Garden inspection method and device based on gait prediction and route matching, and medium

Abstract

The application discloses a park inspection method and device based on gait prediction and route matching, and a medium, and relates to the fields of intelligent parks and mobile security monitoring. The method comprises the following steps: acquiring satellite positioning data in real time during the travel along the inspection task route, and monitoring whether the preset signal requirement is met; when the requirement is met, collecting real-time inertial measurement data in the travel process, determining real-time gait parameters; recursively predicting a plurality of candidate position points of the travel estimated trajectory point at the current time, calculating the perpendicular distance of deviation from the preset inspection route; assigning corresponding position weights, generating new candidate position points, and taking the weighted average coordinates as the current trajectory point at the current time; calculating the confidence degree of the current trajectory point, comparing the confidence degree with a preset confidence threshold, marking the confidence state based on the comparison result; and calculating the coverage degree to generate an inspection completion degree record. Through gait prediction and route matching, the trajectory is continuously tracked and the inspection completion degree is quantified when there is no GPS.

Description

Technical Field

[0001] This application relates to the fields of smart park and mobile security monitoring technology, specifically to park inspection methods, equipment and media based on gait prediction and route matching. Background Technology

[0002] With the continuous deepening of smart park construction, security patrol systems have become a core component of the security and prevention systems of industrial parks and science and technology parks. Mobile patrol apps, as important execution tools for these systems, undertake key tasks such as real-time location tracking of patrol personnel, environmental status reporting, and abnormal event collection. As parks expand and security requirements increase, patrol areas now cover multiple indoor and outdoor scenarios, patrol routes dynamically change, and multiple patrol personnel operate concurrently, becoming the norm. In mobile environments, issues such as background location interruptions and data loss are prone to occur. Therefore, achieving continuous, accurate, and multi-dimensional security patrols in the non-steady-state operating environment of mobile devices has become particularly important.

[0003] Existing mobile security patrol technologies for industrial parks require management to issue patrol tasks containing preset locations and fixed routes to patrol personnel's mobile terminals. Patrol personnel then sequentially arrive at each location and use software (APP) to scan QR codes, take photos with watermarks, and report anomalies. The system records and replays the trajectory based on the GPS location of each location and uploads the data to the cloud via a real-time network. However, mobile terminals are prone to problems such as location interruptions and discontinuous trajectories in unstable operating environments such as when the terminal is off, running in the background, or in weak network conditions. Furthermore, relying on discrete check-in points makes it difficult to continuously collect reliable location and environmental data, resulting in incomplete security situational awareness and hindering accurate control of the park's security status. Summary of the Invention

[0004] To address the aforementioned issues, this application proposes a park inspection method based on gait prediction and route matching, applied to mobile terminals. The method includes: Receive inspection tasks, load the corresponding inspection task routes, and acquire satellite positioning data in real time while traveling along the inspection task routes, and monitor whether the signal quality of the satellite positioning data meets the preset signal requirements. When the signal quality meets the preset signal requirements, real-time inertial measurement data during the movement is collected to determine real-time gait parameters; the gait parameters include at least stride length, heading angle, and gait timestamp. Based on real-time step size and real-time heading angle, multiple candidate positions of the current estimated trajectory point are recursively predicted, and the vertical distance of each candidate position point from the preset inspection route is calculated. Based on the vertical deviation distance, a corresponding position weight is assigned to each candidate position point. Based on the position weight, a new candidate position point for the estimated trajectory point is generated. The weighted average coordinates of the new candidate position points are used as the current trajectory point at the current moment. Based on the current positioning mode and current operating status, calculate the confidence level of the current trajectory point, compare the confidence level with a preset confidence threshold, and mark the confidence status of the current trajectory point based on the comparison result; Once the movement is determined to have stopped, the coverage of key sections of the preset inspection route by the movement process is calculated based on all the track points that have been marked as reliable, and an inspection completion record is generated.

[0005] On the other hand, this application also proposes a park inspection device based on gait prediction and route matching, including: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform a park inspection method based on gait prediction and route matching as described in the above example.

