Space adjustment method and device for financial network point, electronic equipment and program product

Abstract

The invention discloses a space adjustment method and device for a financial network, electronic equipment and a program product, and relates to the technical field of artificial intelligence, and the space adjustment method comprises the steps: obtaining structured point cloud data and sensor data of a plurality of persons in the financial network, each person having a moving path in the financial network; processing all the point cloud data and all the sensor data by adopting a preset fusion positioning algorithm to obtain target movement track data of each person and a target map of the financial network; inputting each piece of target movement track data into a preset behavior classification model to obtain a target behavior tag of the person; and based on the target map and the point cloud data, determining behavior data in the preset area, and based on the behavior data and the target behavior tag, adjusting the space of the financial network. The technical problem that the service efficiency of the financial network is relatively low due to the fact that the space resources of the financial network cannot be accurately adjusted in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and more specifically, to a method, apparatus, electronic device, and program product for spatial adjustment of financial outlets. Background Technology

[0002] In behavioral monitoring and spatial optimization within financial institutions, operations teams typically utilize tools such as video analytics, wireless positioning, and sensors for spatial optimization. While video analytics tools can achieve pixel-level positioning (accuracy of approximately 0.5 to 1 meter), when interactions such as customers and tellers delivering documents cause obstructions, the target switching rate can exceed 25%, leading to frequent issues with broken personnel trajectories. Furthermore, they can only identify basic actions and cannot interpret business semantics; for example, they cannot distinguish between the behavioral essence of "waiting in line" and "processing business," resulting in a false judgment rate exceeding 50%. Wireless positioning tools are limited by multipath interference, making it difficult to effectively distinguish the positions of customers close to the counter. Additionally, the coordinate data they generate is detached from the semantic information of the spatial map, making it impossible to analyze whether the customer's detour path is obstructed by physical obstacles. As for sensor tools, their hardware (such as ground pressure sensors) usually requires trenching and modification of the ground during configuration, resulting in high costs for physical environmental modifications.

[0003] Because the aforementioned space optimization technologies struggle to accurately understand customer behavior and adjust the spatial resources of financial outlets precisely, service efficiency is low and customer experience is poor.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method, apparatus, electronic device, and program product for adjusting the space of financial outlets, in order to at least solve the technical problem in related technologies that the spatial resources of financial outlets cannot be accurately adjusted, resulting in low service efficiency of financial outlets.

[0006] According to one aspect of the embodiments of this application, a method for spatial adjustment of a financial outlet is provided, comprising: acquiring structured point cloud data and sensor data of multiple personnel within the financial outlet, wherein the personnel have movement paths within the financial outlet; processing all point cloud data and all sensor data using a preset fusion positioning algorithm to obtain target movement trajectory data of each personnel and a target map of the financial outlet; inputting each target movement trajectory data into a preset behavior classification model to obtain target behavior labels for the personnel; determining behavior data within a preset area based on the target map and point cloud data, and adjusting the space of the financial outlet based on the behavior data and target behavior labels.

[0007] Furthermore, before acquiring structured point cloud data and sensor data from multiple personnel within the financial outlet, the process also includes: deploying LiDAR and pre-set sensors at the financial outlet, wherein the pre-set sensors are fixed to the main body of the LiDAR; establishing a spatial coordinate system and determining the initial pose of the LiDAR and pre-set sensors in the spatial coordinate system.

[0008] Furthermore, after determining the initial pose of the lidar and preset sensors in the spatial coordinate system, the process also includes: collecting spatial data of financial outlets based on a preset time synchronization strategy to obtain point cloud data and sensor data of multiple personnel synchronized in time; and encapsulating all point cloud data and all sensor data to obtain structured point cloud data and sensor data.

[0009] Furthermore, the steps of processing all point cloud data and all sensor data using a preset fusion positioning algorithm to obtain target movement trajectory data for each person and target map of the financial outlet include: inputting all point cloud data and all sensor data into the first preset module of the preset fusion positioning algorithm to generate multiple sub-maps of the financial outlet; inputting all sub-maps into the second preset module of the preset fusion positioning algorithm to obtain target map and target movement trajectory data.

[0010] Furthermore, the step of inputting all point cloud data and all sensor data into the first preset module of the preset fusion positioning algorithm to generate multiple sub-maps of financial outlets includes: filtering all point cloud data and all sensor data to obtain filtered data; and generating multiple sub-maps based on the filtered data.

[0011] Furthermore, before inputting the trajectory data of each target into the preset behavior classification model to obtain the target behavior label of the person, the method further includes: constructing a trajectory dataset and constructing an initial behavior classification model, wherein the structure of the initial behavior classification model includes at least: a convolutional layer and a bidirectional recurrent neural network; and training the initial behavior classification model based on the trajectory dataset to obtain the preset behavior classification model.

