Method, device and storage medium for optimizing commercial space layout based on neural data

Abstract

The application discloses a commercial space layout optimization method and device based on neural data and a storage medium, relates to the technical field of electric digital data processing, and comprises the following steps: acquiring exhibition space layout data; after the electroencephalogram data, eye movement data and space trajectory data of each collection target are aligned according to timestamps, the space position information in the exhibition space layout data is subjected to space registration, and interest correlation data of each collection target at each space position unit is obtained; according to the interest correlation data of each collection target at each space position unit, a group interest measurement value at each space position unit is determined; and according to the group interest measurement value at each space position unit and the exhibition space layout data, a layout optimization scheme of an exhibition physical space is generated. The application improves the flow of the optimized exhibition.

Description

Technical Field

[0001] This application relates to the field of electrical digital data processing technology, and in particular to a method, device and storage medium for optimizing commercial spatial layout based on neural data. Background Technology

[0002] Offline booths, brand pop-up stores, and large-scale exhibitions are key physical venues for brands to showcase their products and drive commercial conversion. The spatial arrangement of exhibits within a booth, the relative positions of various display areas, and the physical flow of visitor movement directly determine the movement path and visual sampling range of visitors within a limited space.

[0003] Currently, the main evaluation methods for booth space layout in related technologies are based on video surveillance or infrared pyroelectric sensors to count visitor flow, and then select booth numbers with high visitor flow based on the visitor flow. Specifically, video cameras or infrared pyroelectric sensors are deployed at the entrances and passages of each area of ​​the booth. Moving targets are identified and counted by inter-frame difference method or background subtraction method. The number of people entering each booth area per unit time and the average number of people staying in each area are counted. The visitor flow value is then used as an evaluation index of booth attractiveness to serve subsequent booth location decisions.

[0004] However, the relevant technologies assess the number of people and the rough density distribution in each area. The actual physical interactions of visitors within the booth, including changes in three-dimensional spatial coordinates and accompanying changes in gaze direction, cannot be reflected in the pedestrian flow data. Therefore, relying solely on pedestrian flow data limits optimization to a single spatial strategy: placing exhibits next to the physical aisles with the highest pedestrian traffic. It fails to distinguish between two drastically different scenarios: visitors passing by the booth without looking at the exhibit and visitors actively approaching and looking at the exhibit. This results in site selection decisions based on pedestrian counts essentially being equivalent to choosing the location with the highest physical aisle traffic, leading to insufficient visitor traffic at the booth. Summary of the Invention

[0005] The main purpose of this application is to provide a method, device and storage medium for optimizing the layout of commercial spaces based on neural data, which aims to solve the technical problem that the optimization scheme for booth space layout is based on the coordinate position with the largest physical channel traffic, resulting in poor optimization effect.

[0006] To achieve the above objectives, this application provides a method for optimizing commercial spatial layout based on neural data, the method comprising: Obtain booth space layout data, which includes the coordinate range of the physical space of the booth and the spatial location information of each exhibit within the coordinate range; Acquire multimodal interest data of multiple acquisition targets within the coordinate range, wherein the multimodal interest data includes EEG data, eye movement data, and spatial trajectory data of the acquisition targets; After aligning the EEG data, eye movement data, and spatial trajectory data of each of the aforementioned acquisition targets according to timestamps, spatial registration is performed with the spatial location information in the booth spatial layout data to obtain the interest association data of each of the aforementioned acquisition targets at each spatial location unit; Based on the interest association data of each of the collected targets at each spatial location unit, determine the group interest metric value at each spatial location unit; Based on the group interest metric value at each spatial location unit and the booth space layout data, a layout optimization scheme for the physical space of the booth is generated.

[0007] In one embodiment, obtaining the booth space layout data includes: A three-dimensional spatial coordinate system covering the physical space of the booth is established with the designated location of the booth as the origin, and the coordinate range of the physical space of the booth is determined. In the three-dimensional spatial coordinate system, the coordinates of the outer contour boundary points of each exhibit within the physical space of the booth are measured by a ranging device. The spatial area enclosed by the coordinates of the outer contour boundary points is determined as the spatial location information of the corresponding exhibit, and an exhibit identifier is assigned to each exhibit. The exhibit identifier and corresponding spatial location information of each exhibit are associated and stored with the coordinate range of the physical space of the booth to generate the booth space layout data.

[0008] In one embodiment, acquiring multimodal interest data of multiple acquisition targets within the coordinate range includes: Acquire initial EEG data and initial eye movement data of the target within the coordinate range; The initial spatial trajectory data of the target is collected by a spatial positioning tag, which is configured to receive signals from positioning base stations deployed in the physical space of the booth. The initial EEG data, initial eye movement data, and initial spatial trajectory data of each of the acquisition targets are preprocessed to remove invalid data segments and noise data in each data stream, so as to obtain the EEG data, eye movement data, and spatial trajectory data corresponding to each acquisition target as the multimodal interest data.

[0009] In one embodiment, the preprocessing of the initial EEG data, initial eye-tracking data, and initial spatial trajectory data of each of the acquisition targets to remove invalid data segments and noise data from each data stream, and obtaining the EEG data, eye-tracking data, and spatial trajectory data corresponding to each acquisition target as the multimodal interest data, includes: The initial EEG data is bandpass filtered to retain the EEG signal components within the target frequency band. The EEG data is then obtained by identifying and removing the electrooculogram (EOG) artifacts and electromyogram (EMG) artifacts from the EEG data using an independent component analysis algorithm. The initial eye-tracking data is subjected to invalid frame removal processing. The invalid frames include off-screen data frames and frame data with acquisition confidence lower than a preset confidence threshold. The initial eye-tracking data is then corrected in spatial coordinate system based on the factory calibration parameters of the eye-tracking acquisition device, and the effective fixation point sequence is extracted to obtain the eye-tracking data. The initial spatial trajectory data is subjected to outlier removal processing, and the initial spatial trajectory data after outlier removal is smoothed and filtered to obtain the spatial trajectory data. The EEG data, eye-tracking data, and spatial trajectory data are aligned and bound according to a unified timestamp to obtain the multimodal interest data of each acquisition target.

[0010] In one embodiment, the step of aligning the EEG data, eye-tracking data, and spatial trajectory data of each of the acquisition targets by timestamp, and then spatially registering them with the spatial location information in the booth spatial layout data to obtain interest association data of each acquisition target at each spatial location unit includes: The gaze coordinates in the eye movement data of each of the aforementioned targets are mapped from the coordinate system of the eye movement acquisition device to the three-dimensional spatial coordinate system to obtain the three-dimensional gaze space coordinates of the gaze point at each moment in the physical space of the booth. Spatial matching is performed between the three-dimensional gaze space coordinates at each moment and the spatial location information of each exhibit in the booth space layout data to determine the target exhibit identifier corresponding to the gaze point at each moment; Based on the spatial trajectory data of the target at each moment and the corresponding three-dimensional gaze space coordinates, calculate the gaze direction vector of the target at each moment, and determine whether the angle between the gaze direction vector and the spatial vector from the target to the target exhibit is less than a preset angle threshold. If it is less, then the EEG data, eye movement data and the corresponding target exhibit identifier at that moment are associated and recorded as valid interest association data. The coordinate range of the physical space of the booth is divided into multiple spatial location units, either by the spatial location information corresponding to each exhibit or according to a preset grid size. Remove the associated data corresponding to the moments when the included angle is greater than or equal to the preset angle threshold, and summarize the valid interest-related data of each moment in chronological order according to the dimension of the acquisition target to obtain the interest-related data of each acquisition target at each spatial location unit.

[0011] In one embodiment, determining the group interest metric at each spatial location unit based on the interest association data of each of the collected targets at each spatial location unit includes: For each spatial location unit, acquire all interest-related data of all acquisition targets within that spatial location unit; Population statistics are performed on the EEG data of each of the collected targets in the interest-related data within the spatial location unit, and the population mean of β wave power and α wave power within the spatial location unit are calculated. Group statistics are performed on the eye movement data of each acquisition target in the interest association data within the spatial location unit to calculate the cumulative number of fixation points, the cumulative value of fixation duration, and the number of acquisition targets that generate fixation within the spatial location unit; The group interest metric of the spatial location unit is calculated based on the group mean of β-wave power, group mean of α-wave power, cumulative number of fixation points, cumulative fixation duration, and number of data acquisition targets that generate fixation.

[0012] In one embodiment, generating a layout optimization scheme for the physical space of the booth based on the group interest metric value at each spatial location unit and the booth space layout data includes: The group interest metric values ​​at the spatial location units corresponding to each exhibit are sorted in descending order to generate exhibit interest ranking results. Based on the interest ranking results of the exhibits, the recommended placement priority of each exhibit is determined. The exhibits with the top N group interest metric values ​​are marked as high-priority exhibits, and the high-priority exhibits are adjusted within the coordinate range of the physical space of the booth to a region where the distance between the booth entrance and the booth entrance is less than a preset distance threshold. Obtain the preset functional association attributes between each exhibit, mark exhibits with the same functional association attributes as an associated exhibit group, and adjust each exhibit in the associated exhibit group to an adjacent area where the spatial distance between them is less than a preset spacing threshold. The booth space layout data is updated based on the adjusted spatial location information of each exhibit to generate a layout optimization scheme that includes the spatial location information of each exhibit's objectives.

[0013] In one embodiment, generating a layout optimization scheme for the physical space of the booth based on the group interest metric value at each spatial location unit and the booth space layout data further includes: Obtain user profile tags for each of the aforementioned collection targets, wherein the user profile tags include at least one of gender tags, age group tags, and interest preference tags; Based on the user profile tags of each of the collected targets, all collected targets are divided into multiple subgroups, and the collected targets in the same subgroup have the same user profile tags or the same combination of user profile tags. For each subgroup, the subgroup interest metric value at each spatial location unit is calculated from the interest association data of each target in the subgroup at each spatial location unit. The calculation method of the subgroup interest metric value is the same as the calculation method of the group interest metric value. Based on the dwell time of the spatial trajectory data in each spatial location unit, spatial location units with a dwell time exceeding a preset dwelling threshold are marked as dwelling points; The spatial trajectory data of each acquisition target is extracted in time sequence, and the spatial location unit sequence traversed by each acquisition target is clustered by sequence similarity. The sequence cluster containing the most acquisition targets is determined as the high-frequency path. Based on the subgroup interest metric values ​​of each subgroup at each spatial location unit, as well as the stopping point and the high-frequency path, a movement guidance scheme matching each subgroup is generated. The movement guidance scheme includes: a sequence of recommended path exhibits arranged in spatial coordinate order, and spatial location information of the placement of guide signs between adjacent recommended path exhibits.

[0014] In addition, to achieve the above objectives, this application also provides a business space layout optimization device based on neural data. The business space layout optimization device based on neural data includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the business space layout optimization method based on neural data as described above.

[0015] In addition, to achieve the above objectives, this application also provides a storage medium, which is a computer-readable storage medium, on which a program implementing a business space layout optimization method based on neural data is stored. The program implementing the business space layout optimization method based on neural data is executed by a processor to implement the steps of the business space layout optimization method based on neural data as described above.

[0016] This application provides a method for optimizing commercial space layout based on neural data. First, it acquires booth space layout data and multimodal interest data of multiple target groups within the booth's physical space. This multimodal interest data includes EEG data, eye-tracking data, and spatial trajectory data. This comprehensively collects real interactive behavior data of visitors within the booth from three objective dimensions: visual attention, cognitive engagement, and spatial movement. This overcomes the inherent limitation of traditional methods that rely solely on single-dimensional pedestrian count, which cannot distinguish between those who pass by without looking and those who actively approach and look. Next, by aligning the EEG data, eye-tracking data, and spatial trajectory data of each target group according to timestamps and spatially registering them with the spatial location information in the booth space layout data, it obtains interest association data for each target group at each spatial location unit. This ensures that the physiological response signals and visual attention data at each moment are accurately anchored to the corresponding physical space location. This approach addresses the technical challenge of accurately locating high-interest areas due to the disconnect between sensor data and spatial location in traditional solutions. Based on spatial registration, it determines the group interest metric for each spatial location unit using interest-related data of each collected target. This metric comprehensively reflects the cognitive engagement and visual attention intensity of multiple visitors at each spatial location unit. This shifts the evaluation metric for subsequent layout optimization from the traditional method of counting how many people passed by a location to assessing the actual level of visitor interest response at that location. Finally, based on the group interest metric for each spatial location unit and the booth layout data, it generates an optimized layout plan for the physical space of the booth. This optimization plan no longer selects locations based on the coordinates of the highest physical traffic flow, but rather guides the arrangement of exhibits and the booth layout based on the spatial distribution of visitors' actual interest responses.

[0017] In summary, this application, by introducing a multimodal synchronous acquisition and spatial registration mechanism of EEG data, eye-tracking data, and spatial trajectory data, generates an optimized booth space layout scheme by comprehensively analyzing multi-dimensional objective physiological response data. This overcomes the technical defect of existing technologies that rely solely on visitor counts for booth location selection, where "location decision-making is essentially equivalent to selecting the coordinate position with the highest channel traffic." This allows the optimized booth layout scheme to truly reflect the distribution of visitors' real interests in exhibits and different areas of the booth, thereby improving the effectiveness of booth space layout optimization decisions. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an embodiment of the neural data-based commercial space layout optimization method of this application. Figure 2 This is a flowchart illustrating Embodiment 5 of the commercial space layout optimization method based on neural data in this application. Figure 3 This is a schematic diagram of the hardware operating environment involved in the neural data-based commercial space layout optimization device of this application.

[0021] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of this application and are not intended to limit this application.

[0023] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0024] Currently, relevant technologies assess the number of people and their approximate density distribution in each area. However, the actual physical interactions of visitors within the booth, including changes in three-dimensional spatial coordinates and accompanying changes in gaze direction, cannot be reflected by pedestrian flow data. Therefore, relying solely on pedestrian flow data limits optimization strategies to simply placing exhibits alongside the physical aisles with the highest pedestrian traffic. It fails to distinguish between two drastically different scenarios: visitors passing by the booth without looking at the exhibit and visitors actively approaching and looking at the exhibit. This results in site selection decisions based on pedestrian flow counts essentially being equivalent to choosing the location with the highest physical aisle traffic, leading to insufficient visitor traffic at the booth.

[0025] The main solution of this application is as follows: Acquire booth space layout data, which includes the coordinate range of the booth's physical space and the spatial location information of each exhibit within that coordinate range; acquire multimodal interest data of multiple acquisition targets within the coordinate range, including the acquisition targets' EEG data, eye-tracking data, and spatial trajectory data; align the EEG data, eye-tracking data, and spatial trajectory data of each acquisition target by timestamp and spatially register them with the spatial location information in the booth space layout data to obtain interest association data of each acquisition target at each spatial location unit; determine the group interest metric value at each spatial location unit based on the interest association data of each acquisition target at each spatial location unit; and generate a layout optimization scheme for the booth's physical space based on the group interest metric value at each spatial location unit and the booth space layout data.

[0026] This application introduces a multimodal synchronous acquisition and spatial registration mechanism for EEG data, eye-tracking data, and spatial trajectory data to generate an optimized booth space layout scheme by comprehensively analyzing multi-dimensional objective physiological response data. This overcomes the technical defect of existing technologies that rely solely on visitor counts for booth location selection, which results in "the location decision being essentially equivalent to selecting the coordinate position with the highest channel traffic." This allows the optimized booth layout scheme to truly reflect the distribution of visitors' real interests in the exhibits and different areas of the booth, thereby improving the effectiveness of the optimized booth space layout decision.

[0027] It should be noted that the executing entity in this embodiment can be a commercial space layout optimization system based on neural data, or a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or a commercial space layout optimization device based on neural data capable of performing the above functions. This embodiment does not specifically limit it in this way. The following uses a commercial space layout optimization device based on neural data as the executing entity as an example to describe this embodiment and the following embodiments.