[0006] On the other hand, this application also proposes a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as: a park inspection method based on gait prediction and route matching as described in the above example.

[0007] The park inspection method proposed in this application, based on gait prediction and route matching, can bring the following benefits: By deeply integrating the geometric matching constraints of the preset inspection route with the inertial estimation of pedestrian gait, the system automatically switches to the candidate location point recursion and weight correction mode when the satellite signal quality is below the threshold. This enables the mobile terminal to continuously generate a smooth trajectory along the preset route even when the screen is off, running in the background, or in a weak network environment. This completely eliminates the trajectory breakpoints and perception blind spots caused by positioning interruptions in traditional inspection schemes, and achieves continuous trajectory tracking throughout the entire process.

[0008] Meanwhile, the confidence dynamic model is used to mark the track points with a reliable status, and the coverage of key road sections is quantified based on the reliable track points to generate inspection completion records. This upgrades the discrete check-in points to continuous road section coverage assessment, which significantly improves the completeness of the security situation awareness of the park and the quantifiability of the quality of inspection task completion, providing an objective basis for anomaly tracing and performance evaluation. Attached Figure Description

[0009] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating a park inspection method based on gait prediction and route matching in an embodiment of this application. Figure 2 This is a schematic diagram of a park inspection device based on gait prediction and route matching in an embodiment of this application. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0012] like Figure 1 As shown in the figure, this application provides a park inspection method based on gait prediction and route matching, including: S101. Receive the inspection task, load the corresponding inspection task route, and acquire satellite positioning data in real time during the movement along the inspection task route, and monitor whether the signal quality of the satellite positioning data meets the preset signal requirements.

[0013] In practice, the management terminal issues inspection tasks to the mobile terminals held by the inspectors. Each inspection task includes at least one preset inspection route, which is composed of a series of ordered lines connecting geographic coordinate points. This route can be a polyline or a continuous path fitted by a Bézier curve, and it defines the movement trajectory that the inspector should follow during the task. After receiving the inspection task, the mobile terminal loads the preset inspection route into its local memory and parses the geometric information of the path, including the latitude and longitude of each path point, their order, and the equations of the line segments between adjacent path points.

[0014] In one embodiment of this application, the mobile terminal continuously detects motion acceleration through a built-in motion sensor. When the amplitude of the motion acceleration exceeds a preset trigger threshold, it is determined that the inspector has begun to move, and the satellite positioning module is then activated to continuously acquire satellite positioning data. The satellite positioning data includes at least the following parameters: longitude, latitude, altitude, horizontal positioning error radius, number of satellites participating in the positioning, and accuracy attenuation factor.

[0015] The mobile terminal extracts signal quality parameters from real-time acquired satellite positioning data. These parameters include at least one of the following: the number of satellites involved in the positioning process, the horizontal positioning error radius, and the accuracy attenuation factor. Preset signal requirement thresholds are configured according to the park's security positioning needs; for example, the number of satellites must not be less than a preset threshold, the horizontal positioning error radius must not be greater than a preset error threshold, and the accuracy attenuation factor must not be greater than a preset attenuation threshold.

[0016] The mobile terminal compares the real-time extracted signal quality parameters one by one with the corresponding preset signal requirement thresholds: if the number of satellites is not less than the preset number threshold, the horizontal positioning error radius is not greater than the preset error threshold, and the accuracy attenuation factor is not greater than the preset attenuation threshold, then the signal quality of the current satellite positioning signal is determined to meet the preset signal requirements; otherwise, it is determined not to meet them. This determination result will be used in subsequent steps to decide whether to directly generate trajectory points using satellite positioning or switch to the calculation mode based on gait parameters and route geometric constraints.

[0017] It should be noted that when a mobile terminal initiates an inspection task, the user can be guided to add the application to the system's power-saving policy whitelist and request foreground service permissions to reduce the probability of the operating system forcibly freezing the process while running in the background. However, in this application, even if the terminal device enters the background or is in a screen-off state, subsequent steps still maintain the continuity of the trajectory through continuous acquisition of inertial measurement data and geometric constraint calculation, thereby ensuring the integrity of the inspection task.