[0012] Furthermore, the behavioral data includes at least: agglomeration density value and a dwelling index value. Based on the behavioral data and the target behavioral label, the step of adjusting the space of the financial outlet includes: adjusting the number of windows of the financial outlet when the target behavioral label is a first preset label and the agglomeration density value is greater than a preset agglomeration density threshold; or removing the object at the preset location of the financial outlet indicated by the second preset label when the target behavioral label is a second preset label and the dwelling index value is greater than a preset dwelling index threshold.

[0013] According to another aspect of the embodiments of this application, a spatial adjustment device for a financial outlet is also provided, comprising: an acquisition unit, configured to acquire structured point cloud data and sensor data of multiple persons within the financial outlet, wherein the persons have movement paths within the financial outlet; a processing unit, configured to process all point cloud data and all sensor data using a preset fusion positioning algorithm to obtain target movement trajectory data of each person and a target map of the financial outlet; an input unit, configured to input each target movement trajectory data into a preset behavior classification model to obtain target behavior labels for the persons; and an adjustment unit, configured to determine behavior data within a preset area based on the target map and point cloud data, and adjust the space of the financial outlet based on the behavior data and target behavior labels.

[0014] Furthermore, the spatial adjustment device for financial outlets also includes: a first deployment module, used to deploy a lidar and a preset sensor at the financial outlet before acquiring structured point cloud data and sensor data of multiple personnel within the financial outlet, wherein the preset sensor is fixed on the main body of the lidar; and a first determination module, used to establish a spatial coordinate system and determine the initial pose of the lidar and the preset sensor in the spatial coordinate system.

[0015] Furthermore, the spatial adjustment device for financial outlets also includes: a first acquisition module, used to acquire spatial data of the financial outlet based on a preset time synchronization strategy after determining the initial pose of the lidar and preset sensors in the spatial coordinate system, to obtain point cloud data and sensor data of multiple personnel synchronized in time; and a first encapsulation module, used to encapsulate all point cloud data and all sensor data to obtain structured point cloud data and sensor data.

[0016] Furthermore, the processing unit includes: a first generation module, used to input all point cloud data and all sensor data into a first preset module in a preset fusion positioning algorithm to generate multiple sub-maps of the financial outlets; and a first input module, used to input all sub-maps into a second preset module in a preset fusion positioning algorithm to obtain the target map and target movement trajectory data.

[0017] Furthermore, the input unit includes: a first filtering module for filtering all point cloud data and all sensor data to obtain filtered data; and a second generation module for generating multiple sub-maps based on the filtered data.

[0018] Furthermore, the spatial adjustment device for financial outlets also includes: a first construction module, used to construct a trajectory dataset and an initial behavior classification model before inputting the movement trajectory data of each target into a preset behavior classification model to obtain the target behavior label of the person, wherein the structure of the initial behavior classification model includes at least: a convolutional layer and a bidirectional recurrent neural network; and a first training module, used to train the initial behavior classification model based on the trajectory dataset to obtain the preset behavior classification model.

[0019] Furthermore, the behavioral data includes at least: a clustering density value and a dwelling index value. The adjustment unit includes: a first adjustment module, used to adjust the number of windows of the financial outlet when the target behavioral label is a first preset label and the clustering density value is greater than a preset clustering density threshold; and a first removal module, used to remove the object indicating the preset location of the financial outlet by the second preset label when the target behavioral label is a second preset label and the dwelling index value is greater than a preset dwelling index threshold.

[0020] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-described methods for adjusting the space of financial outlets.

[0021] According to another aspect of the embodiments of this application, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement any of the above-described methods for adjusting the space of financial outlets.

[0022] In this invention, structured point cloud data and sensor data of multiple personnel within a financial branch are acquired. A preset fusion positioning algorithm is used to process all point cloud data and sensor data to obtain target movement trajectory data for each personnel and a target map of the financial branch. Each target movement trajectory data is input into a preset behavior classification model to obtain target behavior labels for the personnel. Based on the target map and point cloud data, behavior data within a preset area is determined. Based on the behavior data and target behavior labels, the space of the financial branch is adjusted, thus solving the technical problem in related technologies where the spatial resources of financial branches cannot be accurately adjusted, resulting in low service efficiency of financial branches.

[0023] In this invention, firstly, LiDAR and IMU (Inertial Measurement Unit) sensors deployed within financial branches are used to collect and structure-process point cloud data and motion status information of each person within the branch. Then, a preset fusion positioning algorithm is used to process and integrate the structured data to obtain precise target movement trajectory data for each person, while simultaneously generating a high-precision target map of the entire financial branch. Subsequently, the target movement trajectory data is input into a preset behavior classification model, which can identify and label the behavior patterns of people within the branch, assigning meaning to each movement node, such as queuing or prolonged stay. By analyzing the behavior data of preset areas in the target map and combining it with the target behavior tags of people, the space of the financial branch can be accurately adjusted, thereby improving the service efficiency and enhancing the customer experience of the financial branch. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0025] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a spatial adjustment method for financial outlets is shown.