[0028] Based on this, Embodiment 1 of this application proposes a commercial space layout optimization method based on neural data. Please refer to... Figure 1 The business spatial layout optimization method based on neural data includes steps S10 to S50: Step S10: Obtain booth space layout data, which includes the coordinate range of the booth's physical space and the spatial location information of each exhibit within the coordinate range.

[0029] In this embodiment, the booth space layout data is a structured data set characterizing the geometric attributes of the booth's physical space and the distribution attributes of the exhibits. The coordinate range of the booth's physical space is a continuous spatial region enclosed by boundary coordinate points in a three-dimensional coordinate system, which defines the usable physical boundaries of the booth. The spatial location information corresponding to each exhibit is a set of three-dimensional coordinates of the spatial region occupied by each exhibit within the booth's physical space, including the coordinate sequence of the boundary points of each exhibit's outer contour.

[0030] As an optional implementation, the control terminal deployed at the exhibition booth first acquires a floor plan of the booth's physical space, which includes two-dimensional coordinate information of the booth boundaries and internal exhibits. The control terminal uses the projection of the booth entrance center point onto the horizontal plane as the origin, and the extension lines along the booth's length as the X-axis, the extension lines along the booth's width as the Y-axis, and the vertical direction as the Z-axis, to establish a three-dimensional spatial coordinate system covering the entire physical space of the booth. The control terminal extracts the inflection point coordinates of the booth boundary lines from the floor plan, converts them into boundary point coordinates in the three-dimensional spatial coordinate system, and the spatial range enclosed by all boundary point coordinates is determined as the coordinate range of the booth's physical space. The control terminal continues to extract the coordinates of the four corner points of the rectangular outline of each exhibit or the coordinates of all vertices of the polygonal outline of each exhibit from the floor plan, converts these outline point coordinates into a coordinate sequence in the three-dimensional spatial coordinate system, and determines the spatial location information of the corresponding exhibit by the spatial area enclosed by each outline point coordinate sequence. The control terminal assigns an exhibit identifier to each exhibit and associates and stores the exhibit identifier and its associated spatial location information with the coordinate range of the physical space of the booth, thereby generating booth space layout data.

[0031] As an alternative implementation, when no floor plan is provided at the exhibition booth, or the provided floor plan does not include the precise coordinates of each exhibit's outline, a control terminal deployed at the exhibition booth acquires the geometric dimensions of the physical space of the booth using a laser rangefinder and sequentially measures the spatial position of each exhibit. The control terminal sets up the laser rangefinder at the exhibition booth, establishes a three-dimensional coordinate system with a designated location in the physical space of the booth as the origin, and sequentially measures the distance and angle of each inflection point of the booth boundary relative to the origin. Triangulation is then performed to obtain the three-dimensional spatial coordinates of each inflection point, and the spatial range enclosed by all inflection point coordinates is determined as the coordinate range of the physical space of the booth. The control terminal sequentially measures the distance and angle of multiple boundary points on the outer contour of each exhibit relative to the origin, and triangulation is performed to obtain the three-dimensional spatial coordinates of each boundary point. The spatial area enclosed by the sequence of coordinates of all boundary points for each exhibit is determined as the spatial position information of that exhibit. The control terminal assigns an exhibit identifier to each exhibit, associates and stores the exhibit identifier and its spatial position information with the coordinate range of the physical space of the booth, and generates booth space layout data.

[0032] Step S20: Acquire multimodal interest data of multiple acquisition targets within the coordinate range. The multimodal interest data includes EEG data, eye movement data, and spatial trajectory data of the acquisition targets.

[0033] In this embodiment, the target of the data collection is a subject wearing a sensing device who enters the physical space of the exhibition booth for free viewing. Multimodal interest data is a dataset simultaneously containing three different types of sensor data: electroencephalographic response, visual attention distribution, and spatial movement trajectory. Electroencephalogram (EEG) data is a time-series signal of scalp potential changes over time, collected via electrodes attached to the scalp of the target; this signal is converted from analog to digital to form a multi-channel discrete-time sequence. Eye-tracking data is a time-series data sequence containing fixation point coordinates, pupil diameter, and fixation duration, collected via an eye-tracking device worn by the target. Spatial trajectory data is a time-series sequence of the target's three-dimensional coordinates within the physical space of the exhibition booth, recorded in real-time by a spatial positioning system.

[0034] As an optional implementation, before the target enters the physical space of the booth, on-site operators equip the target with a portable EEG acquisition device, a head-mounted eye-tracking acquisition device, and a spatial positioning tag. The electrodes of the portable EEG acquisition device are attached to the prefrontal and temporal lobes of the target's scalp according to the international 10-20 system standard, with a sampling rate set at no less than 250 Hz to capture raw EEG signals including alpha, beta, and theta wave frequencies. The head-mounted eye-tracking acquisition device includes two miniature infrared cameras facing the target's eyes, with a sampling rate set at no less than 60 Hz to capture the coordinates of the corneal reflection points and pupil center coordinates of both eyes, and calculates the gaze direction and fixation point coordinates using the pupil-corneal reflection vector method. The spatial positioning tag is an ultra-wideband transceiver. Positioning base stations deployed at the four corners of the booth's physical space transmit pulse signals to the tag at a frequency of no less than 10 Hz. After receiving signals from each base station, the tag calculates the distance between itself and each base station, and uses a time-of-arrival algorithm to calculate the tag's real-time three-dimensional coordinates in a three-dimensional spatial coordinate system as the target's spatial trajectory data. The control terminal simultaneously receives EEG data from a portable EEG acquisition device, eye-tracking data from a head-mounted eye-tracking acquisition device, and spatial trajectory data from a spatial positioning tag via a wireless communication link. It adds a uniformly formatted timestamp to each data point based on a global clock signal. While the target is freely viewing the exhibit within the physical space, the control terminal continuously records these three types of data until the target leaves the physical space. Initial EEG data, initial eye-tracking data, and initial spatial trajectory data are then preprocessed. The preprocessing of the EEG data involves the following steps: first, a bandpass filter from 0.5 Hz to 50 Hz is used to filter out DC drift and high-frequency circuit noise; then, a notch filter is used to filter out 50 Hz power frequency interference; finally, an independent component analysis algorithm is used to identify and remove EEG and EMG artifacts from the multi-channel signal, resulting in clean EEG data. The preprocessing of eye-tracking data involves the following steps: First, the pupil center coordinates and corneal reflector coordinates in each frame of eye-tracking data are detected. Frames whose corneal reflector coordinates exceed the field of view boundary of the eye-tracking acquisition device are marked as off-screen data frames and discarded. Frames whose pupil center coordinate calculation confidence is lower than a preset confidence threshold are marked as invalid data frames and discarded. Based on the intrinsic parameter matrix calibrated at the factory of the eye-tracking acquisition device, the pupil-corneal reflector vectors of the remaining frames are converted into gaze direction vectors and fixation point coordinates, resulting in clean eye-tracking data. The preprocessing of spatial trajectory data involves the following steps: First, the Euclidean distance between the 3D coordinates of each frame in the spatial trajectory data sequence and the 3D coordinates of the previous frame is calculated. Frames whose Euclidean distance exceeds a preset displacement threshold are marked as outliers and discarded. The coordinate sequence after outlier removal is smoothed using Kalman filtering to compensate for multipath propagation errors in the positioning signal, resulting in clean spatial trajectory data.The control terminal re-aligns and binds the cleaned EEG data, eye-tracking data, and spatial trajectory data according to a unified timestamp, generating multimodal interest data for each acquisition target.

[0035] Step S30: Align the EEG data, eye movement data and spatial trajectory data of each of the acquisition targets according to the timestamp, and then perform spatial registration with the spatial location information in the booth spatial layout data to obtain the interest association data of each acquisition target at each spatial location unit.

[0036] In this embodiment, spatial registration is the process of unifying the spatial coordinates or line-of-sight directions carried by various types of data in multimodal interest data to the same three-dimensional spatial coordinate system and establishing a mapping relationship between the data and physical spatial locations. A spatial location unit is a number of non-overlapping and fully covered spatial sub-regions divided according to predetermined rules within the physical space of the exhibition booth. Each spatial sub-region has a defined three-dimensional spatial coordinate range. Interest-related data is a structured data record formed by associating and binding the brainwave data and eye-tracking data of the target at a specific time with the spatial location unit identifiers determined after spatial registration.

[0037] As an optional implementation, the control terminal first reads the gaze coordinates of the target at each moment from the multimodal interest data. These gaze coordinates are initially located in the two-dimensional image coordinate system of the eye-tracking acquisition device itself. The control terminal obtains the factory calibration parameters of the eye-tracking acquisition device, which include the rotation matrix and translation vector between the image coordinate system and the three-dimensional spatial coordinate system. Using the two-dimensional gaze coordinates at that moment and the three-dimensional coordinates in the target's spatial trajectory data at that moment as input, the control terminal maps the two-dimensional gaze coordinates to the three-dimensional spatial coordinate system, obtaining the three-dimensional gaze spatial coordinates of the gaze point within the booth's physical space at that moment. The control terminal then divides the coordinate range of the booth's physical space into multiple spatial location units according to either the exhibit granularity or the grid granularity. If the exhibit granularity is used as the dividing basis, the control terminal directly reads the spatial location information of each exhibit in the booth's spatial layout data, using the spatial area enclosed by the outline boundary of each exhibit as a spatial location unit, and binding the exhibit's identifier to that unit. If grid granularity is used as the dividing line, the control terminal reads the boundary values ​​of the physical space coordinates of the booth along each coordinate axis to obtain preset grid size parameters. These parameters include the grid width in the X-axis direction and the grid depth in the Y-axis direction. Starting from the minimum X-value of the coordinate range, the control terminal divides the booth plane into multiple rectangular grid units with the grid width as the step size and the grid depth as the step size, starting from the minimum Y-value. Each grid unit covers the entire range from the ground to the top of the booth in the Z-axis direction. Each grid unit serves as an independent spatial location unit and is assigned a unique unit number. After the division is completed, the control terminal compares the three-dimensional gaze space coordinates at each moment with the spatial range of each spatial location unit, and identifies the spatial location unit corresponding to the spatial range to which the three-dimensional gaze space coordinates belong as the spatial location unit corresponding to the gaze point at that moment. The control terminal further calculates the gaze direction vector from the target location to the gaze point based on the three-dimensional coordinates in the spatial trajectory data of the target at that moment and the determined three-dimensional gaze spatial coordinates. It also reads the coordinates of the center point of the spatial range of the spatial location unit corresponding to the gaze point at that moment, calculates the spatial vector from the target location to that center point, and then calculates the cosine of the angle between the gaze direction vector and the spatial vector, and inversely calculates the angle value. The control terminal compares this angle value with a preset angle threshold of 30 degrees. If the angle value is less than 30 degrees, it is determined that the gaze direction at that moment points geometrically to the spatial location unit, possessing spatial geometric consistency. The EEG data and eye movement data at that moment are associated with the spatial location unit identifier and recorded to form a valid interest association data. If the angle value is greater than or equal to 30 degrees, it is determined that the gaze direction at that moment is inconsistent with the geometric position of the spatial location unit, and the associated data at that moment is discarded.The control terminal repeatedly performs the above mapping, comparison, angle determination and association recording operations for each acquisition target at all times, and summarizes the effective interest association data at each time in chronological order to obtain the interest association data of each acquisition target at each spatial location unit.

[0038] Step S40: Determine the group interest metric value at each spatial location unit based on the interest association data of each of the collected targets at each spatial location unit.

[0039] In this embodiment, the group interest metric is a numerical indicator used to quantify the overall interest response intensity of multiple acquisition targets at the same spatial location unit. This indicator is jointly determined by the group statistics of EEG data and eye movement data of all acquisition targets within the same spatial location unit.

[0040] As an optional implementation, the control terminal traverses all the divided spatial location units. For the currently processed spatial location unit, it filters out all records whose spatial location unit identifier matches the identifier of the current spatial location unit from the interest association data of each acquisition target, and constructs the interest association data set of the spatial location unit. The control terminal performs group statistics on the EEG data contained in all records in the interest association data set: it applies a fast Fourier transform to the EEG time-series signal in each record to transform the signal from the time domain to the frequency domain, extracts the power spectrum energy of the 8 Hz to 13 Hz frequency band as the alpha wave power value, and extracts the power spectrum energy of the 13 Hz to 30 Hz frequency band as the beta wave power value. It sums the alpha wave power values ​​of all records in the current spatial location unit at their respective times and divides the sum by the total number of records to obtain the group mean of the alpha wave power of the spatial location unit. It sums the beta wave power values ​​of all records in the current spatial location unit and divides the sum by the total number of records to obtain the group mean of the beta wave power of the spatial location unit. The control terminal continues to perform group statistics on the eye-tracking data contained in all records of the interest-related data set: After deduplicating the fixation points of each acquisition target within the spatial location unit at each recording time, the cumulative number of fixation points is obtained. The fixation duration of each acquisition target within the spatial location unit at each recording time is accumulated separately according to the target dimension, and then the cumulative values ​​of all acquisition targets are summed to obtain the cumulative fixation duration value. The number of acquisition targets that have generated at least one fixation record within the spatial location unit is counted to obtain the number of acquisition targets that generated fixation. The control terminal obtains a preset set of weight parameters, which includes a first weight corresponding to the group mean of β-wave power, a second weight corresponding to the group mean of α-wave power, a third weight corresponding to the cumulative number of fixation points, a fourth weight corresponding to the cumulative fixation duration value, and a fifth weight corresponding to the number of acquisition targets that generated fixation. The control terminal calculates the group interest metric value of the spatial location unit according to the following formula: Group interest metric = First weight multiplied by the ratio of the group mean β wave power to the preset β baseline value, plus the second weight multiplied by the ratio of the preset α baseline value to the group mean α wave power, plus the third weight multiplied by the ratio of the cumulative number of fixation points to the preset fixation number baseline value, plus the fourth weight multiplied by the ratio of the cumulative fixation duration to the preset fixation duration baseline value, plus the fifth weight multiplied by the ratio of the number of data collection targets that generated fixation to the preset number baseline value.

[0041] The control terminal repeatedly performs the above data filtering, EEG group statistics, eye-tracking group statistics, and weighted summation calculation for each spatial location unit to obtain the group interest metric value corresponding to each spatial location unit.

[0042] Step S50: Generate a layout optimization scheme for the physical space of the booth based on the group interest metric value at each spatial location unit and the booth space layout data.

[0043] In this embodiment, the layout optimization scheme is a digital scheme file containing the adjustment results of the spatial location information of each exhibit within the physical space of the booth.

[0044] As an optional implementation, the control terminal associates the calculated group interest metric value at each spatial location unit with the unique unit number of the spatial location unit. Based on the attribution relationship between the spatial location unit identifier and the exhibit identifier, it aggregates the average group interest metric values ​​of all spatial location units occupied by the same exhibit to obtain the comprehensive interest metric value for each exhibit. The control terminal sorts all exhibits from highest to lowest comprehensive interest metric value, generating an exhibit interest ranking result. The control terminal reads the spatial coordinates of the booth entrance center point from the booth spatial layout data and calculates the spatial distance between the center point of the current spatial location information of each exhibit in the ranking list and the center point of the booth entrance. The control terminal obtains a preset pre-quantity parameter N and marks the exhibits ranked 1st to Nth in the exhibit interest ranking result as high-priority exhibits. The control terminal obtains a preset entrance distance threshold and adjusts the current spatial location information of each high-priority exhibit to a region where the spatial distance between its center point and the center point of the booth entrance is less than the entrance distance threshold, and the adjusted spatial location information does not exceed the coordinate range of the booth's physical space. The control terminal then reads the preset functional association attributes of each exhibit from the booth space layout data. These preset functional association attributes are functional category tags pre-labeled by the booth designer for each exhibit, including product display category, interactive experience category, and information consultation category. The control terminal groups all exhibits according to their functional category tags and marks exhibits with the same functional category tag as related exhibit groups. For each related exhibit group, the control terminal obtains a preset spacing threshold and adjusts each exhibit within the related exhibit group to an adjacent area where the spatial distance between them is less than the spacing threshold, while ensuring that the spatial position information of each exhibit after adjustment does not exceed the coordinate range of the booth's physical space. The control terminal writes the adjusted spatial position information of each exhibit into the corresponding exhibit's spatial position information field in the booth space layout data, overwriting the original data, and generating a layout optimization scheme that includes the spatial position information of each exhibit's tags.