[0018] S102. When the signal quality meets the preset signal requirements, collect real-time inertial measurement data during the movement process to determine real-time gait parameters; the gait parameters include at least stride length, heading angle, and gait timestamp.

[0019] In practice, when the signal quality of the current satellite positioning signal is determined to be lower than the preset signal requirement threshold, such as insufficient satellite quantity, excessive positioning error radius, or excessive accuracy attenuation factor, it indicates that satellite positioning is unavailable or the accuracy cannot meet the requirements for continuity and accuracy of the inspection trajectory. At this time, the mobile terminal automatically switches to the assisted positioning mode based on inertial measurement and route geometric constraints.

[0020] The mobile terminal uses its built-in inertial measurement unit, including an IMU containing a three-axis accelerometer and a three-axis gyroscope, to continuously collect real-time inertial measurement data during movement at a preset sampling frequency. The inertial measurement data includes the three-axis linear acceleration values ​​from the accelerometer and the three-axis angular velocity values ​​from the gyroscope.

[0021] Based on real-time inertial measurement data, real-time gait parameters are determined. In practice, the mobile terminal preprocesses the collected real-time inertial measurement data by removing gravity components and performing low-pass filtering, extracting the vertical acceleration change waveform. A peak detection algorithm is used to identify the ground contact and take-off times during walking. The time interval between two adjacent ground contact times is one gait cycle.

[0022] Based on the integral of acceleration amplitude within the gait cycle, combined with a pre-defined stride length estimation model, such as an empirical model established based on the relationship between height, stride frequency, and acceleration variance, the current stride length is calculated. Simultaneously, the three-axis angular velocity data from the gyroscope are integrated to obtain attitude angles, including roll, pitch, and yaw. Accelerometer data is used to perform gravity-assisted corrections on the pitch and roll angles, and magnetometer data is used to perform geomagnetic corrections on the yaw angle, ultimately outputting a stable heading angle. The detected ground contact moment is used as a gait timestamp to identify the start time of each step.

[0023] S103. Based on the real-time step size and real-time heading angle, recursively predict multiple candidate positions of the current estimated trajectory point, and calculate the vertical distance of each candidate position point from the preset inspection route.

[0024] In practice, the mobile terminal first checks its local cache for the existence of a set of candidate location points generated in the previous moment. If not, meaning this is the first calculation after the satellite signal interruption, the most recently output reliable trajectory point before the current moment is used as the reference position, typically the last high-confidence satellite positioning point. Around this reference position, a preset number of initial candidate location points are randomly generated according to a preset Gaussian distribution rule, and each candidate location point is assigned an equal initial weight.

[0025] For each predicted candidate location point, the mobile terminal assigns a location weight based on its vertical deviation according to a preset distance-weight mapping function. The basic principle of this distance-weight mapping function is: the smaller the vertical deviation of a candidate location point, the closer it is to the preset inspection route, and its location weight is higher; the larger the vertical deviation of a candidate location point, the more likely it is caused by inertial recursion error or abnormal motion, and its location weight is lower.

[0026] In one embodiment of this application, if a set of candidate positions from the previous time step exists in the local cache, then for each candidate position in the previous time step set, the position is recursively calculated based on the current step size and heading angle. The recursion method is as follows: using the step size as the moving distance and the heading angle as the moving direction, the new coordinates of each candidate position after movement are calculated. Let the coordinates of the i-th candidate position from the previous time step be... The current step size is The current heading angle is Then the new coordinates after recursion Calculate using the following formula: .

[0027] Where i is the index of the candidate location point, and its value ranges from 1 to the preset total number N; The unit is meters. The unit is radians. All the recursively derived new coordinates are used as the predicted candidate location points for the current time, forming the candidate location point set for the current time.