[0026] Figure 2 This is a flowchart of the spatial adjustment method for financial outlets according to Embodiment 1 of this application;

[0027] Figure 3 This is a schematic diagram of an optional lidar workflow according to an embodiment of this application;

[0028] Figure 4 This is a schematic diagram illustrating the workflow of an optional preset fusion localization algorithm according to an embodiment of this application;

[0029] Figure 5 This is a schematic diagram of an optional visual monitoring panel according to an embodiment of this application;

[0030] Figure 6 This is a schematic diagram of an optional space adjustment device for a financial outlet according to an embodiment of this application;

[0031] Figure 7 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0033] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0034] It should be noted that all related information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) collected and involved in this invention are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data comply with the relevant laws, regulations, and standards of the relevant regions, necessary confidentiality measures have been taken, and it does not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse. For example, this system has an interface with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface, and the relevant information is obtained only after receiving consent from the aforementioned user or organization.

[0035] Due to the current limitations of financial institution branch space optimization technologies, such as insufficient positioning accuracy and high system deployment or modification costs, this invention addresses these issues by combining the Cartographer algorithm (a positioning and mapping technology) for high-precision positioning and mapping. This improves positioning accuracy to the centimeter level, eliminating positioning deviations. The Cartographer algorithm generates continuous semantic maps that integrate environmental features (such as "queue area" and "form filling station"), enabling precise correlation between personnel trajectories and specific business scenarios. This solves the problem of converting raw data into business semantics. Furthermore, a single LiDAR device replaces multiple types of distributed sensors (such as ground pressure sensors and infrared counters), reducing hardware deployment costs. Through the real-time construction of a dynamic occupancy map, it simultaneously monitors the distribution of personnel in multiple areas (such as waiting chair areas and form filling areas), improving the utilization rate of space resources.

[0036] The present invention will now be described in detail with reference to various embodiments.

[0037] Example 1

[0038] According to an embodiment of this application, an embodiment of a method for adjusting the space of financial outlets is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a spatial adjustment method for financial outlets is shown. Figure 1 As shown, computer terminal 10 (or mobile device) may include one or more ( Figure 1 The processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions may also be included. In addition, it may include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera, wherein the network interface can be connected to wired and / or wireless networks. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0040] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0041] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the financial outlet space adjustment method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned financial outlet space adjustment method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0042] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0043] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0044] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for adjusting the spatial distribution of financial outlets is shown. Figure 2 This is a flowchart of the spatial adjustment method for financial outlets according to Embodiment 1 of this application, as follows: Figure 2 As shown, the method includes the following steps:

[0045] Step S201: Obtain structured point cloud data and sensor data of multiple personnel within the financial branch, including the movement paths of the personnel within the financial branch.

[0046] In this embodiment of the invention, structured point cloud data (three-dimensional point data obtained by LiDAR scanning and organized into an ordered list or matrix) and sensor data (data generated by IMU sensors, including three-axis angular velocity, acceleration, etc.) of multiple people within a financial branch are acquired. By acquiring structured data, high accuracy and real-time performance of subsequent analysis are ensured.

[0047] Step S202: A preset fusion positioning algorithm is used to process all point cloud data and all sensor data to obtain the target movement trajectory data of each person and the target map of the financial outlet.

[0048] In this embodiment of the invention, a preset fusion positioning algorithm (such as the Cartographer algorithm) can construct and optimize the environmental map in real time, while simultaneously locating personnel positions. The preset fusion positioning algorithm can be used to process all point cloud data and all sensor data to obtain target movement trajectory data for each person and a target map of the financial branch (i.e., a high-precision grid map of the internal structure and layout of the financial branch). For example, a detailed trajectory of customer A from the entrance to the self-service area and then to the counter area can be generated, while simultaneously constructing a detailed branch map including walls, counters, self-service areas, etc.

[0049] Step S203: Input the movement trajectory data of each target into the preset behavior classification model to obtain the target behavior label of the person.

[0050] In this embodiment of the invention, a model capable of identifying specific behavioral patterns based on human movement trajectories is trained through machine learning. For example, by analyzing historical data, the model can identify behavioral patterns such as queuing, moving, and detouring. By inputting the data of each target movement trajectory into a preset behavior classification model, a target behavior label for the person can be obtained. For example, the model identifies customer A's lingering behavior in front of the self-service area, and therefore assigns a lingering behavior label to customer A.

[0051] Step S204: Based on the target map and point cloud data, determine the behavioral data within the preset area, and adjust the space of the financial outlet based on the behavioral data and target behavioral labels.

[0052] In this embodiment of the invention, based on the target map and point cloud data, behavioral data (such as cluster density (which can be determined by the distribution of point cloud data and the frequency of queuing behavior), dwell index, etc.) within a preset area can be determined. Based on the behavioral data and target behavioral tags, the space of the financial outlet can be adjusted. For example, if it is identified that the number of customers queuing in front of a certain window is too long (i.e., there is queuing behavior and the cluster density is high), more staff can be added or the window opening strategy can be adjusted.