[0045] As an alternative implementation, after sorting the exhibits, the control terminal further acquires the spatial location information of preset fixed facilities within the physical space of the booth. These fixed facilities include the locations of the booth's load-bearing columns, fire-fighting equipment, and power supply interfaces. When adjusting the spatial location of high-priority exhibits, the control terminal first performs overlap detection between the spatial location information of each high-priority exhibit and the spatial location information of the fixed facilities. If an overlap is detected between the adjusted exhibit spatial location and the fixed facility spatial location, the target spatial location of the exhibit is redefined within a preset safe avoidance distance around the fixed facility until there is no overlap between the exhibit spatial location and the fixed facility spatial location, and the spatial distance between the center point of the exhibit and the center point of the booth entrance is still less than the entrance distance threshold. This operation method solves the technical problem that fixed facilities in the actual physical space of the booth constitute physical constraints on the placement of exhibits, enabling the generated layout optimization scheme to meet the needs of interest optimization while also possessing physical space feasibility.

[0046] For example, a beauty brand set up a 36-square-meter pop-up store in the atrium of a shopping mall. The booth included four sections: a makeup trial area, a product display area, an interactive mirror area, and an information desk. A control terminal established a three-dimensional coordinate system with the center point of the booth entrance as the origin. A laser rangefinder measured the coordinates of the booth's boundary inflection points and the outer contour boundary points of the four sections, generating spatial layout data for the booth. Thirty recruited consumers, acting as the data collection targets, entered the booth in turn and freely explored. Each consumer wore a portable EEG headband, eye-tracking glasses, and an ultra-wideband positioning tag suspended from their waist, allowing them to move freely and experience the various sections at their own discretion. The control terminal collected real-time prefrontal cortex EEG signals, gaze coordinates, and three-dimensional spatial position data for each consumer. After filtering, noise reduction, and coordinate mapping, the three-dimensional coordinates of the gaze point at each moment were obtained within the booth's spatial coordinate system. The control terminal divides the booth layout into 0.5m x 0.5m grid units. It spatially matches the gaze coordinates of each consumer with the grid units and calculates the angle between the gaze direction and the target grid unit. Only gaze records with an angle less than 30 degrees are retained as valid interest-related data. The control terminal statistically analyzes the average beta-wave power, average alpha-wave power, cumulative number of gaze points, cumulative gaze duration, and the number of people who gaze at each consumer within each grid unit. Weighted calculations yield the group interest metric for each grid unit. The two grid units with the highest group interest metrics correspond to the center positions of the makeup trial area and the interactive mirror area, respectively, while the group interest metric for the product display area is lower than the average level of all grid units. Based on this result, the control terminal marks the makeup trial area and the interactive mirror area as high-priority exhibits. While ensuring they do not exceed the booth's coordinate range, these two exhibits are moved to the area closest to the entrance. The product display area is also moved adjacent to the makeup trial area to leverage the high foot traffic in the makeup trial area for customer attraction. The control terminal ultimately outputs a layout optimization plan containing the spatial coordinates of the four exhibits after adjustments. After the booth builder rearranged the booth according to the plan, the average dwell time of subsequent batches of consumers in the makeup trial area and interactive mirror area increased by about 40% compared to before the adjustment.

[0047] This embodiment spatially registers the gaze coordinates from eye-tracking data with the three-dimensional spatial coordinates from spatial trajectory data and the spatial location information of each exhibit from the booth layout data. It then determines the angle between the gaze direction vector at each moment and the spatial vector from the target to the exhibit. Gaze records that satisfy spatial geometric consistency are bound to EEG data as interest-related data, ensuring that subsequent calculations of group interest metrics are based solely on the physical spatial location where visual interaction actually occurs. The group interest metric is calculated by weighting and fusing the group mean of beta-wave power, group mean of alpha-wave power, cumulative number of gaze points, cumulative gaze duration, and number of people exhibiting gaze within each spatial location unit. This integrates the cognitive input represented by EEG physiological response and the visual attention represented by eye-tracking data into a unified quantitative indicator. By prioritizing high-priority exhibits based on the ranking results of group interest metrics and keeping the adjusted positions within the coordinate range, the layout optimization scheme directly maps the distribution of visitor interest responses to the physical coordinates of exhibits. This avoids the single strategy of traditional schemes that only select locations based on the maximum channel traffic, resulting in a more balanced distribution of group interest metrics in each spatial unit of the optimized booth layout.

[0048] Based on any of the above embodiments, in Embodiment 2 of this application, obtaining booth space layout data includes: Step S11: Using the designated location of the physical space of the booth as the origin, establish a three-dimensional spatial coordinate system covering the physical space of the booth, and determine the coordinate range of the physical space of the booth.

[0049] In this embodiment, the three-dimensional spatial coordinate system uses a fixed physical point within the booth's physical space as the origin, and three orthogonal spatial directions as the X-axis, Y-axis, and Z-axis, respectively, to uniquely describe the spatial measurement benchmark of any position within the booth's physical space. The coordinate range of the booth's physical space is the set of continuous intervals occupied by the booth's physical space along each coordinate axis in the three-dimensional spatial coordinate system, defined by the minimum and maximum values ​​of the booth boundary in the X-axis, Y-axis, and Z-axis directions.

[0050] As an optional implementation, the control terminal uses the vertical projection of the center point of the booth's physical entrance onto the ground as the origin of the coordinate system. The component of the normal direction pointing inwards from the booth's entrance plane is taken as the positive X-axis, the direction parallel to the entrance plane and lying on the ground plane is taken as the Y-axis, and the direction perpendicular to the ground and upwards is taken as the positive Z-axis, establishing a right-handed three-dimensional coordinate system. The control terminal acquires the maximum and minimum extension distances of the booth boundary along the X-axis and Y-axis, as well as the maximum height of the booth's top structure along the Z-axis, measured by a laser rangefinder. The intervals defined by the minimum and maximum extension distances along the X-axis, Y-axis, and the Z-axis from zero to the maximum height are collectively determined as the coordinate range of the booth's physical space, where the zero value of the Z-axis is the plane containing the ground. In related technologies, a coordinate system is typically established using the center point of the booth or a corner point on the booth floor plan as a reference. However, this method does not consider the actual entry direction of visitors, resulting in the relative orientation between the visitor's line of sight and the booth entrance requiring additional rotation and translation transformations for proper interpretation during subsequent spatial registration. This increases the computational load and error accumulation paths in the data processing stage. This embodiment establishes a coordinate system using the entrance center point as the origin. This ensures that the spatial trajectory data of the target and the coordinates of the gaze point, after spatial registration, naturally possess orientation information with the entrance as a reference, avoiding additional coordinate system rotation transformations and reducing error propagation paths in the data registration process.

[0051] Step S12: In the three-dimensional spatial coordinate system, the coordinates of the outer contour boundary points of each exhibit within the physical space of the booth are measured by a ranging device. The spatial area enclosed by the coordinates of the outer contour boundary points is determined as the spatial location information of the corresponding exhibit, and an exhibit identifier is assigned to each exhibit.

[0052] In this embodiment, the ranging device is an electronic measuring instrument capable of measuring the distance and azimuth of a target point relative to a measurement reference point and outputting the measurement results as spatial coordinates. The coordinates of the outer contour boundary points of the exhibit are the coordinate values ​​in a three-dimensional spatial coordinate system of a set of discrete sampling points located on the physical boundary of the exhibit. These sampling points are distributed along the exhibit boundary and are sufficient to describe the projected shape of the exhibit on the horizontal plane and its coverage area in the height direction. The exhibit identifier is a numerical number or string code used to distinguish different exhibits during data storage and processing.

[0053] As an optional implementation, after establishing the three-dimensional spatial coordinate system, the control terminal uses a laser rangefinder to sequentially perform the following measurement operations on each exhibit within the physical space of the booth: The operator of the control terminal fixes the tripod of the laser rangefinder at the origin of the coordinate system and completes horizontal calibration. Then, the operator holds the reflective target of the laser rangefinder and places it sequentially at each boundary inflection point of the outer contour of the current exhibit. The reflective target is attached to the ground or the outer edge of the exhibit's surface. At each boundary inflection point, the laser rangefinder emits a laser pulse towards the reflective target and receives the reflected signal. The spatial distance between the rangefinder and the reflective target is calculated based on the flight time of the laser pulse. Simultaneously, the horizontal and vertical deflection angles provided by the attitude sensor inside the rangefinder are read. The control terminal then calculates the X-axis, Y-axis, and Z-axis coordinates of the boundary inflection point in the three-dimensional spatial coordinate system through trigonometric transformation based on the spatial distance, horizontal deflection angle, and vertical deflection angle. The operator moves the reflective target sequentially along the outer contour of the exhibit, while the control terminal records the coordinates of all boundary inflection points. The closed spatial area formed by connecting the coordinates of all boundary inflection points in sequence is then defined as the spatial location information of the exhibit. After measuring one exhibit, the operator repeats the above process for the next exhibit until the coordinates of all boundary inflection points of all exhibits have been measured. After completing the measurement of all exhibits, the control terminal assigns a unique exhibit identifier to each exhibit from a preset coding rule. This identifier is then bound and stored with the exhibit's spatial location information in subsequent steps. Related technologies typically use an approximation method to record the exhibit's position as the coordinates of a center point or a simple rectangular frame. This approximation method loses the boundary details of the actual space occupied by the exhibit in the layout optimization of large-sized or irregularly shaped exhibits, making it impossible to accurately determine whether the gaze point falls inside or outside the exhibit's outer contour during subsequent spatial registration. This embodiment measures the coordinates of the boundary points of the outer contour of the exhibit and uses the spatial area enclosed by the boundary points as spatial location information. This provides a precise spatial geometric basis for determining whether the gaze point points to the exhibit during the spatial registration stage, avoiding misjudgments of gaze attribution caused by overly coarse location information.

[0054] Step S13: Associate and store the exhibit identifier and corresponding spatial location information of each exhibit with the coordinate range of the physical space of the booth to generate the booth space layout data.

[0055] In this embodiment, the booth space layout data is a structured data file containing the coordinate range of the booth's physical space and the mapping relationship between the exhibit identifiers and spatial location information of each exhibit within the coordinate range.

[0056] As an optional implementation, the control terminal writes the coordinate range of the booth's physical space determined in step S11 into the spatial range field of a preset data template file in the form of a coordinate interval array. The control terminal then iterates through all exhibits that were measured in step S12. For the currently iterated exhibit, it writes the exhibit's identifier into the identifier field of a newly added record entry in the data template file. It then sequentially writes the coordinates of all outer contour boundary points contained in the exhibit's spatial location information into the contour coordinate array field of that record entry. The spatial location information field value of this record entry is automatically associated with the coordinate range field value of the booth's physical space through a predefined hierarchical relationship in the data template file; that is, the spatial location information of this record entry is constrained to a subset of the coordinate range of the booth's physical space. The control terminal repeats the above writing operation for each exhibit until the exhibit identifiers and corresponding spatial location information of all exhibits have been written into the data template file. The control terminal stores the data template file with all data filled in its local non-volatile memory, generating booth space layout data.

[0057] This embodiment establishes a three-dimensional spatial coordinate system with a designated location in the physical space of the exhibition booth as the origin and defines the coordinate range. This provides a unified geometric reference for all subsequent spatial registration operations. This geometric reference covers the entire usable area of ​​the physical space of the exhibition booth, ensuring that the eye-tracking gaze point coordinates and spatial trajectory coordinates of the acquired targets can obtain a clear physical location meaning under this reference. By measuring the coordinates of the outer contour boundary points of each exhibit using a ranging device and defining the spatial area enclosed by the outer contour boundary point coordinates as the spatial location information of the exhibit, each exhibit has an accurate and complete spatial occupancy description in the three-dimensional spatial coordinate system. This description includes the actual shape and size information of the exhibit, enabling the spatial attribution judgment between the gaze point and the exhibit in subsequent spatial registration to be performed at the level of the real geometric boundary, avoiding spatial attribution deviations caused by simplifying the exhibit location information to a single coordinate point or a simplified rectangle. By associating and storing exhibit identifiers and corresponding spatial location information with coordinate ranges to generate booth spatial layout data, a structured mapping relationship between exhibits and their physical spatial locations is established. This data structure can be directly indexed and retrieved in subsequent steps, reducing the repeated reading and parsing of original measurement data during the layout optimization process and improving data processing efficiency.

[0058] Based on any of the above embodiments, in Embodiment 3 of this application, acquiring multimodal interest data of multiple acquisition targets within the coordinate range includes: Step S21: Collect initial EEG data and initial eye movement data of the target within the coordinate range.

[0059] In this embodiment, the initial EEG data is the raw, multi-channel voltage timing signal acquired by a portable EEG acquisition device from the scalp surface of the target acquisition point without any filtering or artifact removal processing. The initial eye-tracking data is the raw, unprocessed sensor data sequence acquired by an eye-tracking acquisition device from the target eye, including pupil center coordinates, corneal reflector coordinates, and infrared image frames.

[0060] As an optional implementation, after the target enters the coordinate range of the booth's physical space, the on-site operator equips the target with a portable EEG acquisition device and a head-mounted eye-tracking acquisition device. The four electrodes of the portable EEG acquisition device are attached to the target's scalp at the left prefrontal lobe (Fp1), right prefrontal lobe (Fp2), left temporal lobe (T3), and right temporal lobe (T4) according to the international 10-20 system standard. The reference electrode is attached to the mastoid process behind the target's ear. The operator activates the portable EEG acquisition device, which continuously acquires scalp potential signals from four channels at a sampling rate of 250 Hz. At each sampling moment, a four-dimensional voltage vector is acquired, with each component corresponding to the instantaneous potential amplitude at an electrode location, measured in microvolts. Simultaneously, the operator activates the head-mounted eye-tracking acquisition device. The device's two built-in miniature infrared cameras are pointed towards the target's left and right eyes, respectively, continuously acquiring infrared image frames of the bilateral area at a sampling rate of 60 Hz. Each frame contains two-dimensional pixel coordinates of the pupil center and corneal reflection point. After the target began freely viewing the exhibit within the physical space of the booth, the portable EEG acquisition device continuously transmitted raw voltage timing signals to the control terminal as initial EEG data, and the head-mounted eye-tracking acquisition device continuously transmitted infrared image frame sequences to the control terminal as initial eye-tracking data. The control terminal added a timestamp to each received data point using a global clock signal. In related technologies, the acquisition of eye-tracking and EEG data is usually conducted in scenarios where the subject's head is fixed or their body is restricted, and the device calibration parameters remain unchanged during the acquisition process. However, in this embodiment, the target was moving freely within the physical space of the booth, and their head posture and body orientation were constantly changing. If the acquisition method for a fixed scenario were directly used, the coordinates of the pupil center and corneal reflex point in the eye-tracking data would be systematically offset due to head movement, failing to reflect the true direction of gaze. At the same time, the amplitude of electromyography artifacts caused by body movement in the EEG signal would be much higher than the amplitude of the EEG signal, completely masking the true EEG rhythm components. This embodiment configures a portable EEG acquisition device and a head-mounted eye-tracking acquisition device for the target and continuously collects raw data during free viewing. This allows the initial EEG data and initial eye-tracking data to fully record the physiological responses corresponding to the target's natural viewing behavior in real physical space. This avoids data distortion caused by the fixed scene acquisition method due to the restriction of the subject's activities. At the same time, subsequent preprocessing steps are used to remove noise and artifacts caused by movement.

[0061] Step S22: Collect initial spatial trajectory data of the target using a spatial positioning tag, wherein the spatial positioning tag is configured to receive signals from positioning base stations deployed within the physical space of the booth.