[0028] For each candidate location point in the current candidate location point set, the mobile terminal calculates the shortest perpendicular distance between the current location of that point and the preset inspection route. The preset inspection route can be mathematically represented as a series of sequentially connected line segments. Let the coordinates of the candidate location point be... The preset inspection route consists of M line segments, with the endpoint of the k-th line segment being... and Calculation points perpendicular distance to the k-th line segment If the perpendicular falls within the range of the line segment, then the perpendicular distance is taken; otherwise, the minimum Euclidean distance from the point to the two endpoints of the line segment is taken. The vertical deviation of the candidate location point from the entire route is then calculated. for: This is the minimum distance among all line segments. (Distance from vertical) The unit is meters. The smaller the value, the closer the candidate location is to the preset inspection route. The larger the value, the farther the deviation.

[0029] It should be noted that the vertical deviation will serve as the basis for assigning weights to candidate location points in subsequent steps, thereby enabling the correction of trajectory estimation by route geometric constraints.

[0030] S104. Based on the vertical deviation distance, assign corresponding position weights to each candidate position point, generate new candidate position points for the estimated trajectory point based on the position weights, and use the weighted average coordinates of the new candidate position points as the current trajectory point at the current moment.

[0031] Specifically, the mobile terminal uses the geometric information of the preset inspection route to correct the distribution of the candidate location point set, so that candidate location points that deviate far from the route are suppressed and candidate location points that deviate close to the route are enhanced. Finally, the reliable trajectory point at the current moment is obtained by weighted averaging.

[0032] In practical implementation, let's assume there are a total of [number] at the current moment. For the nth candidate location point, Candidate location points The calculated vertical deviation distance is Mobile terminals according to Assign an initial weight to each candidate location point The allocation rule is as follows: the smaller the deviation from the vertical distance, the higher the weight; the larger the deviation from the vertical distance, the lower the weight. This application uses an exponential distance-weight mapping function, and the specific calculation formula is as follows: .

[0033] in, For the natural constant An exponential function with base 0; For the first The vertical deviation of each candidate location point from the preset inspection route, in meters; This is a preset scale parameter, in meters, used to control the rate at which the weight decays with distance. The smaller the value, the stronger the spatial selectivity of the weights. This mapping function makes when The time weight is the maximum value of 1, when As the weight increases, the weight monotonically decreases and approaches 0, thus effectively suppressing candidate locations that are far from the route.

[0034] The mobile terminal normalizes the initial weights to make the sum of the weights of all candidate locations equal to 1. The normalized weights are then... Calculate using the following formula: Among them, the denominator This is the sum of the initial weights for all candidate locations. After normalization, each... satisfy and , Reflects the first Each candidate location point is a probability estimate of the inspector's actual location.

[0035] Mobile terminal based on normalized weight distribution The current set of candidate locations is resampled. The purpose of resampling is to generate a new set of candidate locations according to their weights; regions with higher weights generate more candidate locations, and regions with lower weights generate fewer candidate locations. This application uses a systematic resampling method, the specific process of which is as follows: First, generate a value between 0 and... Random numbers between Then for Calculate the cumulative weight And judge in turn Which cumulative weight interval does it fall into? At that time, the first The coordinates of each candidate location point are copied to form a new candidate location point t. This process is repeated N times to obtain a set of N new candidate location points. After resampling, the weights of all new candidate location points are reset to be equal, but the distribution of the point set is concentrated in the high-weight region.

[0036] The mobile terminal performs a weighted average on the new candidate location points obtained after resampling, and uses the weighted average coordinates as the trajectory point output at the current time. Since the weights of each new candidate location point are equal after resampling, the weighted average degenerates into an arithmetic average. Let the coordinates of the t-th new candidate location point be... Then the coordinates of the trajectory point at the current moment. Calculate using the following formula:

[0037] Where N is the preset total number of candidate location points. Here, x and y are the x and y coordinates of the t-th new candidate position, respectively. This weighted average coordinate is the current trajectory point at the current moment.

[0038] It should be noted that, through the process of weight allocation, resampling, and weighted averaging, the mobile terminal can continuously generate a continuous and smooth sequence of trajectory points along the route direction by utilizing gait parameters and the geometric constraints of the preset inspection route, even when satellite signals are missing.

[0039] S105. Calculate the confidence level of the current trajectory point based on the current positioning mode and current operating status, compare the confidence level with a preset confidence threshold, and mark the confidence status of the current trajectory point based on the comparison result.