[0053] In summary, by deploying LiDAR and IMU sensors and combining them with the Cartographer fusion positioning algorithm, high-precision real-time positioning and behavioral pattern recognition of personnel within financial branches are achieved. Specifically, structured point cloud data and motion state information can be acquired first, and the Cartographer algorithm can be used to efficiently process this data to generate accurate personnel movement trajectories and branch spatial maps. Then, the trajectory data is input into a pre-trained behavior classification model to obtain semantic labels for personnel behavior. Subsequently, based on the behavior data and labels, the space of the financial branch can be adjusted, such as adjusting the counter layout and adding self-service equipment, thereby improving the efficiency of spatial resource allocation and customer experience. This solves the technical problem in related technologies where the spatial resources of financial branches cannot be accurately adjusted, resulting in low service efficiency at financial branches.

[0054] In order to achieve high-precision real-time perception and spatial intelligent optimization of personnel behavior within the branch, in the spatial adjustment method of financial branch provided in Embodiment 1 of this application, a lidar and a preset sensor are deployed in the financial branch, wherein the preset sensor is fixed on the main body of the lidar; a spatial coordinate system is established, and the initial pose of the lidar and the preset sensor in the spatial coordinate system is determined.

[0055] In this embodiment of the invention, firstly, the LiDAR is positioned high up at the site to cover as much of the field of view as possible, while ensuring that the IMU sensor is tightly integrated with the LiDAR to achieve synchronous acquisition of attitude data. Next, spatial coordinate system calibration is used to determine the initial pose of the fused LiDAR and IMU data, ensuring the consistency and accuracy of data acquisition. For example, the LiDAR is installed above the entrance, covering key areas including the self-service area, form-filling station, and queuing area, while the IMU sensor is rigidly connected to the LiDAR, together forming a high-precision spatial perception system. By implementing coordinate system calibration, the operations team can determine the initial position and attitude of the LiDAR and IMU.

[0056] In order to accurately obtain structured point cloud data and sensor data, in the spatial adjustment method of financial outlets provided in Embodiment 1 of this application, spatial data of financial outlets are collected based on a preset time synchronization strategy to obtain point cloud data and sensor data of multiple personnel synchronized in time; all point cloud data and all sensor data are encapsulated to obtain structured point cloud data and sensor data.

[0057] In this embodiment of the invention, a preset time synchronization strategy (i.e., achieving microsecond-level synchronization of data streams between the lidar and sensor via a hardware-triggered pulse mechanism) is employed to collect spatial data of financial outlets. This ensures strict alignment of data acquisition between the lidar and sensor, avoiding the negative impact of time deviation on positioning and trajectory construction. The lidar generates high-precision point clouds in real time at a frequency of 55.5 microseconds per 360-degree scan (300,000 points per second, ranging 150 meters ± 2 centimeters), while the IMU continuously outputs three-axis angular velocity and acceleration data at a frequency of no less than 200 Hz (yaw angle accuracy ± 0.1 degrees). The collected point cloud data and sensor data need to be transmitted through specific protocols (such as Ethernet or UART (Universal Asynchronous Receiver / Transmitter)). Before transmission, the data can be encapsulated into structured data packets with timestamps (i.e., all point cloud data and all sensor data are encapsulated to obtain structured point cloud data and sensor data). Through time synchronization strategies and data encapsulation, a highly coordinated and structured data source is provided for subsequent spatial intelligent adjustment, which effectively improves the accuracy of positioning and the reliability of spatial optimization decisions.

[0058] Figure 3 This is a schematic diagram of an optional lidar workflow according to an embodiment of this application, such as... Figure 3 As shown, the equipment is first deployed and fixed in a certain location (e.g., an RS-LiDAR-16 lidar is installed high up in a financial institution, and the IMU sensor is rigidly fixed to the lidar body to ensure attitude synchronization), and spatial coordinate system calibration is performed. The initial pose of the lidar and IMU combination in the global coordinate system is determined by measurement. After the physical deployment is completed, hardware initialization is performed, that is, the lidar's TOF (Time of Flight) ranging module and 16-channel transmitter are started. At the same time, the IMU is activated to warm up and zero-bias calibration of the gyroscope and accelerometer. After that, multi-source data can be collected synchronously, and the collected multi-source data is spatiotemporally aligned to obtain aligned data. Then, data encapsulation can be performed.

[0059] In order to accurately obtain the target map and target movement trajectory data, in the spatial adjustment method of financial outlets provided in Embodiment 1 of this application, all point cloud data and all sensor data are input into the first preset module of the preset fusion positioning algorithm to generate multiple sub-maps of the financial outlets; all sub-maps are input into the second preset module of the preset fusion positioning algorithm to obtain the target map and target movement trajectory data.