[0062] In this embodiment, the spatial positioning tag is an active electronic tag carried by the target and capable of receiving external positioning signals and calculating its own spatial position based on the received signals. The positioning base station is a signal transmitting device fixedly deployed within the physical space of the exhibition booth, with known spatial coordinates. Multiple positioning base stations work collaboratively to achieve three-dimensional spatial positioning of the spatial positioning tag. The initial spatial trajectory data is a sequence of three-dimensional spatial coordinates continuously recorded and output by the spatial positioning tag in chronological order during the target's free viewing of the exhibition.

[0063] As an optional implementation, before the target enters the physical space of the booth, on-site operators attach a spatial positioning tag to the waist of the target. This spatial positioning tag has a built-in ultra-wideband signal receiving module and data processing unit. A first, second, third, and fourth positioning base station are deployed at the four corners of the physical space of the booth, respectively. The precise spatial coordinates of the four positioning base stations within the coordinate range of the physical space of the booth have been pre-calibrated and stored in the control terminal. After the target enters the coordinate range of the physical space of the booth, the spatial positioning tag simultaneously transmits an ultra-wideband pulse signal at a frequency of 10 Hz to the four positioning base stations. Each positioning base station receives the pulse signal, records the pulse arrival time, and immediately reports the arrival time to the control terminal via a wired or wireless link. Based on the coordinate values ​​of the four positioning base stations and the recorded pulse arrival times, the control terminal uses a time difference of arrival algorithm to calculate the X-axis, Y-axis, and Z-axis coordinates of the spatial positioning tag in the three-dimensional spatial coordinate system. Specifically, the control terminal uses the first positioning base station as a reference base station, calculates the pulse arrival time difference between the second and first positioning base stations, and multiplies this time difference by the speed of pulse signal propagation in the air to obtain the distance difference. Geometrically, this distance difference represents the difference between the distance from the spatial positioning tag to the second positioning base station and the distance to the first positioning base station. Based on this distance difference and the known coordinates of the first and second positioning base stations, the control terminal constructs a hyperbolic equation. The control terminal repeats the above calculation for the third and fourth positioning base stations respectively, obtaining two more hyperbolic equations. The coordinates of the intersection of the three hyperbolic equations are the current position of the spatial positioning tag in the three-dimensional spatial coordinate system. The control terminal repeats the above pulse transmission, arrival time recording, and hyperbolic equation solving process at a frequency of 10 Hz at each sampling moment to obtain a continuously changing three-dimensional spatial coordinate sequence of the spatial positioning tag as the target moves. This coordinate sequence, arranged in chronological order, constitutes the initial spatial trajectory data.

[0064] Step S23: The initial EEG data, initial eye movement data and initial spatial trajectory data of each of the acquisition targets are preprocessed to remove invalid data segments and noise data in each data stream, so as to obtain the EEG data, eye movement data and spatial trajectory data corresponding to each acquisition target as the multimodal interest data.

[0065] In this embodiment, preprocessing involves a series of operations to clean and format the raw sensor data to eliminate noise components and invalid data segments introduced during the acquisition process. Invalid data segments are data fragments whose quality is lower than a preset usable standard due to excessive movement of the target, brief device disconnection, or environmental interference. Noise data consists of non-target signal components superimposed on valid physiological and position signals.

[0066] Optionally, step S23 includes: Step S231: Bandpass filtering is performed on the initial EEG data to retain the EEG signal components within the target frequency band. The electrooculography (EOG) artifacts and electromyography (EMG) artifacts in the EEG data are identified and removed using an independent component analysis algorithm to obtain the EEG data.

[0067] In this embodiment, the independent component analysis algorithm is a blind source separation algorithm that decomposes a multi-channel mixed signal into several statistically independent source signal components. The electrooculography (EOG) artifact component is an interference component mixed into the EEG signal due to potential changes caused by the target's blinking or eye movements. This component typically has an amplitude above 100 microvolts and a characteristic pulse shape. The electromyography (EMG) artifact component is an interference component mixed into the EEG signal due to potential activity caused by the tension and contraction of the target's facial or neck muscles. This component typically has a frequency above 30 Hz and a wide energy distribution.

[0068] As an optional implementation, the control terminal reads four-channel voltage timing signals from initial EEG data. Each channel contains N sampling points, and the four-channel signals are arranged into a 4xN matrix M. The control terminal loads preset bandpass filter coefficients. This bandpass filter has a frequency response with half-power attenuation at 0.5 Hz and half-power attenuation at 50 Hz, and a flat gain in the 8 Hz to 30 Hz frequency band. The control terminal sequentially passes the timing signals of each row in matrix M through this bandpass filter. The filtering operation uses a zero-phase digital filtering method to prevent phase shift of the signal. After bandpass filtering, DC drift and baseline fluctuation components below 0.5 Hz in matrix M are filtered out, as are high-frequency circuit noise and some electromyographic components above 50 Hz. The remaining signal components are mainly concentrated in the 0.5 Hz to 50 Hz frequency band. The control terminal then applies a fast independent component analysis algorithm to the filtered matrix M. The specific execution process of the algorithm is as follows: First, the matrix M is centered by subtracting the mean of each row, making the mean of the signals in each row zero. Then, the centered matrix is ​​whitened by obtaining the eigenvalue matrix and eigenvector matrix through principal component analysis, projecting the signals onto the directions of each principal component, and scaling them to unit variance to obtain the whitening matrix Z. The control terminal initializes a 4x4 separation matrix W, whose initial value is a randomly generated fourth-order identity matrix. The control terminal enters an iterative optimization loop, updating each column vector of the separation matrix W one by one according to a preset negative entropy maximization criterion in each iteration, maximizing the negative entropy of each independent component after projection, i.e., maximizing the statistical independence between components. The iteration continues until the change in the W matrix is ​​less than the preset convergence threshold of 0.0001. After convergence, the control terminal calculates Y = W × Z to obtain four independent components, each of which is a four-row vector representing a source signal separated from the mixed signal. The control terminal calculates the correlation coefficient between each independent component and a pre-established electrooculogram (EOG) artifact template, as well as the correlation coefficient between each independent component and a pre-established electromyogram (EMG) artifact template. The EOG artifact template is obtained by time-domain averaging and amplitude normalization of a large number of independently acquired blink signals, while the EMG artifact template is obtained by frequency-domain feature extraction of independently acquired large-amplitude EMG signals. If the correlation coefficient between an independent component and the EOG artifact template is greater than a preset first correlation coefficient threshold of 0.75, or the correlation coefficient with the EMG artifact template is greater than a preset second correlation coefficient threshold of 0.70, the control terminal marks that independent component as an artifact component. The control terminal removes one or more independent components marked as artifact components from the Y matrix, and applies the inverse of the separation matrix W to the remaining unmarked independent components for spatial projection, remapping the signals in the independent component space back to the channel space, obtaining clean EEG time-series signals for four channels. This signal is the pre-processed EEG data.Related technologies typically use fixed-band filters to process EEG signals, which cannot distinguish between EEG components with overlapping frequency bands and artifact components. This results in the inability to effectively remove residual electrooculography (EOG) and electromyography (EMG) artifacts located within the same frequency band while retaining target frequency band signals such as alpha and beta waves. Consequently, the calculated alpha and beta wave power values ​​in subsequent feature extraction include the energy contribution of artifact components. This embodiment further decomposes the multi-channel signal into independent source signal components using an independent component analysis algorithm after bandpass filtering. Based on a preset artifact template, correlation discrimination is performed to accurately identify and remove artifact components. This ensures that the alpha and beta wave components retained in the final obtained EEG data truly reflect the EEG activity of the target rather than artifact interference.

[0069] Step S232: The initial eye-tracking data is subjected to invalid frame removal processing. The invalid frames include off-screen data frames and frame data with acquisition confidence lower than a preset confidence threshold. The initial eye-tracking data is then corrected in spatial coordinate system based on the factory calibration parameters of the eye-tracking acquisition device, and the effective gaze point sequence is extracted to obtain the eye-tracking data.

[0070] In this embodiment, the off-screen data frame is the infrared image frame captured when the gaze direction of the target exceeds the field of view of the eye-tracking acquisition device. Valid pupil center coordinates or corneal reflector coordinates cannot be extracted from this frame. The valid fixation point sequence is the set of fixation point coordinates arranged in chronological order after removing invalid frames and performing spatial coordinate correction on the remaining data frames.

[0071] As an optional implementation, the control terminal reads the two-dimensional pixel coordinates of the pupil center and the two-dimensional pixel coordinates of the corneal reflector point corresponding to each frame of the initial eye-tracking data. For the currently read frame, the control terminal first checks whether the pupil center coordinates in the frame are within the pixel range of the horizontal and vertical directions of the image. If the horizontal coordinate value of the pupil center coordinates is less than the lower limit of the image width (0) or greater than the upper limit of the image width, or the vertical coordinate value of the pupil center coordinates is less than the lower limit of the image height (0) or greater than the upper limit of the image height, then the frame is marked as an off-screen data frame. The control terminal continues to check whether the corneal reflector point coordinates in the frame are within the pixel range of the image. If the horizontal or vertical coordinates of the corneal reflector point coordinates exceed the image boundary, the frame is also marked as an off-screen data frame. After completing the off-screen determination, the control terminal reads the confidence value output to the frame by the confidence evaluation module built into the eye-tracking acquisition device. This confidence value is calculated by the eye-tracking acquisition device based on the clarity of the current frame's infrared image, the convergence degree of the pupil detection algorithm, and the recognizability of the corneal reflector point, and its value ranges from 0 to 1. The control terminal compares the confidence value with a preset confidence threshold of 0.7. If the confidence value of the frame is less than 0.7, the frame is marked as a low-confidence data frame. All frames marked as off-screen or low-confidence data frames are discarded and not included in subsequent fixation point coordinate calculations. For each remaining frame, the control terminal obtains the x and y coordinates of the pupil center and the corneal reflection point. The pupil-corneal reflection vector is obtained by subtracting the corneal reflection point coordinates from the pupil center coordinates. The control terminal reads the intrinsic parameter matrix stored in the eye-tracking acquisition device's factory calibration file. This intrinsic parameter matrix is ​​a 3x3 matrix containing the device's optical center coordinates, pixel focal length, and distortion parameters. The control terminal expands the pupil-corneal reflection vector into three-dimensional homogeneous coordinates and multiplies it by the inverse of the intrinsic parameter matrix to obtain the unit direction vector of the gaze direction in the eye-tracking device's coordinate system. The control terminal calculates the three-dimensional coordinates of the fixation point in the eye-tracking device's coordinate system at that moment based on the unit direction vector and the preset gaze projection distance parameters. The control terminal reads the rotation matrix and translation vector between the eye-tracking device's coordinate system and the three-dimensional spatial coordinate system of the booth's physical space. This rotation matrix and translation vector are obtained through multiple observations of the calibration board by the eye-tracking device during on-site deployment at the booth. The control terminal multiplies the three-dimensional coordinates of the fixation point by the aforementioned rotation matrix and adds the translation vector to obtain the three-dimensional fixation spatial coordinates of the fixation point in the three-dimensional spatial coordinate system. The control terminal repeats the above pupil-corneal reflection vector calculation, intrinsic parameter matrix correction, and extrinsic parameter matrix transformation operations for each remaining frame to obtain the corresponding three-dimensional fixation spatial coordinates for each frame. The three-dimensional fixation spatial coordinates of all frames are arranged in chronological order to form a valid fixation point sequence, which is the preprocessed eye-tracking data. In related technologies, the preprocessing of eye-tracking data typically ignores invalid frame removal and spatial coordinate system correction, directly using the fixation point coordinates output by the device for subsequent analysis.In scenarios where the target is freely walking, frequent head turns generate a large number of off-screen data frames. If these frames are not removed, the invalid or low-quality coordinate data contained within them will directly contaminate the matching results between gaze points and exhibits in subsequent spatial registration. This embodiment eliminates data points that cannot accurately reflect the gaze direction of the target by removing off-screen and low-confidence data frames. This ensures that only data frames with clear gaze directions and reliable coordinate accuracy are retained in the effective gaze point sequence. Simultaneously, the pixel-level pupil-corneal reflection vector is converted into a gaze direction vector in physical space using the intrinsic parameter matrix in the factory calibration parameters. Then, the gaze direction vector is mapped to the three-dimensional spatial coordinate system of the exhibit's physical space using the extrinsic parameter matrix. This ensures that the spatial physical meaning of each coordinate value in the effective gaze point sequence is consistent and completely matches the coordinate system in the exhibit's spatial layout data.

[0072] Step S233: Outlier removal is performed on the initial spatial trajectory data, and smoothing filtering is applied to the initial spatial trajectory data after outlier removal to obtain the spatial trajectory data.

[0073] In this embodiment, outliers are data points whose spatial coordinates deviate significantly from the actual physical location of the target due to positioning signal reflection, multipath interference, or momentary loss of device lock. The degree of deviation is manifested in the displacement between the coordinates of the point and the coordinates of the adjacent points before and after it far exceeds the maximum physical distance that the target may move within the adjacent sampling interval.

[0074] As an optional implementation, the control terminal reads the three-dimensional spatial coordinate sequence of the target at each sampling time from the initial spatial trajectory data. Each coordinate includes an X-axis value, a Y-axis value, and a Z-axis value. The control terminal sets the sliding window length to 7 frames and processes the data sequentially starting from the first coordinate in the coordinate sequence. At the current sliding window position, the control terminal acquires the seven three-dimensional coordinates of frames 1 to 7 within the window, and calculates the average values ​​of the coordinates of the remaining six frames (excluding the middle frame, i.e., frame 4) in the X-axis, Y-axis, and Z-axis directions to obtain the average values ​​of the reference coordinates. The control terminal calculates the Euclidean distance between the coordinates of frame 4 within the window and the aforementioned reference coordinates. This Euclidean distance is the straight-line distance between the coordinates of frame 4 and the reference coordinates in three-dimensional space. The control terminal compares the Euclidean distance with a preset displacement abruptness threshold. This threshold is the maximum physical distance the target can move within two adjacent sampling intervals. The specific value is calculated based on the positioning tag's sampling frequency of 10 Hz and the upper limit of normal human walking speed of 2 meters per second. Dividing the upper limit of human walking speed by the sampling frequency and multiplying by a safety factor of 3 yields a displacement abruptness threshold of 0.6 meters. If the Euclidean distance is greater than 0.6 meters, the control terminal marks the 4th frame as an outlier and removes it from the coordinate sequence. If the Euclidean distance is less than or equal to 0.6 meters, the control terminal retains the 4th frame. The control terminal moves the sliding window forward one frame, forming a new sliding window from the 2nd to the 8th frame, and repeats the above calculation and judgment until the sliding window reaches the end of the coordinate sequence. After completing the detection and removal of all outliers, the control terminal obtains the spatial coordinate sequence after removing outliers. The control terminal then applies a Kalman filter to smooth this spatial coordinate sequence after removing outliers. The state vector of the Kalman filter is set as a six-dimensional vector, with the first three dimensions being the X-axis, Y-axis, and Z-axis coordinates, and the last three dimensions being the X-axis velocity, Y-axis velocity, and Z-axis velocity. The control terminal initializes the covariance matrix of the state vector as an identity matrix. In the processing of each frame, the control terminal first performs a prediction step: calculating the predicted state vector of the current frame based on the state vector and state transition matrix of the previous frame, where the state transition matrix is ​​determined by the sampling interval; calculating the prediction covariance matrix of the current frame based on the covariance matrix of the previous frame and the process noise covariance matrix. Subsequently, the control terminal performs an update step: calculating the observation residual based on the three-dimensional coordinate observation values ​​output by the spatial positioning tag of the current frame, i.e., the difference between the observed value and the coordinate part of the predicted state vector; calculating the Kalman gain matrix based on the prediction covariance matrix, the observation matrix, and the observation noise covariance matrix; multiplying the observation residual by the Kalman gain matrix and adding it to the predicted state vector to obtain the estimated state vector of the current frame; and updating the covariance matrix.The control terminal sequentially performs the above prediction and update steps on the coordinates of each frame after outlier removal. It extracts and arranges the coordinate components from the estimated state vector of each frame in chronological order to obtain a smoothed and filtered three-dimensional spatial coordinate sequence, which is the preprocessed spatial trajectory data. In related technologies, the preprocessing of spatial trajectory data typically only uses simple mean filtering or median filtering to smooth the coordinate values. While mean and median filtering have a certain smoothing effect on slowly changing drift errors, for instantaneous outliers caused by signal reflection, mean filtering will spread the energy of the outlier to multiple adjacent frames, and median filtering can only weaken but not completely eliminate the influence of outliers. In this embodiment, the mean value of the surrounding frames within the window is first calculated using a sliding window as a reference, and the Euclidean distance between the intermediate frame and the reference is calculated to determine outliers. Outliers with instantaneous jumps are accurately identified and eliminated. Then, the remaining coordinate sequence is smoothed by Kalman filtering. In each frame, Kalman filtering uses the estimated state of the previous frame to predict the current frame and then performs weighted correction based on the observed values. This can maintain the continuity and smoothness of the target motion trajectory while eliminating random noise introduced by multipath propagation of the positioning signal.