[0040] Specifically, after generating the current trajectory point, the mobile terminal needs to quantitatively evaluate the reliability of the trajectory point to distinguish whether it is generated directly based on high-precision satellite positioning or calculated from candidate location points when satellite signals are missing. At the same time, the foreground or background operation status of the mobile terminal also affects the stability of positioning; therefore, the confidence calculation needs to comprehensively consider both the positioning mode and the operating status.

[0041] In practical implementation, the positioning modes include two types: satellite positioning mode and candidate location point estimation mode. Satellite positioning mode refers to the working state where trajectory points are generated directly using satellite positioning data at the current moment, typically occurring when the signal quality meets preset requirements in step S101. Candidate location point estimation mode refers to the working state where trajectory points are generated at the current moment through the candidate location point recursion, weight allocation, and weighted averaging method described in steps S103 to S104, typically occurring when the signal quality is below a preset threshold. The mobile terminal internally maintains a current positioning mode identifier, which is recorded synchronously during trajectory point generation.

[0042] The operating status includes foreground and background operating states. Foreground operating state refers to the inspection terminal application being visible and interactive to the user. In this state, the application is not significantly restricted by the operating system's resource scheduling policies and can execute sensor data acquisition and computation tasks with higher priority. Background operating state refers to the application being switched to the background or the screen being off. In this state, the operating system may limit the application's CPU usage, sensor data acquisition frequency, network access permissions, etc., to reduce power consumption. The mobile terminal detects and records the current operating status in real time by calling the lifecycle callback interface provided by the operating system.

[0043] When the positioning mode is satellite positioning mode and the application is in the foreground running state, the confidence level approaches the upper limit (i.e., 1) based on the confidence level of the previous moment. Specifically, the current confidence level equals the confidence level of the previous moment plus an increment, which is the product of the preset recovery rate and the distance between the confidence level of the previous moment and the upper limit. In this state, due to the high accuracy of satellite positioning and the fact that the application is in the foreground and not limited by resources, the confidence level gradually recovers to the highest value at a relatively fast rate.

[0044] When the positioning mode is candidate location point estimation mode and the running state is in the foreground running state, the confidence level decays downwards over time based on the confidence level at the previous moment. Specifically, the current confidence level is equal to the confidence level at the previous moment multiplied by a decay factor that decreases exponentially with the extension of the estimation duration. In this state, due to the cumulative error of the estimation mode that relies on inertial measurement and route geometric constraints, the confidence level gradually decreases over time.

[0045] When the application is running in the background, regardless of the current location mode, additional attenuation is applied on top of the corresponding rules. Specifically, the confidence level in the background is equal to the confidence level calculated according to the foreground rules multiplied by an accelerated attenuation coefficient, which is greater than 0 and less than 1. During background operation, the operating system imposes stricter resource constraints on the application, and sensor data acquisition may be delayed or lost; therefore, the confidence attenuation rate needs to be accelerated.

[0046] When the inspection task begins and there is no trajectory point from the previous moment, the initial confidence value is set to the preset initial value.

[0047] In one embodiment of this application, the mobile terminal presets a confidence threshold. After calculating the confidence level of the current trajectory point, the confidence level value is compared with the confidence threshold. If the confidence level is greater than or equal to the threshold, the current trajectory point is marked as a confident state; if the confidence level is less than the threshold, the current trajectory point is marked as a low-confidence state. The marked state information, along with the trajectory point coordinates and confidence level, is stored for subsequent inspection completion recording and visualization rendering by the management terminal.

[0048] This also includes dynamically adjusting the sampling frequency of the mobile terminal based on the confidence level. The mobile terminal has a pre-set first threshold and a second threshold, with the first threshold being greater than the second threshold. When the confidence level is higher than the first threshold, it indicates that the current positioning accuracy is high and stable. In this case, the satellite positioning data acquisition interval is reduced, and the inertial measurement data sampling frequency is also reduced to save device power. When the confidence level is lower than the second threshold, it indicates that the current positioning reliability is low, requiring more intensive observation data to improve positioning. In this case, the satellite positioning data acquisition interval is increased, and the inertial measurement data sampling frequency is also increased. When the confidence level is between the first and second thresholds, the current sampling frequency remains unchanged.