[0060] In this embodiment of the invention, all point cloud data and all sensor data are input into the first preset module (i.e., the local SLAM (Simultaneous Localization and Mapping) module) in the preset fusion positioning algorithm, which is responsible for real-time positioning and local map construction, and can generate multiple sub-maps of financial outlets. All sub-maps are input into the second preset module (i.e., the global SLAM module) in the preset fusion positioning algorithm, which is responsible for global optimization, map construction and accurate trajectory tracking of personnel, and can obtain target map and target movement trajectory data.

[0061] In order to accurately generate multiple sub-maps, in the spatial adjustment method for financial outlets provided in Embodiment 1 of this application, all point cloud data and all sensor data are filtered to obtain filtered data; based on the filtered data, multiple sub-maps are generated.

[0062] In this embodiment of the invention, redundant or inaccurate data points can be removed through algorithmic processing (i.e., filtering all point cloud data and all sensor data to obtain filtered data), ensuring data quality. The first preset module in the preset fusion positioning algorithm can generate a sub-map reflecting local environmental features in real time based on the filtered data.

[0063] Figure 4 This is a schematic diagram illustrating the workflow of an optional preset fusion localization algorithm according to an embodiment of this application, such as... Figure 4As shown, the input data (i.e., time-synchronized point cloud and IMU data) is input into the Local SLAM module of the preset fusion localization algorithm (i.e., Cartographer). The point cloud data is frame-matched using the Ceres scan matcher (a scan matching algorithm), and the matched data is filtered to obtain the filtered data. At the same time, the IMU data is transformed to the radar coordinate system, so that the two types of data can be fused under the same reference frame. Through data matching and transformation, sub-maps are generated. Each sub-map includes the point cloud data of the local environment and the corresponding pose estimation. Then, multiple sub-maps are input into the Glocal SLAM module of the preset fusion localization algorithm. The constraint relationship between sub-maps is calculated through the mapping algorithm, and the pose of all sub-maps in the world is optimized using the Sparse Pose Adjustment (SPA) technique. As the sub-map generation tends to stabilize, the accumulated errors in the scan matching and sub-map construction process are eliminated through loop closure detection, and the output data (i.e., target map) is generated.

[0064] In order to accurately obtain the preset behavior classification model, in the spatial adjustment method of financial outlets provided in Embodiment 1 of this application, a trajectory dataset is constructed and an initial behavior classification model is constructed. The structure of the initial behavior classification model includes at least a convolutional layer and a bidirectional recurrent neural network. Based on the trajectory dataset, the initial behavior classification model is trained to obtain the preset behavior classification model.

[0065] In this embodiment of the invention, a trajectory dataset (the precise locations and action patterns of personnel at different time points within a network, after preprocessing and annotation) is constructed, and an initial behavior classification model is built. Its network structure includes at least convolutional layers and a bidirectional recurrent neural network. Taking the 6-dimensional features of 30 historical trajectory points as input, key features are extracted after processing by the convolutional layers and the bidirectional recurrent neural network. Finally, the probability distribution of the behavior is output. The categorical_focal_loss function can be used to address the sample imbalance problem. During the training of the initial behavior classification model, an EarlyStopping callback can be set to prevent overfitting, resulting in a preset behavior classification model. By constructing the trajectory dataset and using a deep learning architecture to train the behavior classification model, the accuracy and applicability of behavior recognition are enhanced.

[0066] The behavioral data includes at least: cluster density value and dwell index value. In order to accurately adjust the space of financial outlets, in the space adjustment method of financial outlets provided in Embodiment 1 of this application, when the target behavioral label is a first preset label and the cluster density value is greater than a preset cluster density threshold, the number of windows of the financial outlet is adjusted; or, when the target behavioral label is a second preset label and the dwell index value is greater than a preset dwell index threshold, the object indicating the preset location of the financial outlet by the second preset label is removed.

[0067] In this embodiment of the invention, the distribution of personnel within a preset area is analyzed, and the number of people per unit area (i.e., the cluster density value) and the proportion of personnel whose stay time in the area exceeds the average (i.e., the dwelling index value) are calculated. If the target behavior label is a first preset label (e.g., queuing), and the cluster density value is greater than a preset cluster density threshold (e.g., 3.0 people / square meter), the number of service windows at the financial branch can be adjusted. If the target behavior label is a second preset label (crowd congestion), and the dwelling index value is greater than a preset dwelling index threshold (which can be set manually, e.g., 0.7), objects indicating preset locations (e.g., safety passages) at the financial branch indicated by the second preset label are removed. This adjustment strategy can be pushed to the financial branch's service system in real time via an IoT protocol. Based on the dynamic adjustment of the number of service windows and the removal of obstacles, the space utilization bottleneck is effectively solved, and the service efficiency and customer experience of the branch are improved.