[0075] Step S234: Align and bind the EEG data, eye movement data, and spatial trajectory data according to a unified timestamp to obtain the multimodal interest data of each acquisition target.

[0076] In this embodiment, alignment binding is an integration operation that performs time-series matching and fusion of three modal data with different sampling frequencies and different data recording start times based on their respective timestamp fields.

[0077] As an optional implementation, the control terminal acquires clean EEG data, clean eye-tracking data, and clean spatial trajectory data in steps S231, S232, and S233, respectively. The sampling rate of the EEG data is 250 Hz, meaning the sampling interval between two adjacent frames is 4 milliseconds; the sampling rate of the eye-tracking data is 60 Hz, meaning the sampling interval between two adjacent frames is approximately 16.67 milliseconds; and the sampling rate of the spatial trajectory data is 10 Hz, meaning the sampling interval between two adjacent frames is 100 milliseconds. The three types of data have different sampling frequencies, and their respective start timestamps for the first frame also differ. The control terminal first uses the timeline of the EEG data as a reference timeline, reads the timestamps of the first and last frames of the EEG data, and determines the start and end times covering the entire acquisition period. The control terminal then filters out all frames in the eye-tracking data whose timestamps fall within this start and end time range, and also filters out all frames in the spatial trajectory data whose timestamps fall within this start and end time range. For each sampling moment in the EEG data, the control terminal checks whether a frame with the exact same timestamp exists in the eye-tracking data. If it does, the eye-tracking data of that frame is directly associated with the EEG data at that moment. If not, the control terminal searches for the closest preceding and following frames in the eye-tracking data and calculates the corresponding eye-tracking data interpolation value using a linear interpolation method. The control terminal performs the same interpolation processing on the spatial trajectory data, mapping each sampling moment of the EEG data to the current EEG signal value, the interpolated eye-tracking data value, and the interpolated spatial trajectory coordinate value, thus forming an integrated record. This record simultaneously contains the four-channel EEG amplitude, the three-dimensional spatial coordinates of the fixation point, and the three-dimensional spatial coordinates of the acquisition target at a single time point. The control terminal traverses all sampling moments of the EEG data, generating an integrated record for each sampling moment, and arranges all integrated records in chronological order to form multimodal interest data for each acquisition target. After alignment and binding, the generated multimodal interest data includes EEG data, eye-tracking data, and spatial trajectory data, all sharing the same timeline reference and corresponding data values ​​at each moment. In related technologies, different modal sensor data are typically stored separately in independent data files after acquisition. During subsequent analysis, coarse time alignment is performed manually or semi-automatically. However, this method suffers from timing drift due to differences in the built-in clocks and sampling frequencies of different devices, resulting in time deviations of several milliseconds to tens of milliseconds between the aligned data across different modalities. This embodiment uses the high-density timeline of the EEG data as a reference, performing linear interpolation to complete the eye-tracking and spatial trajectory data. This achieves frame-level precise alignment of the three modalities on a unified timeline, ensuring that each gaze point and spatial location point in subsequent spatial registration operations matches the EEG response signal at the correct moment.

[0078] This embodiment simultaneously acquires initial EEG data, initial eye-tracking data, and initial spatial trajectory data during the free viewing process of the exhibit. This ensures that the three types of data completely record the neurophysiological response, visual attention orientation, and spatial position at the moment the viewing behavior occurs. Compared to the method of manually stitching together data after separate acquisition in related technologies, the synchronous acquisition mechanism ensures the accurate correspondence of the three types of data in the time dimension. By performing bandpass filtering to remove DC drift and high-frequency noise, notch filtering to eliminate power frequency interference, and independent component analysis to remove EEG and EMG artifacts from the initial EEG data, clean EEG data that retains the true neural rhythm components is obtained. This ensures that the subsequent calculation results of alpha and beta wave power are not contaminated by motion artifacts and power supply environment noise. By removing off-screen and low-confidence data frames from the initial eye-tracking data and correcting the spatial coordinates of the intrinsic parameter matrix, a gaze point coordinate sequence with a clear physical orientation in the three-dimensional spatial coordinate system is obtained. This ensures the spatial accuracy of the geometric determination between the gaze point and the exhibit in subsequent spatial registration. By removing outliers and smoothing the initial spatial trajectory data with Kalman filtering, the multipath effect and signal jump of the positioning system are eliminated, so that the clean spatial trajectory data accurately reflects the actual movement route of the target in the physical space of the booth, thus providing real position change input for subsequent path clustering.

[0079] Based on any of the above embodiments, in Embodiment 4 of this application, the EEG data, eye-tracking data, and spatial trajectory data of each of the acquisition targets are aligned by timestamps and then spatially registered with the spatial location information in the booth spatial layout data to obtain interest association data of each acquisition target at each spatial location unit, including: Step S31: Map the gaze coordinates in the eye movement data of each of the acquired targets from the coordinate system of the eye movement acquisition device to the three-dimensional spatial coordinate system to obtain the three-dimensional gaze space coordinates of the gaze point at each moment in the physical space of the booth.

[0080] In this embodiment, the coordinate system of the eye-tracking acquisition device is a local coordinate system with the optical center of the eye-tracking acquisition device as the origin and the optical axis direction of the device and the horizontal and vertical directions of the image sensor as the coordinate axes.

[0081] As an optional implementation, the control terminal reads eye-tracking data from the multimodal interest data of the currently acquired target. This eye-tracking data contains a sequence of gaze coordinates of the acquired target at each sampling time, with each gaze coordinate located in the coordinate system of the eye-tracking acquisition device. The control terminal reads the gaze coordinates of each frame in the gaze coordinate sequence frame by frame. For the gaze coordinates of the currently read frame, the control terminal synchronously reads the three-dimensional spatial coordinates of the acquired target's spatial trajectory data corresponding to the timestamp of that frame from the multimodal interest data. These three-dimensional spatial coordinates are located in the three-dimensional spatial coordinate system of the booth's physical space. The control terminal loads the factory calibration file of the eye-tracking acquisition device. This calibration file stores the rotation matrix and translation vector from the eye-tracking acquisition device's coordinate system to the three-dimensional spatial coordinate system. This rotation matrix and translation vector are obtained by least-squares fitting after observing multiple known spatial coordinate calibration points on the calibration board using the eye-tracking acquisition device during the booth deployment phase. The control terminal expands the gaze coordinates of that frame from homogeneous coordinates to four-dimensional homogeneous coordinates, that is, adding a dimension with a value of 1 after the X, Y, and Z values ​​of the gaze coordinates. The control terminal multiplies the four-dimensional homogeneous coordinates sequentially by the augmented matrices of the rotation and translation vectors to obtain the mapped coordinates of the gaze point in the three-dimensional spatial coordinate system. The control terminal performs the above mapping operation on the gaze point coordinates at each sampling time to obtain the three-dimensional gaze space coordinate sequence of the gaze point in the physical space of the booth at each time.

[0082] Step S32: Spatial match is performed between the three-dimensional gaze space coordinates at each time point and the spatial location information of each exhibit in the booth spatial layout data to determine the target exhibit identifier corresponding to the gaze point at each time point.

[0083] In this embodiment, the target exhibit identifier is a unique identifier of the exhibit that has a spatial correspondence with the current gaze point. This correspondence is determined by the fact that the three-dimensional gaze space coordinates fall within the spatial area enclosed by the exhibit's spatial location information.

[0084] As an optional implementation, the control terminal reads the exhibit identifiers and spatial location information of each exhibit from the booth space layout data. The spatial location information of each exhibit includes the coordinate sequence of boundary points of its outer contour. The control terminal iterates through the three-dimensional gaze space coordinate sequence obtained in step S31 frame by frame. For the current three-dimensional gaze space coordinates, the control terminal sequentially uses the spatial location information of each exhibit as the current matching object. The control terminal obtains the coordinates of all outer contour boundary points contained in the spatial location information of the current exhibit, and connects these boundary point coordinates sequentially in the three-dimensional spatial coordinate system to form a closed surface. This closed surface defines the entire spatial area occupied by the current exhibit in the physical space of the booth. The control terminal determines whether the current three-dimensional gaze space coordinates are located inside the spatial area enclosed by the closed surface. This determination is achieved by the ray method: the control terminal emits a ray from the three-dimensional gaze space coordinate point in any direction, calculates the number of intersections between the ray and the closed surface, and if the number of intersections is odd, the point is determined to be inside the surface; if the number of intersections is even, the point is determined to be outside the surface. If the 3D gaze space coordinates are determined to be inside the closed surface of the current exhibit, the control terminal identifies the exhibit identifier of the current exhibit as the target exhibit identifier corresponding to the gaze point at that moment, and terminates the traversal of the remaining exhibits. If the 3D gaze space coordinates are determined not to be inside the closed surface of the current exhibit, the control terminal uses the next exhibit as the current matching object and repeats the above judgment until an exhibit containing the 3D gaze space coordinates is found or all exhibits have been traversed. If, after traversing all exhibits, the 3D gaze space coordinates are not contained within the closed surface of any exhibit, the gaze point at that moment is marked as not matching any exhibit.

[0085] Step S33: Based on the spatial trajectory data of the target at each time point and the corresponding three-dimensional gaze space coordinates, calculate the gaze direction vector of the target at each time point, and determine whether the angle between the gaze direction vector and the spatial vector from the target to the target exhibit is less than a preset angle threshold. If it is less than the threshold, then the EEG data, eye movement data and the corresponding target exhibit identifier at that time point are associated and recorded as valid interest association data.

[0086] In this embodiment, the gaze direction vector is the vector corresponding to the directed line segment pointing from the spatial location of the target to the location of the gaze point at that moment. The preset angle threshold is an angular boundary value used to determine whether the gaze direction of the target points to the target exhibit in geometric space. This angular boundary value is determined based on the size of the central region of the human eye's field of vision.

[0087] As an optional implementation, after determining the target exhibit identifier corresponding to the current gaze point in step S32, the control terminal only performs this step for moments where the target exhibit identifier is successfully matched. The control terminal synchronously reads the three-dimensional spatial coordinates of the target's spatial trajectory data from the multimodal interest data at that moment. These coordinates represent the target's actual position in the physical space of the booth. The control terminal reads the corresponding three-dimensional gaze spatial coordinates from the mapping result obtained in step S31. Using the target's three-dimensional spatial coordinates as the starting point and the target's three-dimensional gaze spatial coordinates as the ending point, the control terminal calculates the direction vector from the starting point to the ending point, and uses this direction vector as the gaze direction vector of the target at that moment. The control terminal reads the spatial position information corresponding to the target exhibit identifier from the booth spatial layout data, calculates the average of the coordinates of all outer contour boundary points in the spatial position information, and obtains the center point coordinates of the target exhibit. The control terminal uses the direction vector from the target's three-dimensional spatial coordinates to the center point coordinates of the target exhibit as the spatial vector from the target to the target exhibit. The control terminal calculates the angle between the gaze direction vector and the spatial vector from the target to the target exhibit. The included angle is obtained through the dot product operation of vectors: the gaze direction vector and the spatial vector are normalized to unit vectors, and the dot product of the two unit vectors is calculated. This dot product value is the cosine of the angle between the two vectors. Taking the inverse cosine of this cosine value yields the included angle value. The control terminal compares this included angle value with a preset angle threshold of 30 degrees. If the included angle value is less than 30 degrees, it is determined that the gaze direction at that moment points geometrically to the target exhibit, and the control terminal associates and records the EEG data, eye movement data, and corresponding target exhibit identifier at that moment as a valid interest association data. If the included angle value is greater than or equal to 30 degrees, it is determined that the geometric consistency between the gaze direction at that moment and the target exhibit is insufficient, and the record at that moment is not considered valid interest association data.

[0088] Step S34: Divide the coordinate range of the physical space of the booth into multiple spatial location units, using the spatial location information corresponding to each exhibit as the unit, or according to the preset grid size.

[0089] In this embodiment, the grid size is a fixed length value used when dividing the spatial location units in two directions along the horizontal plane of the physical space coordinate range of the booth.

[0090] As an optional implementation, the control terminal receives a spatial location unit division mode selection command submitted by the user through the input interface. If the command indicates that the division unit is based on exhibits, the control terminal reads the exhibit identifiers of all exhibits and the corresponding spatial location information of each exhibit from the booth space layout data. The spatial area enclosed by the spatial location information of each exhibit is directly taken as a spatial location unit, and the exhibit identifier of that exhibit is bound to this spatial location unit as a unit identifier. Simultaneously, the spatial boundary coordinates of this spatial location unit are recorded. If the command indicates that the division unit is based on a grid, the control terminal reads the minimum and maximum coordinate ranges of the booth physical space in the X-axis direction and the minimum and maximum coordinate ranges in the Y-axis direction. The control terminal obtains a preset grid size, which includes the grid width value in the X-axis direction and the grid depth value in the Y-axis direction, both of which are 0.5 meters. The control terminal delineates the grid boundary lines starting from the minimum value of the X-axis within the coordinate range, proceeding in 0.5-meter increments along the positive X-axis until it crosses the maximum value. Similarly, it delineates the grid boundary lines starting from the minimum value of the Y-axis within the coordinate range, proceeding in 0.5-meter increments along the positive Y-axis until it crosses the maximum value. All grid boundary lines along the X-axis and Y-axis intersect on the horizontal plane to form multiple rectangular grids. Each rectangular grid extends along the Z-axis from the minimum to the maximum value, forming a spatial location unit. The control terminal assigns a unique number to each spatial location unit as its identifier and records its coordinate intervals along the X-axis, Y-axis, and Z-axis, serving as the spatial boundary coordinates for that unit.

[0091] Step S35: Remove the associated data corresponding to the time when the included angle is greater than or equal to the preset angle threshold, and summarize the effective interest-related data of each time according to the dimension of the acquisition target in chronological order to obtain the interest-related data of each acquisition target at each spatial location unit.

[0092] As an optional implementation, during the execution of step S33, for each gaze point at any given time, if the angle calculated in step S33 is greater than or equal to 30 degrees, the control terminal marks the gaze point data at that time as invalid and does not generate a valid interest-related data record. If the angle is less than 30 degrees, the control terminal generates a valid interest-related data record, which includes the timestamp of that time, the target exhibit identifier, EEG data segments, and eye-tracking data segments. After the control terminal has traversed all time points, it obtains a set of interest-related data records for the target at all valid time points. The control terminal reads the multiple spatial location units divided in step S34 and maps the target exhibit identifier or three-dimensional gaze spatial coordinates in each interest-related data record to the corresponding spatial location unit. Specifically, if the spatial location unit is divided by exhibit, the control terminal directly reads the target exhibit identifier from the interest-related data record and assigns the record to a spatial location unit with the same unit identifier as the target exhibit identifier. If the spatial location unit is divided by grid, the control terminal obtains the three-dimensional gaze space coordinates of the interest-related data record at the corresponding time, compares these three-dimensional gaze space coordinates with the spatial region boundary coordinates of each grid spatial location unit, and assigns the record to a grid spatial location unit whose spatial region boundary coordinates include the three-dimensional gaze space coordinates. After performing the above mapping and assignment operation on all interest-related data records of the current acquisition target, the control terminal arranges all records assigned to the same spatial location unit in chronological order, forming a subset of interest-related data for the acquisition target under that spatial location unit. The control terminal summarizes the subsets of interest-related data for the acquisition target under all spatial location units to obtain the interest-related data for the acquisition target at each spatial location unit. The control terminal repeats steps S31 to S35 for each acquisition target until the interest-related data for all acquisition targets has been summarized.