[0049] S106. After determining that the movement has stopped, based on all the track points that have been marked as reliable, calculate the coverage of the key sections of the preset inspection route during the movement process, and generate an inspection completion record.

[0050] Specifically, the mobile terminal continuously monitors the execution status of the inspection task. When the conditions for stopping the movement are met, such as the motion sensor detecting that the acceleration is below the stationary threshold for a continuous preset period of time, or receiving an instruction from the inspector to manually end the task, the inspection process is determined to be over.

[0051] In practice, the mobile terminal acquires key sections of the preset inspection route. These key sections fall into two categories: one is the key inspection points along the preset route, such as mandatory checkpoints, equipment inspection points, and risk monitoring points; the other is the connecting sections between adjacent key inspection points, i.e., the path from one key point to the next. Each key section is configured with a tolerance radius in the system, which is used to determine whether the trajectory points effectively cover the section.

[0052] The mobile terminal iterates through all trajectory points marked as "trusted". For each critical road segment, it checks whether there exists at least one trusted trajectory point whose shortest distance to the critical road segment is less than or equal to the preset tolerance radius of the road segment. If it exists, the critical road segment is determined to be covered; if it does not exist, the critical road segment is determined to be uncovered.

[0053] Count the number of all covered critical road segments, calculate the ratio of this number to the total number of critical road segments, and use this ratio as the coverage of the preset inspection route during this inspection. Coverage can be expressed as a percentage.

[0054] The mobile terminal generates an inspection completion record based on the coverage data. The completion record includes at least the coverage percentage, a list of covered key road segments, a list of uncovered key road segments, and the corresponding coverage determination distance information. The mobile terminal can temporarily store this record locally and upload it to the management terminal when the network is connected, allowing managers to conduct inspection quality assessments, performance evaluations, and arrange supplementary inspections for uncovered road segments.

[0055] It should be noted that this percentage is displayed on the management interface, and the key road sections that are not covered are also output in list form for managers or inspectors to conduct supplementary inspections.

[0056] In an optional embodiment of this application, the mobile terminal encapsulates the coordinates, confidence level, and corresponding timestamp of the current trajectory point, along with the raw satellite positioning observations and inertial measurement data collected within the time window, into a spatiotemporal capsule data unit. This spatiotemporal capsule is a self-contained data structure containing all the information required to complete a positioning calculation. The spatiotemporal capsule is stored in the terminal device's local cache queue, which employs a first-in, first-out (FIFO) strategy. When the length of the local cache queue exceeds a preset capacity threshold, the earliest generated spatiotemporal capsule at the head of the queue is automatically deleted to prevent storage overflow. When the terminal device detects that a network signal has been restored, it uploads the spatiotemporal capsules that have not yet been uploaded from the local cache queue to the cloud in batches according to chronological order. This mechanism ensures that data collected in weak or no-network environments is not lost, making offline inspection possible.

[0057] Based on the spatiotemporal capsule uploads, as a preferred implementation, after receiving spatiotemporal capsules uploaded from one or more terminal devices, the cloud performs fusion processing on multiple spatiotemporal capsules from different terminal devices within a preset time window. The confidence level and timestamp of each spatiotemporal capsule are extracted, and the fusion confidence level is calculated. The fusion confidence level is equal to one minus the product of the disconfidence levels of all individual spatiotemporal capsules, where the disconfidence level of an individual spatiotemporal capsule is equal to one minus its confidence level. The physical meaning of this fusion formula is that the probability of multiple independent observation sources simultaneously claiming low confidence is extremely low; therefore, as long as there is even one high-confidence observation source, the fusion confidence level will approach 1. If the fusion confidence level is higher than a preset fusion threshold, the trajectory points corresponding to all spatiotemporal capsules within that time window are merged into a single high-confidence positioning point and included in the security situation analysis; otherwise, it is marked as a low-confidence segment and manual review is requested. Through multi-source fusion, the impact of single-terminal positioning errors on the overall security situation assessment is effectively reduced.