[0068] Figure 5 This is a schematic diagram of an optional visual monitoring panel according to an embodiment of this application, such as... Figure 5 As shown, the visualization monitoring panel includes a spatial map module, a behavior analysis module, and an optimization decision-making workbench. In the spatial map module, maintenance personnel can view the generated centimeter-level raster map, which can be overlaid using the layer controller. In section D1, static structural layers such as walls or counters are displayed in dark colors, while section D2 displays real-time point cloud layers with adjustable transparency. Dynamic targets are presented as colored clustered point clouds, and personnel density is mapped using red, yellow, and blue gradients, refreshing 20 frames per second. The behavior analysis module can display a trajectory tracking view. On the D3 map, which includes network equipment such as walls or counters, the movement path of personnel is displayed as a colored streamline with arrows in the D4 window. Combined with the spatial map module, the current area can be marked, for example: "Stalled_Waiting Area_4 People". When the number of markings reaches a certain limit, an event alarm board can be added to the D5 window of the optimization decision-making workbench. The alarm can be pushed out in a scrolling manner according to preset rules. For example, if the number of people gathered at window 3 in the cash area exceeds the limit of 5 for 120 seconds, the associated target will be highlighted to remind the maintenance personnel of the personnel gathering situation in the current area. The maintenance personnel can optimize the scenario based on the actual situation.

[0069] The spatial adjustment method for financial outlets provided in this application can achieve accurate classification of personnel behavior and intelligent optimization of the internal space of financial outlets by combining time-synchronous acquisition and structured encapsulation of high-precision LiDAR and IMU sensor data, as well as data filtering, sub-map and global map construction based on the Cartographer algorithm. It can accurately identify behavioral patterns such as abnormal gathering and long-term stay, and automatically adjust the number of windows or remove obstacles based on preset thresholds, effectively improving space utilization efficiency and service process speed.

[0070] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0071] Example 2

[0072] This application also provides a space adjustment device for financial outlets. It should be noted that the space adjustment device for financial outlets in this application can be used to execute the space adjustment method for financial outlets provided in this application. The space adjustment device for financial outlets provided in this application is described below.

[0073] According to an embodiment of this application, an apparatus for implementing the above-described method for adjusting the space of financial outlets is also provided. Figure 6 This is a schematic diagram of an optional space adjustment device for a financial branch according to an embodiment of this application, such as... Figure 6 As shown, the space adjustment device for the financial outlet may include: an acquisition unit 60, a processing unit 61, an input unit 62, and an adjustment unit 63.

[0074] The acquisition unit 60 is used to acquire structured point cloud data and sensor data of multiple people within the financial outlet, where the people have movement paths within the financial outlet.

[0075] Processing unit 61 is used to process all point cloud data and all sensor data using a preset fusion positioning algorithm to obtain target movement trajectory data of each person and target map of financial outlets.

[0076] Input unit 62 is used to input the movement trajectory data of each target into a preset behavior classification model to obtain the target behavior label of the person;

[0077] Adjustment unit 63 is used to determine behavioral data within a preset area based on the target map and point cloud data, and to adjust the space of the financial outlet based on the behavioral data and target behavioral labels.

[0078] The spatial adjustment device for financial outlets provided in this application embodiment can acquire structured point cloud data and sensor data of multiple personnel within the financial outlet through the acquisition unit 60. The processing unit 61 can process all point cloud data and all sensor data using a preset fusion positioning algorithm to obtain target movement trajectory data of each person and target map of the financial outlet. The input unit 62 can input each target movement trajectory data into a preset behavior classification model to obtain target behavior labels of the personnel. The adjustment unit 63 can determine the behavior data within a preset area based on the target map and point cloud data, and adjust the space of the financial outlet based on the behavior data and target behavior labels.

[0079] Optionally, the spatial adjustment device for financial outlets further includes: a first deployment module, used to deploy a lidar and a preset sensor at the financial outlet before acquiring structured point cloud data and sensor data of multiple personnel within the financial outlet, wherein the preset sensor is fixed on the main body of the lidar; and a first determination module, used to establish a spatial coordinate system and determine the initial pose of the lidar and the preset sensor in the spatial coordinate system.

[0080] Optionally, the spatial adjustment device for financial outlets further includes: a first acquisition module, used to acquire spatial data of the financial outlet based on a preset time synchronization strategy after determining the initial pose of the lidar and preset sensors in the spatial coordinate system, to obtain point cloud data and sensor data of multiple personnel synchronized in time; and a first encapsulation module, used to encapsulate all point cloud data and all sensor data to obtain structured point cloud data and sensor data.

[0081] Optionally, the processing unit 61 includes: a first generation module, used to input all point cloud data and all sensor data into a first preset module in a preset fusion positioning algorithm to generate multiple sub-maps of financial outlets; and a first input module, used to input all sub-maps into a second preset module in a preset fusion positioning algorithm to obtain target map and target movement trajectory data.

[0082] Optionally, the input unit 62 includes: a first filtering module for filtering all point cloud data and all sensor data to obtain filtered data; and a second generation module for generating multiple sub-maps based on the filtered data.