[0093] This embodiment maps the gaze coordinates from the eye-tracking acquisition device coordinate system to a three-dimensional spatial coordinate system, accurately defining the spatial position of the gaze point under a unified geometric reference in the physical space of the exhibition booth, eliminating spatial misalignment caused by differences in coordinate systems of different devices. By matching the three-dimensional gaze spatial coordinates with the spatial position information of each exhibit to determine the target exhibit identifier, each gaze point can be assigned to a specific exhibit, avoiding the defect of gaze points being suspended and unable to be located. At the same time, using the outer contour boundary of the exhibit as the basis for spatial assignment preserves the actual geometry of the exhibit, unlike the method of simplifying it to a single coordinate point, thus reducing the misjudgment rate in gaze point assignment determination. By comparing the angle between the gaze direction vector and the spatial vector from the acquisition target to the target exhibit with a preset angle threshold, it distinguishes between the acquisition target actively gazing at the target exhibit and the case where the gaze direction merely passes through the direction of the target exhibit, effectively eliminating invalid gaze data generated by the gaze sweeping over exhibits during head rotation, ensuring that the gaze records in the effective interest association data all correspond to the acquisition target's active gaze at the exhibit. By dividing spatial locations into units based on exhibit location information or grid size, a spatial analysis granularity adapted to different booth layout characteristics is provided. In scenarios with sparse exhibits, using exhibits as the unit of division reduces redundant computation, while in scenarios with dense exhibits, fine-grained grid division enhances spatial resolution. By removing data with abnormal angles and summarizing interest-related data chronologically according to the dimension of the collection target, a complete association record containing EEG, eye-tracking data, and exhibit identifiers is obtained for each collection target at each spatial location unit. This ensures that subsequent calculations of group interest metrics are based solely on physiological data corresponding to truly valid gaze behaviors in spatial geometry.

[0094] Based on any of the above embodiments, in Embodiment 5 of this application, referring to Figure 2 Based on the interest association data of each collected target at each spatial location unit, a group interest metric value is determined at each spatial location unit, including: Step S41: For each spatial location unit, acquire all interest association data of all acquisition targets within the spatial location unit.

[0095] As an optional implementation, the control terminal reads all data records that have completed spatial classification from the interest association data of each acquisition target at each spatial location unit obtained in step S35. The control terminal traverses all spatial location units and takes the currently traversed spatial location unit as the target unit. The control terminal retrieves all data records whose unit identifier or spatial region boundary matches the target unit from the interest association data corresponding to each acquisition target, and aggregates all retrieved data records into a data set for the target unit. Each record in this data set contains a fragment of EEG data, a fragment of eye movement data, and the corresponding exhibit identifier or spatial location coordinates of an acquisition target at a certain moment, and the three-dimensional gaze spatial coordinates at that moment are located within the spatial region boundary of the target unit. The control terminal performs the above aggregation operation for each spatial location unit until each spatial location unit obtains the corresponding complete interest association data set.

[0096] Step S42: Perform group statistics on the EEG data of each of the collected targets in the interest-related data within the spatial location unit, and calculate the group mean of β wave power and α wave power within the spatial location unit.

[0097] As an optional implementation, the control terminal uses the interest-related data set of the current spatial location unit obtained in step S41 as the processing object. Each valid interest-related data record in this set contains an EEG data segment, which is a clean EEG signal at the corresponding time of the record, stored in the form of multi-channel time-series voltage values. The control terminal reads the EEG data segments contained in all records in the interest-related data set. Each EEG data segment contains a continuous EEG signal acquired at a sampling rate of 250 Hz, and the duration corresponds to the duration of a single gaze event of the acquisition target. The control terminal performs the following operations on each EEG data segment: performs a Fast Fourier Transform on the time-series signal in the segment to transform the signal from the time domain to the frequency domain, obtaining the power distribution spectrum of the signal on the frequency axis. The control terminal extracts the power values ​​of all frequency points in the 8 Hz to 13 Hz frequency band from the power distribution spectrum and sums them to obtain the alpha wave power value of the segment. The control terminal extracts and sums the power values ​​of all frequency points within the 13 Hz to 30 Hz band in the power distribution spectrum to obtain the β-wave power value of that segment. The control terminal performs the above extraction of α-wave and β-wave power values ​​on all records in the interest-related data set, obtaining a set of α-wave power value sequences and a set of β-wave power value sequences. The control terminal groups these sequences according to the acquisition target dimension, grouping the power values ​​generated by gaze records of the same acquisition target into the same group. The average α-wave power value of the same acquisition target across all gaze records is used to obtain the individual α-wave power mean of that acquisition target in the current spatial location unit, and the average β-wave power value of the same acquisition target across all gaze records is used to obtain the individual β-wave power mean of that acquisition target in the current spatial location unit. After obtaining the individual α-wave power mean and individual β-wave power mean of all acquisition targets in the current spatial location unit, the control terminal sums the individual α-wave power mean of all acquisition targets and divides by the number of acquisition targets generating gazes to obtain the group α-wave power mean of the current spatial location unit. The control terminal sums the individual beta-wave power averages of all acquired targets and divides the sum by the number of acquired targets that are fixated, to obtain the population beta-wave power average of the current spatial location unit. The control terminal performs the above processing sequentially for each spatial location unit to obtain the population beta-wave power average and the population alpha-wave power average for each spatial location unit.

[0098] Step S43: Perform group statistics on the eye movement data of each acquisition target in the interest association data within the spatial location unit, and calculate the cumulative number of fixation points, the cumulative value of fixation duration, and the number of acquisition targets that generate fixation within the spatial location unit.

[0099] As an optional implementation, the control terminal uses the interest-related data set of the current spatial location unit obtained in step S41 as the processing object. Each valid interest-related data record in this set contains eye-tracking data, and each record corresponds to a gaze event of the acquisition target. The control terminal counts the total number of records in the interest-related data set and uses this total number as the cumulative number of gaze points for the current spatial location unit. This cumulative number of gaze points represents the sum of the number of valid gazes generated by all acquisition targets at the current spatial location unit. The control terminal reads the gaze duration value in each record of the interest-related data set. This gaze duration value is the duration of the gaze event extracted from the initial eye-tracking data during the eye-tracking data preprocessing stage, in milliseconds. The control terminal adds up the gaze duration values ​​of all records in the interest-related data set to obtain the cumulative gaze duration value of the current spatial location unit. The control terminal counts the number of different target identifiers in the interest-related data set. Specifically, it iterates through all records in the set, reads the target identifier carried in each record, adds identifiers that have already appeared to the target set, and ignores duplicate identifiers. The final number of elements in the target set represents the number of people exhibiting gaze at the target in the current spatial location unit. The control terminal performs the above statistical operation sequentially for each spatial location unit to obtain the cumulative number of gaze points, cumulative gaze duration, and the number of people exhibiting gaze at the target for each spatial location unit.

[0100] Step S44: Calculate the group interest metric of the spatial location unit based on the group mean of β-wave power, group mean of α-wave power, cumulative number of fixation points, cumulative fixation duration, and number of acquisition targets generating fixation.

[0101] As an optional implementation, the control terminal uses the average beta-wave power group and the average alpha-wave power group obtained in step S42, as well as the cumulative number of fixation points, cumulative fixation duration, and number of acquisition targets generating fixation obtained in step S43, as input parameters. The control terminal loads a preset set of normalized parameters, which includes the beta-wave baseline value corresponding to the average beta-wave power group, the alpha-wave baseline value corresponding to the average alpha-wave power group, the baseline value for the number of fixations corresponding to the cumulative number of fixation points, the baseline value for the fixation duration corresponding to the cumulative fixation duration, and the baseline value for the number of acquisition targets generating fixation. Each baseline value is obtained by statistically analyzing the corresponding parameter in all spatial location units within the booth's physical space. Specifically, the maximum and minimum values ​​of the parameter are extracted from all spatial location units, and the difference between the maximum and minimum values ​​is used as the baseline value for the parameter, or a fixed value is preset based on historical acquisition data. The control terminal performs a normalization operation on the average beta-wave power group: calculating the ratio of the average beta-wave power group to the beta-wave baseline value to obtain the normalized beta-wave engagement level. The control terminal performs a normalization operation on the average alpha wave power: calculating the ratio of the alpha wave baseline value to the average alpha wave power to obtain the normalized alpha wave relaxation. The control terminal performs a normalization operation on the cumulative number of fixations: calculating the ratio of the cumulative number of fixations to the baseline number of fixations to obtain the normalized number of fixation sessions. The control terminal performs a normalization operation on the cumulative fixation duration: calculating the ratio of the cumulative fixation duration to the baseline fixation duration to obtain the normalized fixation duration. The control terminal performs a normalization operation on the number of target subjects generating fixations: calculating the ratio of the number of target subjects generating fixations to the baseline number of subjects to obtain the normalized number of subjects covered. The control terminal loads a preset set of weight parameters, which includes a first weight corresponding to the normalized beta wave engagement, a second weight corresponding to the normalized alpha wave relaxation, a third weight corresponding to the normalized number of fixation sessions, a fourth weight corresponding to the normalized fixation duration, and a fifth weight corresponding to the normalized number of subjects covered. The values ​​of each weight are preset by the user based on the booth optimization goals. Beta-wave engagement reflects the cognitive engagement level of the target at the current spatial location unit; alpha-wave relaxation reflects the attentional relaxation state of the target at the current spatial location unit; the number of times the target is focused and the fixation duration reflect the visual attention intensity of the target at the current spatial location unit; and the number of people covered reflects the overall attractiveness range of the current spatial location unit to the group. The control terminal calculates the group interest metric of the current spatial location unit according to the following formula: multiply the normalized beta-wave engagement by the first weight to obtain the first weighted value; multiply the normalized alpha-wave relaxation by the second weight to obtain the second weighted value; multiply the normalized number of times the target is focused by the third weight to obtain the third weighted value; multiply the normalized fixation duration by the fourth weight to obtain the fourth weighted value; multiply the normalized number of people covered by the fifth weight to obtain the fifth weighted value; and add the first, second, third, fourth, and fifth weighted values ​​to obtain the group interest metric of the current spatial location unit.The magnitude of this group interest metric is positively correlated with the overall interest attraction intensity of the target group to the current spatial location unit. The control terminal repeats the above normalized weighted calculation for each spatial location unit to obtain the group interest metric value corresponding to each spatial location unit.

[0102] For example, in the spatial location unit division of a pop-up store booth, spatial location unit A, where the makeup trial area is located, gathers valid interest-related data from thirty target individuals. This unit's interest-related data set contains 240 valid gaze records, which the control terminal uses as the cumulative gaze point count. The control terminal extracts EEG data segments from each record and performs Fast Fourier Transform (FFT) on each, obtaining 240 alpha wave power values ​​and 240 beta wave power values. After grouping and averaging by target dimension, the average alpha wave power and average beta wave power of the thirty target individuals are obtained. The control terminal then averages the average alpha wave power of all individuals to obtain a group average alpha wave power of 8.2 microvolts squared for this unit, and averages the average beta wave power of all individuals to obtain a group average beta wave power of 6.7 microvolts squared for this unit. The control terminal reads the gaze duration values ​​of 240 records within this unit. The gaze duration of each record ranges from 300 milliseconds to 1200 milliseconds, and after summing them, the cumulative gaze duration of this unit is 156,000 milliseconds. The control terminal counted 30 people as the target for gaze acquisition within this unit. The control terminal loaded a preset set of normalized parameters, where the baseline value for beta waves was 10.0 μV², the baseline value for alpha waves was 12.0 μV², the baseline value for the number of gazes was 500, the baseline value for gaze duration was 300,000 milliseconds, and the baseline value for the number of people was 50. The control terminal calculated the normalized beta wave engagement as 0.67, the normalized alpha wave relaxation as 1.46, the normalized number of gazes as 0.48, the normalized gaze duration as 0.52, and the normalized number of people covered as 0.60. The control terminal loaded a preset set of weight parameters, where the first weight was 0.30, the second weight was 0.05, the third weight was 0.25, the fourth weight was 0.25, and the fifth weight was 0.15. The control terminal multiplied each normalized value by its corresponding weight and summed the results to obtain a group interest metric of 0.671 for spatial location unit A. After the control terminal performs the same parameter statistics and weighted calculations on all spatial units within the booth, it obtains the group interest metric value corresponding to each spatial unit. The group interest metric value of the unit where the makeup trial area is located is the highest, the group interest metric value of the unit where the product display area is located is at a medium level, and the group interest metric value of the unit where the information desk in the corner of the booth is located is the lowest.

[0103] This embodiment aggregates interest-related data from all acquisition targets for each spatial location unit. It then calculates five statistics on the aggregated data: the population mean of beta-wave power, the population mean of alpha-wave power, the cumulative number of fixations, the cumulative fixation duration, and the number of acquisition targets generating fixations. These five statistics are then normalized and weighted to transform into a single population interest metric. This integrates individual-level interest-related data scattered across various acquisition targets and time points into a population-level quantitative indicator at the spatial location unit level. The population mean of beta-wave power and the population mean of alpha-wave power characterize the overall cognitive engagement and attentional relaxation state of all acquisition targets at the current spatial location from a neurophysiological perspective. The cumulative number of fixations, the cumulative fixation duration, and the number of acquisition targets generating fixations characterize the total frequency of fixation, sustained attention depth, and breadth of acquisition targets covered at the current spatial location from a visual behavior perspective. By merging the population statistics of two objective dimensions—electrophysiological response and visual attention behavior—into a single quantitative indicator, this population interest metric can simultaneously reflect both the objective aspects of whether a spatial location unit "attracted visual attention" and "stimulated cognitive engagement." By normalizing, five statistical quantities with different dimensions are converted into comparable values ​​on the same numerical scale, enabling comparisons between different spatial units based on the same benchmark. By weighted summation, the five statistical quantities are merged into a single value according to their respective contribution weights. This single value is used as the final output form of the group interest metric. This allows subsequent layout optimization steps to eliminate the need to process the five independent statistical quantities separately, and to directly sort and compare based on the single value to determine the priority of exhibits and make spatial location adjustment decisions.

[0104] Based on any of the above embodiments, in Embodiment Six of this application, a layout optimization scheme for the physical space of the booth is generated according to the group interest metric value at each spatial location unit and the booth space layout data, including: Step S71: Sort the group interest metric values ​​at the spatial location units corresponding to each exhibit in descending order to generate exhibit interest ranking results.

[0105] In this embodiment, the exhibit interest ranking result is a sequence formed by arranging all exhibits in the physical space of the booth from high to low according to their respective group interest metric values.

[0106] As an optional implementation, the control terminal reads the exhibit identifiers of all exhibits from the booth space layout data and obtains the group interest metric values ​​corresponding to each spatial location unit calculated in step S40. For the currently traversed exhibit, the control terminal determines all the spatial location units it covers based on the exhibit's spatial location information, reads the group interest metric values ​​corresponding to each of these spatial location units from the calculation results of step S40, and performs an arithmetic average of the group interest metric values ​​of all spatial location units belonging to the same exhibit to obtain the comprehensive interest metric value of the exhibit. The control terminal repeats the above operation for each exhibit, and after obtaining the comprehensive interest metric value corresponding to each exhibit, it arranges all exhibits in descending order of comprehensive interest metric values ​​to generate an exhibit interest ranking result. Each exhibit entry in this ranking result contains the exhibit identifier and its corresponding comprehensive interest metric value.

