image blocks that completely match the rectangular boundary of the node display area. Combine all the extracted sub-image blocks in frame order to generate a local guide response interface sub-sequence frame that matches the size of the edge node display area. The continuity of the optical flow particle motion trajectory between adjacent sub-image blocks in the local guide response interface sub-sequence frame is detected to restore the continuity of the motion direction of the optical flow particles when they cross the clipping boundary. Using the display color profile corresponding to the edge node, each frame of the repaired local wayfinding response interface sub-sequence frame is processed for color space mapping and gamma correction, and encoded and packaged according to the data transmission protocol supported by the edge node to generate node push data packets. Each edge node transmits its push data packet to the corresponding edge node. After receiving the corresponding push data packet, each edge node parses the frame sequence timestamp alignment information and inter-frame decoding dependency description in the push data packet. Based on the inter-frame decoding dependency description, it sequentially decodes each frame of the local guide response interface sub-sequence frame and presents each frame in the display area of ​​the edge node according to the timestamp order.

9. A user behavior response server, characterized in that, include: A processor; a storage device having a computer program stored thereon; a network interface for providing network communication functions; and when the computer program is executed by the processor, causing the processor to implement the user behavior response method for a signage and wayfinding system as described in any one of claims 1-8.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the user behavior response method for a signage and wayfinding system as described in any one of claims 1-8.

Images