[0083] Optionally, the spatial adjustment device for financial outlets further includes: a first construction module, used to construct a trajectory dataset and an initial behavior classification model before inputting the movement trajectory data of each target into a preset behavior classification model to obtain the target behavior label of the person, wherein the structure of the initial behavior classification model includes at least: a convolutional layer and a bidirectional recurrent neural network; and a first training module, used to train the initial behavior classification model based on the trajectory dataset to obtain the preset behavior classification model.

[0084] Optionally, the behavioral data includes at least: agglomeration density value and a retention index value. The adjustment unit 63 includes: a first adjustment module, used to adjust the number of windows of the financial outlet when the target behavioral label is a first preset label and the agglomeration density value is greater than a preset agglomeration density threshold; and a first removal module, used to remove the object indicating the preset location of the financial outlet by the second preset label when the target behavioral label is a second preset label and the retention index value is greater than a preset retention index threshold.

[0085] The aforementioned space adjustment device for financial outlets may also include a processor and a memory. The aforementioned acquisition unit 60, processing unit 61, input unit 62, adjustment unit 63, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0086] The aforementioned processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured. By adjusting kernel parameters, behavioral data within a preset area can be determined based on the target map and point cloud data. Based on this behavioral data and target behavioral labels, the spatial layout of the financial branch can be adjusted.

[0087] The aforementioned memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0088] It should be noted that the aforementioned acquisition unit 60, processing unit 61, input unit 62, and adjustment unit 63 correspond to steps S201 to S204 in Embodiment 1. The instances and application scenarios implemented by these units and their corresponding steps are the same, but they are not limited to the content disclosed in Embodiment 1. It should also be noted that these units can be hardware or software components stored in a memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). These units can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0089] Example 3

[0090] Embodiments of this application may provide a computer terminal, which may be any computer terminal device in a group of computer terminals. Optionally, in this embodiment, the aforementioned computer terminal may also be replaced with a mobile terminal or an electronic device, etc.

[0091] Optionally, in this embodiment, the computer terminal may be located in at least one of a plurality of network devices in a computer network.

[0092] In this embodiment, the aforementioned computer terminal can execute the program code for the following steps in the spatial adjustment method for financial outlets: acquiring structured point cloud data and sensor data of multiple personnel within the financial outlet, wherein the personnel have movement paths within the financial outlet; processing all point cloud data and all sensor data using a preset fusion positioning algorithm to obtain target movement trajectory data for each personnel and a target map of the financial outlet; inputting each target movement trajectory data into a preset behavior classification model to obtain target behavior labels for the personnel; determining behavior data within a preset area based on the target map and point cloud data, and adjusting the space of the financial outlet based on the behavior data and target behavior labels.

[0093] Optionally, the aforementioned computer terminal can execute program code for the following steps in the spatial adjustment method for financial outlets: deploying a lidar and a preset sensor at the financial outlet, wherein the preset sensor is fixed on the main body of the lidar; establishing a spatial coordinate system and determining the initial pose of the lidar and the preset sensor in the spatial coordinate system.

[0094] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the spatial adjustment method for financial outlets: based on a preset time synchronization strategy, collect spatial data of the financial outlets to obtain point cloud data and sensor data of multiple personnel synchronized in time; encapsulate all point cloud data and all sensor data to obtain structured point cloud data and sensor data.

[0095] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the spatial adjustment method for financial outlets: inputting all point cloud data and all sensor data into the first preset module of the preset fusion positioning algorithm to generate multiple sub-maps of the financial outlets; inputting all sub-maps into the second preset module of the preset fusion positioning algorithm to obtain the target map and target movement trajectory data.

[0096] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the spatial adjustment method for financial outlets: filtering all point cloud data and all sensor data to obtain filtered data; and generating multiple sub-maps based on the filtered data.

[0097] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the spatial adjustment method for financial outlets: constructing a trajectory dataset and constructing an initial behavior classification model, wherein the structure of the initial behavior classification model includes at least: a convolutional layer and a bidirectional recurrent neural network; training the initial behavior classification model based on the trajectory dataset to obtain a preset behavior classification model.

[0098] Optionally, the aforementioned computer terminal can execute program code for the following steps in the spatial adjustment method for financial outlets: adjusting the number of windows in the financial outlet when the target behavior label is a first preset label and the aggregation density value is greater than a preset aggregation density threshold; or removing the object at the preset location of the financial outlet indicated by the second preset label when the target behavior label is a second preset label and the dwell index value is greater than a preset dwell index threshold.

[0099] Optionally, Figure 7 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 7 As shown, the electronic device may include: one or more ( Figure 7 (Only one is shown) processor 702, memory 704, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0100] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the financial outlet space adjustment method and apparatus in this application embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the aforementioned financial outlet space adjustment method. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0101] The processor can access the information and application programs stored in the memory via the transmission device to execute the steps described above in the spatial adjustment method for financial outlets.