[0107] Step S72: Based on the exhibit interest ranking results, determine the recommended layout priority of each exhibit, mark the exhibits with the top N group interest metric values ​​as high priority exhibits, and adjust the high priority exhibits within the coordinate range of the physical space of the booth to an area where the distance between them and the booth entrance is less than a preset distance threshold.

[0108] As an optional implementation, the control terminal reads the exhibit interest ranking results and obtains a preset priority quantity parameter N. This priority quantity parameter N is pre-configured by the user based on the total number of exhibits in the booth and the booth area, and is used to specify the number of exhibits marked as high priority. The control terminal extracts exhibits ranked from 1st to Nth from the exhibit interest ranking results and marks these exhibits as high priority exhibits. The control terminal reads the coordinates of the booth entrance center point in the three-dimensional spatial coordinate system from the booth spatial layout data. For each exhibit marked as a high priority exhibit, the control terminal obtains the current spatial location information of the exhibit and calculates its center point coordinates, and calculates the spatial Euclidean distance between the center point coordinates and the booth entrance center point coordinates. The control terminal obtains a preset distance threshold, which is one-third of the total length of the booth physical space in the X-axis direction. If the spatial Euclidean distance between the current center point of the exhibit and the entrance is greater than the preset distance threshold, the control terminal searches for the target spatial location of the exhibit within the coordinate range of the booth physical space, within a spatial area centered on the booth entrance center point coordinates and with the preset distance threshold as the radius. During the search process, the control terminal ensures that the target spatial location of the exhibit does not exceed the coordinate range of the physical space of the booth, and that the spatial Euclidean distance between it and the center point of the booth entrance is less than a preset distance threshold. After determining the target spatial location, the control terminal translates the coordinates of the outer contour boundary points in the exhibit's spatial location information to the target spatial location, thus updating the center point coordinates of the exhibit to match the center point coordinates of the target spatial location. If the spatial Euclidean distance between the current center point of the exhibit and the entrance is less than or equal to the preset distance threshold, the control terminal maintains the current spatial location of the exhibit unchanged.

[0109] Step S73: Obtain the preset functional association attributes between each exhibit, mark exhibits with the same functional association attributes as an associated exhibit group, and adjust each exhibit in the associated exhibit group to an adjacent area where the spatial distance between them is less than a preset spacing threshold.

[0110] In this embodiment, the preset functional association attribute is a functional category label that the booth designer pre-marks for each exhibit, which is used to characterize the business function type to which the exhibit belongs.

[0111] As an optional implementation, the control terminal reads the preset functional association attributes corresponding to each exhibit from the booth space layout data. These preset functional association attributes are stored in the exhibit data records in the form of functional category tags. The control terminal groups all exhibits according to their functional category tags, with exhibits having the same functional category tag grouped into the same exhibit set. Each exhibit set constitutes an associated exhibit group. The control terminal sequentially traverses each associated exhibit group. For the currently processed associated exhibit group, the control terminal obtains the current spatial position information of all exhibits within that group. The control terminal reads a preset spacing threshold, which is one-tenth of the total length of the booth's physical space along the X-axis. The control terminal determines whether the spatial distance between any two exhibits within the group is less than the preset spacing threshold. If the spatial distance between any two exhibits is greater than or equal to the preset spacing threshold, the control terminal, within the coordinate range of the physical space of the booth, uses the current spatial position of the exhibit with the highest comprehensive interest metric value in the group as the reference position, and moves the remaining exhibits in the group one by one towards the reference position. The spatial position information of each exhibit after movement must meet the following requirements: the Euclidean distance between any two exhibits in the group is less than the preset spacing threshold, the adjusted spatial position of each exhibit does not exceed the coordinate range of the physical space of the booth, and the spatial position information of different exhibits does not overlap. During the adjustment of exhibits in the associated exhibit group, if the available space around the reference position is insufficient to accommodate all exhibits in the group, the control terminal expands outward from the reference position in descending order of the comprehensive interest metric value of the exhibits in the group. The expansion distance is limited to not exceeding the preset spacing threshold, ensuring that all exhibits in the group are within the coordinate range while satisfying the spacing constraints. The control terminal performs the above adjustment operation sequentially for each associated exhibit group until all exhibits in the associated exhibit groups are adjusted to positions that satisfy the spacing constraints.

[0112] Step S74: Update the booth space layout data according to the adjusted spatial location information of each exhibit, and generate a layout optimization scheme containing the spatial location information of each exhibit.

[0113] In this embodiment, the layout optimization scheme is a digital scheme file containing the adjusted target spatial location information of each exhibit within the physical space of the booth.

[0114] As an optional implementation, after completing all spatial position adjustment operations in steps S72 and S73, the control terminal obtains the adjusted spatial position information for each exhibit. This spatial position information includes the adjusted coordinate sequence of the outer contour boundary points of the exhibit in the three-dimensional spatial coordinate system. The control terminal opens the booth spatial layout data file generated in step S13 and iterates through the record entries corresponding to each exhibit in the data file one by one. For the currently iterated exhibit record entry, the control terminal replaces the outer contour boundary point coordinate sequence in the original spatial position information field of the entry with the adjusted outer contour boundary point coordinate sequence of the exhibit, while keeping the exhibit identifier and other attribute fields of the entry unchanged. The control terminal performs the above replacement operation on each exhibit record entry in sequence until the spatial position information of all exhibits is updated to the adjusted target spatial position information. The control terminal saves the booth spatial layout data file after completing all replacement operations as a new file, which is the layout optimization scheme containing the target spatial position information of each exhibit. In this layout optimization scheme, the coordinate range of the physical space of the booth remains unchanged, the exhibit identifiers remain unchanged, and the coordinates of the outer contour boundary points in the spatial location information of each exhibit have all been updated to the adjusted target values.

[0115] This embodiment sorts exhibits by group interest metrics and places high-priority exhibits near the entrance, making layout decisions based on objective interest response data of the visitor group rather than empirical judgment, thus reducing the physical walking distance required for visitors to reach high-interest exhibits. By aggregating functionally related exhibits into adjacent areas, the path length between related exhibits is shortened, reducing the unnecessary walking distance caused by visitors searching for related exhibits across areas. The booth space layout data is updated based on the adjusted spatial location information, and a layout optimization plan is generated. The optimization results are solidified into executable spatial coordinate data, allowing booth builders to directly complete exhibit placement according to the plan. The above methods work together to overcome the technical deficiency of existing technologies that only select sites based on the maximum channel traffic, failing to reflect the actual distribution of visitor interests.

[0116] Based on any of the above embodiments, in Embodiment 7 of this application, generating a layout optimization scheme for the physical space of the booth based on the group interest metric value at each spatial location unit and the booth space layout data further includes: Step S81: Obtain user profile tags for each of the collected targets. The user profile tags include at least one of gender tags, age group tags, and interest preference tags.

[0117] In this embodiment, user profile tags are classification tag data used to describe the demographic attributes and interests of the target population.

[0118] As an optional implementation, before the target enters the physical space of the booth, on-site operators display an electronic questionnaire interface to the target via a tablet terminal. This electronic questionnaire includes gender selection, age group selection, and interest preference selection. The gender selection includes two options: male and female. The age group selection includes four options: 18-25 years old, 26-35 years old, 36-45 years old, and 46 years and older. The interest preference selection includes five options: technology and digital products, beauty and fashion, automobiles and machinery, parenting, and comprehensive experiences. After the target completes the selection on the electronic questionnaire, the tablet terminal associates and stores the target's answer with the target identifier bound to their worn sensing device, generating a user profile tag for that target. The control terminal receives and stores the user profile tags for each target from the tablet terminal.

[0119] Step S82: Based on the user profile tags of each of the collected targets, all collected targets are divided into multiple subgroups, and the collected targets in the same subgroup have the same user profile tags or the same combination of user profile tags.

[0120] As an optional implementation, the control terminal reads the user profile tags of all data collection targets and obtains a preset grouping strategy. This grouping strategy is pre-configured by the user according to analysis needs and includes two types: single-tag grouping mode and multi-tag combination grouping mode. In single-tag grouping mode, the control terminal selects a specified tag dimension from the gender, age group, and interest preference tags included in the user profile tags, and divides all data collection targets into multiple subgroups according to different values ​​of that tag dimension. Data collection targets within the same subgroup have the same value on that tag dimension. For example, the control terminal selects the gender tag as the grouping dimension, dividing all data collection targets into male and female subgroups. In multi-tag combination grouping mode, the control terminal selects at least two specified tag dimensions from the three tag dimensions included in the user profile tags, and divides all data collection targets into multiple subgroups according to the combinations of values ​​of these tag dimensions. Data collection targets within the same subgroup have the same value on each selected tag dimension. For example, the control terminal selects gender and age group labels as grouping dimensions, dividing all data collection targets into subgroups such as males aged 18-25, females aged 18-25, males aged 26-35, and females aged 26-35. The control terminal performs the division operation according to the user-specified grouping strategy, generating multiple subgroups. Each subgroup contains one or more data collection targets, and each data collection target belongs to only one subgroup.

[0121] Step S83: For each subgroup, calculate the subgroup interest metric value at each spatial location unit from the interest association data of each target collected within the subgroup at each spatial location unit. The calculation method of the subgroup interest metric value is the same as the calculation method of the group interest metric value.

[0122] As an optional implementation, the control terminal sequentially selects the currently processed subgroup from the multiple subgroups divided in step S82. The control terminal obtains a list of target identifiers for all targets included in the subgroup. From the interest association data of each target at each spatial location unit obtained in step S35, the control terminal filters out all interest association data whose target identifiers fall into the identifier list of the current subgroup, forming a subset of interest association data for that subgroup. The control terminal performs the same statistical and calculation operations as steps S41 to S44 on a spatial location unit basis on the subset of interest association data for that subgroup. Specifically, for the currently processed spatial location unit, the control terminal filters out all records belonging to that spatial location unit from the interest-related data subset of that subgroup. It then performs group statistics on the EEG data of each acquisition target within that subgroup in that spatial location unit, calculating the group mean of beta wave power and alpha wave power. Finally, it performs group statistics on the eye movement data of each acquisition target within that subgroup in that spatial location unit, calculating the cumulative number of fixations, the cumulative fixation duration, and the number of acquisition targets exhibiting fixation. These five statistics are then normalized and weighted to obtain the subgroup interest metric value for that subgroup at the current spatial location unit. The control terminal performs the above calculations for each subgroup across all spatial location units to obtain the subgroup interest metric value for each subgroup at each spatial location unit.

[0123] Step S84: Based on the dwell time of the spatial trajectory data in each spatial location unit, mark the spatial location unit whose dwell time exceeds the preset dwell threshold as a dwell point.

[0124] In this embodiment, the stopping point is a spatial location unit where the target group stays for more than a preset threshold during the exhibition, indicating that the location has the characteristic of attracting the target group to stop and pay attention.

[0125] As an optional implementation, the control terminal reads the spatial trajectory data of all acquisition targets at each time moment. This spatial trajectory data includes the three-dimensional spatial coordinates of each acquisition target at each sampling time moment. Based on the spatial region boundary coordinates of the spatial location units divided in step S34, the control terminal maps the spatial trajectory coordinates of each acquisition target at each time moment to the corresponding spatial location unit. For each acquisition target, the control terminal traverses the spatial trajectory coordinate sequence of the acquisition target in chronological order, recording the time when the acquisition target enters each spatial location unit and the time when it leaves the spatial location unit. The exit time is subtracted from the entry time to obtain the single dwell time of the acquisition target within that spatial location unit. If the same acquisition target has multiple entry and exit records within the same spatial location unit, the control terminal accumulates the dwell times of each entry and exit to obtain the cumulative dwell time of the acquisition target within that spatial location unit. The control terminal performs the above mapping and dwell time calculation on all acquisition targets, summing the cumulative dwell times of all acquisition targets within each spatial location unit to obtain the group cumulative dwell time of that spatial location unit. The control terminal obtains a preset dwell threshold. If the cumulative dwell time of the current spatial location unit is greater than or equal to the preset dwelling threshold, the control terminal marks the spatial location unit as a dwelling point. The control terminal performs the above judgment on all spatial location units in sequence, and stores the spatial location units marked as dwelling points into the dwelling point set.

[0126] Step S85: Extract the spatial location unit sequence that each acquisition target has passed through in time sequence from the spatial trajectory data of each acquisition target, perform sequence similarity clustering on all spatial location unit sequences of acquisition targets, and determine the sequence cluster containing the most acquisition targets as the high-frequency path.

[0127] In this embodiment, the spatial location unit sequence is the arrangement sequence of spatial location units that a single acquisition target enters sequentially during the exhibition process. The high-frequency path is the spatial location unit access order pattern commonly followed by the most numerous acquisition targets among all acquisition targets.

[0128] As an optional implementation, the control terminal, for each acquisition target, traverses the spatial trajectory coordinate sequence of the target in chronological order, mapping the spatial trajectory coordinates at each moment to the corresponding spatial location unit. If multiple consecutive moments are mapped to the same spatial location unit, only the moment of the first entry is retained as an access record. The spatial location unit access sequence of the acquisition target is formed by the chronological order of the access records. After obtaining the spatial location unit access sequences corresponding to all acquisition targets, the control terminal randomly selects one sequence from all sequences as the initial cluster center. For each sequence not selected as a cluster center, the sequence similarity between the sequence and all current cluster centers is calculated. The sequence similarity is calculated based on the longest common subsequence algorithm, that is, the length of the longest common subsequence that maintains the same relative order between two sequences is calculated, and this length is divided by the maximum value of the lengths of the two sequences to obtain the normalized similarity value. The control terminal assigns the current sequence to the sequence cluster to which it belongs. If the similarity value of the current sequence with all cluster centers is lower than the preset similarity threshold of 0.5, then the sequence is used as a new cluster center and a new sequence cluster is created. After the control terminal completes one round of merging for all sequences, it recalculates the mean sequence of all sequences in each sequence cluster as the new cluster center for that cluster, and repeats the merging operation until the cluster centers no longer change or the number of iterations reaches a preset limit. After clustering is completed, the control terminal counts the number of collection targets contained in each sequence cluster, identifies the sequence cluster containing the most collection targets as the high-frequency path cluster, and determines the cluster center sequence of the high-frequency path cluster as the high-frequency path.

[0129] Step S86: Based on the subgroup interest metric value of each subgroup at each spatial location unit, as well as the stopping point and the high-frequency path, generate a movement guidance scheme matching each subgroup. The movement guidance scheme includes: a sequence of recommended path exhibits arranged in spatial coordinate order, and spatial location information of the placement of guide signs between adjacent recommended path exhibits.

[0130] In this embodiment, the circulation guidance scheme is a digital scheme file generated for a specific subgroup, which includes recommended viewing routes and the locations of guide signs along the routes.

[0131] As an optional implementation, the control terminal sequentially selects the currently processed subgroup from the multiple subgroups divided in step S82. The control terminal obtains the subgroup interest metric value of the subgroup at each spatial location unit. The control terminal sorts the subgroup interest metric values ​​of the current subgroup at all spatial location units from high to low, and selects the top M spatial location units as candidate interest units of the subgroup, where M is the user-preset recommended exhibit quantity parameter. The control terminal obtains the set of stopping points marked in step S84 and the high-frequency path determined in step S85. Using the high-frequency path as the basic path skeleton, the control terminal replaces each spatial location unit in the sequence of spatial location units traversed by the high-frequency path with the candidate interest unit that is spatially closest to that spatial location unit. If the replaced candidate interest unit belongs to the same exhibit area in the physical space of the booth as the preceding and following units in the high-frequency path, the replacement result is directly retained. If they do not belong to the same exhibit area, the position is replaced with the candidate interest unit with the smallest sum of spatial distances to the preceding and following units. After obtaining the replaced spatial location unit sequence, the control terminal extracts the exhibit identifiers corresponding to each spatial location unit in the sequence sequentially, resulting in a recommended route exhibit sequence arranged in spatial coordinate order. For each pair of adjacent exhibits in the recommended route exhibit sequence, the control terminal obtains the coordinates of the last boundary point appearing along the viewing direction from the spatial location information of the preceding exhibit and the coordinates of the first boundary point appearing along the viewing direction from the spatial location information of the following exhibit. The midpoint of the line connecting these two boundary point coordinates is determined as the spatial location for the placement of the guide sign. The control terminal associates and stores the recommended route exhibit sequence and the spatial location information of all guide signs, generating a movement guidance scheme for the current subgroup. The control terminal repeats the above operation for each subgroup to obtain a movement guidance scheme matching each subgroup.

[0132] For example, during the generation of a path guidance scheme for a certain exhibit signage location, the control terminal acquires user profile tags for all target users. Users are pre-selected to group by gender. The control terminal divides all thirty target users into male and female subgroups, with the male subgroup containing sixteen target users and the female subgroup containing fourteen. The control terminal calculates the subgroup interest metric for each subgroup at each spatial location unit. Based on the spatial trajectory data of all target users, the control terminal identifies spatial location unit A as a stopping point. This unit corresponds to the intelligent cockpit simulator exhibit, and the group's cumulative dwell time is twenty-eight minutes, exceeding the preset dwell time threshold of twenty minutes. The control terminal clusters the spatial location unit access sequences of all thirty target users. The sequence cluster containing the most target users contains eleven target users, and the access order of this sequence cluster is spatial location unit C, spatial location unit A, spatial location unit E, and spatial location unit B. The control terminal determines this order as the high-frequency path. The control terminal generates a circulation guidance scheme for the male subgroup: the spatial units with the highest interest metric for the male subgroup are the engine display stand and the intelligent cockpit simulator. Using high-frequency paths as the framework, the control terminal generates a recommended sequence of exhibits: engine display stand, intelligent cockpit simulator, brand history wall. The control terminal places directional signs along the line connecting the engine display stand and the intelligent cockpit simulator in the middle of the two exhibits, and along the line connecting the intelligent cockpit simulator and the brand history wall, places directional signs at the corners of the passageway. The control terminal also generates a circulation guidance scheme for the female subgroup: the spatial units with the highest interest metric for the female subgroup are the interior experience cabin and the children's interactive area. The control terminal generates a recommended sequence of exhibits: interior experience cabin, children's interactive area, intelligent cockpit simulator. The control terminal places directional signs along the line connecting the interior experience cabin and the children's interactive area in the middle, slightly inside the two exhibits, and along the line connecting the children's interactive area and the intelligent cockpit simulator, places directional signs at the midpoint of the straight section of the passageway.

[0133] This embodiment acquires user profile tags of the target audience and divides them into multiple subgroups. This allows the subsequent path guidance scheme to differentiate between target groups with different demographic attributes, eliminating the problem of insufficient group adaptability caused by using the same guidance path for all targets with a uniform path. By independently calculating the subgroup's interest metric, the path guidance scheme is based on the subgroup's own interest response data rather than the average response data of all targets, preventing the interests of a high-proportion group from overshadowing the differentiated needs of a low-proportion group. By marking stopping points based on the cumulative dwell time of the group in each spatial location unit according to spatial trajectory data, objective dwell time data serves as the basis for determining the actual dwelling behavior of the target group at each spatial location unit. This ensures that the identification of stopping points is based on the spatial movement behavior of the target audience rather than the subjective preset of the exhibit designer. By performing sequence similarity clustering on the spatial location unit sequences of each collection target and identifying the sequence cluster containing the most collection targets as the high-frequency path, and using the actual path sequence traversed by all collection targets as the basis for determining the high-frequency path, the high-frequency path truly reflects the natural viewing order of most collection targets, rather than an idealized circulation route subjectively conceived by the booth designer. By combining the subgroup interest metric, stopping points, and high-frequency paths to generate a circulation guidance scheme matching each subgroup, the recommended route exhibit sequence retains the natural traversal order of most collection targets while incorporating exhibits of interest to each subgroup into the path. Furthermore, the spatial location information of the guidance signage is calculated based on the spatial boundary coordinates of adjacent recommended exhibits, ensuring that the guidance signage can be placed at the actual physical locations where visitors need to turn or choose paths.

[0134] This application provides a business space layout optimization device based on neural data. The business space layout optimization device based on neural data includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the business space layout optimization method based on neural data in the first embodiment described above.

[0135] The following is for reference. Figure 3 The diagram illustrates a structural schematic suitable for implementing the neural data-based commercial space layout optimization device in the embodiments of this application. The neural data-based commercial space layout optimization device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, personal digital assistants (PDAs), tablets, and in-vehicle terminals, as well as fixed terminals such as digital TVs and desktop computers. Figure 3The commercial spatial layout optimization device based on neural data shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0136] like Figure 3 As shown, a neural data-based commercial space layout optimization device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the neural data-based commercial space layout optimization device. The processing unit 1001, the ROM 1002, and the RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the neural data-based commercial space layout optimization device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a neural data-based commercial space layout optimization device with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0137] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0138] The neural data-based commercial space layout optimization device provided in this application, employing the neural data-based commercial space layout optimization method described in the above embodiments, can solve the technical problem that the optimization scheme for booth space layout based on the coordinate position with the largest physical channel traffic leads to poor optimization results. Compared with the prior art, the beneficial effects of the neural data-based commercial space layout optimization device provided in this application are the same as those of the neural data-based commercial space layout optimization device provided in the above embodiments, and other technical features in this neural data-based commercial space layout optimization device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0139] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0140] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0141] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the business space layout optimization method based on neural data in the above embodiments.

[0142] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, radio frequency (RF), or any suitable combination thereof.

[0143] The aforementioned computer-readable storage medium may be included in a neural data-based commercial space layout optimization device; or it may exist independently and not be assembled into a neural data-based commercial space layout optimization device.

[0144] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by a neural data-based commercial space layout optimization device, the device performs the following actions: acquires booth space layout data, which includes the coordinate range of the booth's physical space and spatial location information corresponding to each exhibit within the coordinate range; acquires multimodal interest data of multiple acquisition targets within the coordinate range, including the acquisition targets' electroencephalogram (EEG) data, eye-tracking data, and spatial trajectory data; aligns the EEG data, eye-tracking data, and spatial trajectory data of each acquisition target by timestamp and performs spatial registration with the spatial location information in the booth space layout data to obtain interest association data of each acquisition target at each spatial location unit; determines a group interest metric at each spatial location unit based on the interest association data of each acquisition target at each spatial location unit; and generates a layout optimization scheme for the booth's physical space based on the group interest metric at each spatial location unit and the booth space layout data.

[0145] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0146] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation that may be implemented in systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0147] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0148] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described neural data-based commercial space layout optimization method. This addresses the technical problem that booth space layout optimization schemes based on the coordinates of the highest physical channel traffic result in poor optimization performance. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the neural data-based commercial space layout optimization method provided in the above embodiments, and will not be elaborated upon here.

[0149] This application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described business space layout optimization method based on neural data.

[0150] The computer program product provided in this application can solve the technical problem that the optimization scheme for booth space layout is based on the coordinate position with the largest physical channel traffic, resulting in poor optimization effect. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the commercial space layout optimization method based on neural data provided in the above embodiments, and will not be repeated here.

[0151] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.

Claims

1. A commercial spatial layout optimization method based on neural data, characterized in that, The neural data-based commercial spatial layout optimization method includes: Obtain booth space layout data, which includes the coordinate range of the physical space of the booth and the spatial location information of each exhibit within the coordinate range; Acquire multimodal interest data of multiple acquisition targets within the coordinate range, wherein the multimodal interest data includes EEG data, eye movement data, and spatial trajectory data of the acquisition targets; After aligning the EEG data, eye movement data, and spatial trajectory data of each of the aforementioned acquisition targets according to timestamps, spatial registration is performed with the spatial location information in the booth spatial layout data to obtain the interest association data of each of the aforementioned acquisition targets at each spatial location unit; Based on the interest association data of each of the collected targets at each spatial location unit, determine the group interest metric value at each spatial location unit; Based on the group interest metric value at each spatial location unit and the booth space layout data, a layout optimization scheme for the physical space of the booth is generated. The step of generating a layout optimization scheme for the physical space of the booth based on the group interest metric value at each spatial location unit and the booth space layout data includes: The group interest metric values ​​at the spatial location units corresponding to each exhibit are sorted in descending order to generate exhibit interest ranking results. Based on the interest ranking results of the exhibits, the recommended placement priority of each exhibit is determined. The exhibits with the top N group interest metric values ​​are marked as high-priority exhibits, and the high-priority exhibits are adjusted within the coordinate range of the physical space of the booth to a region where the distance between the booth entrance and the booth entrance is less than a preset distance threshold. Obtain the preset functional association attributes between each exhibit, mark exhibits with the same functional association attributes as an associated exhibit group, and adjust each exhibit in the associated exhibit group to an adjacent area where the spatial distance between them is less than a preset spacing threshold. The booth space layout data is updated based on the adjusted spatial location information of each exhibit to generate a layout optimization scheme that includes the spatial location information of each exhibit's objectives.

2. The commercial spatial layout optimization method based on neural data as described in claim 1, characterized in that, The acquisition of booth space layout data includes: A three-dimensional spatial coordinate system covering the physical space of the booth is established with the designated location of the booth as the origin, and the coordinate range of the physical space of the booth is determined. In the three-dimensional spatial coordinate system, the coordinates of the outer contour boundary points of each exhibit within the physical space of the booth are measured by a ranging device. The spatial area enclosed by the coordinates of the outer contour boundary points is determined as the spatial location information of the corresponding exhibit, and an exhibit identifier is assigned to each exhibit. The exhibit identifier and corresponding spatial location information of each exhibit are associated and stored with the coordinate range of the physical space of the booth to generate the booth space layout data.

3. The commercial spatial layout optimization method based on neural data as described in claim 1, characterized in that, The acquisition of multimodal interest data of multiple acquisition targets within the coordinate range includes: Acquire initial EEG data and initial eye movement data of the target within the coordinate range; The initial spatial trajectory data of the target is collected by a spatial positioning tag, which is configured to receive signals from positioning base stations deployed in the physical space of the booth. The initial EEG data, initial eye movement data, and initial spatial trajectory data of each of the acquisition targets are preprocessed to remove invalid data segments and noise data in each data stream, so as to obtain the EEG data, eye movement data, and spatial trajectory data corresponding to each acquisition target as the multimodal interest data.

4. The commercial spatial layout optimization method based on neural data as described in claim 3, characterized in that, The initial EEG data, initial eye-tracking data, and initial spatial trajectory data of each of the aforementioned acquisition targets are preprocessed to remove invalid data segments and noise data from each data stream, resulting in EEG data, eye-tracking data, and spatial trajectory data corresponding to each acquisition target as the multimodal interest data, including: The initial EEG data is bandpass filtered to retain the EEG signal components within the target frequency band. The EEG data is then obtained by identifying and removing the electrooculogram (EOG) artifacts and electromyogram (EMG) artifacts from the EEG data using an independent component analysis algorithm. The initial eye-tracking data is subjected to invalid frame removal processing. The invalid frames include off-screen data frames and frame data with acquisition confidence lower than a preset confidence threshold. The initial eye-tracking data is then corrected in spatial coordinate system based on the factory calibration parameters of the eye-tracking acquisition device, and the effective fixation point sequence is extracted to obtain the eye-tracking data. The initial spatial trajectory data is subjected to outlier removal processing, and the initial spatial trajectory data after outlier removal is smoothed and filtered to obtain the spatial trajectory data. The EEG data, eye-tracking data, and spatial trajectory data are aligned and bound according to a unified timestamp to obtain the multimodal interest data of each acquisition target.

5. The commercial spatial layout optimization method based on neural data as described in claim 1, characterized in that, The step of aligning the EEG data, eye-tracking data, and spatial trajectory data of each of the aforementioned targets according to timestamps, and then spatially registering them with the spatial location information in the booth spatial layout data, yields interest association data for each of the aforementioned targets at each spatial location unit, including: The gaze coordinates in the eye movement data of each of the aforementioned targets are mapped from the coordinate system of the eye movement acquisition device to the three-dimensional spatial coordinate system to obtain the three-dimensional gaze space coordinates of the gaze point at each moment in the physical space of the booth. Spatial matching is performed between the three-dimensional gaze space coordinates at each moment and the spatial location information of each exhibit in the booth space layout data to determine the target exhibit identifier corresponding to the gaze point at each moment; Based on the spatial trajectory data of the target at each moment and the corresponding three-dimensional gaze space coordinates, calculate the gaze direction vector of the target at each moment, and determine whether the angle between the gaze direction vector and the spatial vector from the target to the target exhibit is less than a preset angle threshold. If it is less, then the EEG data, eye movement data and the corresponding target exhibit identifier at that moment are associated and recorded as valid interest association data. The coordinate range of the physical space of the booth is divided into multiple spatial location units, either by the spatial location information corresponding to each exhibit or according to a preset grid size. Remove the associated data corresponding to the moments when the included angle is greater than or equal to the preset angle threshold, and summarize the valid interest-related data of each moment in chronological order according to the dimension of the acquisition target to obtain the interest-related data of each acquisition target at each spatial location unit.

6. The commercial spatial layout optimization method based on neural data as described in claim 1, characterized in that, The step of determining the group interest metric value at each spatial location unit based on the interest association data of each of the collected targets at each spatial location unit includes: For each spatial location unit, acquire all interest-related data of all acquisition targets within that spatial location unit; Population statistics are performed on the EEG data of each of the collected targets in the interest-related data within the spatial location unit, and the population mean of β wave power and α wave power within the spatial location unit are calculated. Group statistics are performed on the eye movement data of each acquisition target in the interest association data within the spatial location unit to calculate the cumulative number of fixation points, the cumulative value of fixation duration, and the number of acquisition targets that generate fixation within the spatial location unit; The group interest metric of the spatial location unit is calculated based on the group mean of β-wave power, group mean of α-wave power, cumulative number of fixation points, cumulative fixation duration, and number of data acquisition targets that generate fixation.

7. The commercial spatial layout optimization method based on neural data as described in claim 1, characterized in that, The step of generating a layout optimization scheme for the physical space of the booth based on the group interest metric value at each spatial location unit and the booth space layout data further includes: Obtain user profile tags for each of the aforementioned collection targets, wherein the user profile tags include at least one of gender tags, age group tags, and interest preference tags; Based on the user profile tags of each of the collected targets, all collected targets are divided into multiple subgroups, and the collected targets in the same subgroup have the same user profile tags or the same combination of user profile tags. For each subgroup, the subgroup interest metric value at each spatial location unit is calculated from the interest association data of each target in the subgroup at each spatial location unit. The calculation method of the subgroup interest metric value is the same as the calculation method of the group interest metric value. Based on the dwell time of the spatial trajectory data in each spatial location unit, spatial location units with a dwell time exceeding a preset dwelling threshold are marked as dwelling points; The spatial trajectory data of each acquisition target is extracted in time sequence, and the spatial location unit sequence traversed by each acquisition target is clustered by sequence similarity. The sequence cluster containing the most acquisition targets is determined as the high-frequency path. Based on the subgroup interest metric values ​​of each subgroup at each spatial location unit, as well as the stopping point and the high-frequency path, a movement guidance scheme matching each subgroup is generated. The movement guidance scheme includes: a sequence of recommended path exhibits arranged in spatial coordinate order, and spatial location information of the placement of guide signs between adjacent recommended path exhibits.

8. A commercial space layout optimization device based on neural data, characterized in that, The neural data-based commercial space layout optimization device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the neural data-based commercial space layout optimization method as described in any one of claims 1 to 7.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the business spatial layout optimization method based on neural data as described in any one of claims 1 to 7.

Images