[0102] The present application provides a solution for adjusting the space of financial outlets. By deploying integrated LiDAR and IMU sensors and processing them using the Cartographer algorithm, trajectory data of personnel and target maps are obtained. The trajectory data can be input into a preset behavior classification model to obtain the behavior patterns of personnel. Based on behavior pattern recognition, spatial resources can be dynamically adjusted, thereby solving the technical problem in related technologies that the spatial resources of financial outlets cannot be accurately adjusted, resulting in low service efficiency of financial outlets.

[0103] Those skilled in the art will understand that Figure 7 The structure shown is for illustrative purposes only. Electronic devices can also be terminal devices such as smartphones, tablets, PDAs, and mobile internet devices (MIDs). Figure 7 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 7 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 7 The different configurations shown.

[0104] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0105] Example 4

[0106] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the financial outlet space adjustment method provided in Embodiment 1.

[0107] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0108] This application also provides a computer program product, which, when executed on a data processing device, is suitable for performing the steps of a method for adjusting the space of financial outlets.

[0109] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0110] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0111] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0112] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0113] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0114] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0115] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for spatial adjustment of financial outlets, characterized in that, include: Obtain structured point cloud data and sensor data of multiple personnel within a financial branch, wherein the personnel have movement paths within the financial branch; A preset fusion positioning algorithm is used to process all the point cloud data and all the sensor data to obtain the target movement trajectory data of each person and the target map of the financial outlet; Each target movement trajectory data is input into a preset behavior classification model to obtain the target behavior label of the person; Based on the target map and the point cloud data, behavioral data within a preset area is determined, and the space of the financial outlet is adjusted based on the behavioral data and the target behavioral tags.

2. The spatial adjustment method for financial outlets according to claim 1, characterized in that, Before acquiring structured point cloud data and sensor data from multiple individuals within a financial branch, the process also includes: A lidar and a preset sensor are deployed at the financial outlet, wherein the preset sensor is fixed to the main body of the lidar; Establish a spatial coordinate system and determine the initial pose of the lidar and the preset sensor in the spatial coordinate system.

3. The spatial adjustment method for financial outlets according to claim 2, characterized in that, After determining the initial poses of the lidar and the preset sensor in the spatial coordinate system, the method further includes: Based on a preset time synchronization strategy, spatial data of the financial outlets are collected to obtain point cloud data of multiple personnel and sensor data that are synchronized in time. All the point cloud data and all the sensor data are encapsulated to obtain the structured point cloud data and sensor data.

4. The spatial adjustment method for financial outlets according to claim 1, characterized in that, The steps of processing all the point cloud data and all the sensor data using a preset fusion positioning algorithm to obtain the target movement trajectory data of each person and the target map of the financial outlet include: All the point cloud data and all the sensor data are input into the first preset module of the preset fusion positioning algorithm to generate multiple sub-maps of the financial outlet; All the sub-maps are input into the second preset module of the preset fusion positioning algorithm to obtain the target map and the target movement trajectory data.

5. The spatial adjustment method for financial outlets according to claim 4, characterized in that, The step of inputting all the point cloud data and all the sensor data into the first preset module of the preset fusion positioning algorithm to generate multiple sub-maps of the financial outlet includes: All the point cloud data and all the sensor data are filtered to obtain filtered data; Based on the filtered data, multiple sub-maps are generated.

6. The spatial adjustment method for financial outlets according to claim 1, characterized in that, Before inputting each of the target movement trajectory data into a preset behavior classification model to obtain the target behavior label of the person, the method further includes: Construct a trajectory dataset and build an initial behavior classification model, wherein the structure of the initial behavior classification model includes at least: a convolutional layer and a bidirectional recurrent neural network; Based on the trajectory dataset, the initial behavior classification model is trained to obtain the preset behavior classification model.

7. The spatial adjustment method for financial outlets according to claim 1, characterized in that, The behavioral data includes at least: cluster density value and retention index value. The step of adjusting the space of the financial outlet based on the behavioral data and the target behavioral label includes: If the target behavior label is a first preset label and the aggregation density value is greater than a preset aggregation density threshold, the number of windows at the financial outlets may be adjusted, or; If the target behavior label is the second preset label and the retention index value is greater than the preset retention index threshold, remove the object at the preset location of the financial outlet indicated by the second preset label.

8. A spatial adjustment device for financial outlets, characterized in that, include: The acquisition unit is used to acquire structured point cloud data and sensor data of multiple people within a financial outlet, wherein the people have a movement path within the financial outlet; The processing unit is used to process all the point cloud data and all the sensor data using a preset fusion positioning algorithm to obtain the target movement trajectory data of each person and the target map of the financial outlet. An input unit is used to input the movement trajectory data of each target into a preset behavior classification model to obtain the target behavior label of the person. The adjustment unit is used to determine behavioral data within a preset area based on the target map and the point cloud data, and to adjust the space of the financial outlet based on the behavioral data and the target behavioral tags.

9. A computer program product, characterized in that, The method includes a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the spatial adjustment method for financial outlets as described in any one of claims 1 to 7.

10. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the spatial adjustment method for financial outlets as described in any one of claims 1 to 7.

Images