[0058] After marking the credibility status of the current trajectory point based on the comparison result between the confidence level and the preset credibility threshold, in a preferred implementation, a real-time alarm is triggered on the terminal device when the confidence level is lower than the preset credibility threshold. The alarm includes at least one of vibration, sound, or notification bar message. Simultaneously, the alarm provides the inspector with operational suggestions, including slowing down the travel speed, switching the terminal device to foreground operation, or moving to an open area to restore satellite signal. When the confidence level remains below the preset credibility threshold for a preset duration, the current trajectory segment is marked as a "forced low-credibility segment" and saved to the local cache for later review. This mechanism can promptly remind inspectors of positioning anomalies, improving the effectiveness of inspection task execution.

[0059] This application deeply integrates the geometric matching constraints of the preset inspection route with the inertial estimation of pedestrian gait. When the satellite signal quality is lower than the threshold, it automatically switches to the candidate location point recursion and weight correction mode, so that the mobile terminal can still continuously generate a smooth trajectory along the preset route even when the screen is off, running in the background or in a weak network environment. This completely eliminates the trajectory breakpoints and perception blind spots caused by positioning interruption in traditional inspection schemes, and realizes continuous trajectory tracking throughout the entire process.

[0060] Meanwhile, this application uses a confidence dynamic model to mark the track points with a credible state, and generates inspection completion records based on the credible track points to quantify the coverage of key road sections. This upgrades discrete check-in points to continuous road section coverage assessment, significantly improving the completeness of the park's security situation awareness and the quantifiability of the quality of inspection task completion, and providing an objective basis for anomaly tracing and performance evaluation.

[0061] like Figure 2 As shown in the figure, this application also proposes a park inspection device based on gait prediction and route matching, including: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform a park inspection method based on gait prediction and route matching as described in any of the above embodiments.

[0062] This application also provides a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as: a park inspection method based on gait prediction and route matching as described in any of the above embodiments.

[0063] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0064] The devices and media provided in this application are one-to-one with the methods. Therefore, the devices and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0065] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0066] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0067] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0068] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0069] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0070] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0071] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0072] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0073] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A yard inspection method based on gait prediction and route matching, characterized in that, Applied to a mobile terminal, the method includes: Receive inspection tasks, load the corresponding inspection task routes, and acquire satellite positioning data in real time while traveling along the inspection task routes, and monitor whether the signal quality of the satellite positioning data meets the preset signal requirements. When the signal quality meets the preset signal requirements, real-time inertial measurement data during the movement is collected to determine real-time gait parameters; the gait parameters include at least stride length, heading angle, and gait timestamp. Based on real-time step size and real-time heading angle, multiple candidate positions of the current estimated trajectory point are recursively predicted, and the vertical distance of each candidate position point from the preset inspection route is calculated. Based on the vertical deviation distance, a corresponding position weight is assigned to each candidate position point. Based on the position weight, a new candidate position point for the estimated trajectory point is generated. The weighted average coordinates of the new candidate position points are used as the current trajectory point at the current moment. Based on the current positioning mode and current operating status, calculate the confidence level of the current trajectory point, compare the confidence level with a preset confidence threshold, and mark the confidence status of the current trajectory point based on the comparison result; Once the movement is determined to have stopped, the coverage of key sections of the preset inspection route by the movement process is calculated based on all the track points that have been marked as reliable, and an inspection completion record is generated.

2. The garden inspection method based on gait prediction and route matching according to claim 1, characterized in that, During the movement along the inspection route, real-time acquisition of satellite positioning data and monitoring of whether the signal quality of the satellite positioning data meets preset signal requirements specifically include: The motion speed is collected in real time by the built-in motion sensor to determine the motion acceleration; When the acceleration exceeds a preset threshold, it is determined that the movement process has begun and satellite positioning data is continuously acquired. Signal quality parameters are extracted from the satellite positioning data, and the signal quality parameters are compared one by one with the corresponding preset signal requirement thresholds; the signal quality parameters include at least one of the following: the number of satellites participating in the positioning, the horizontal positioning error radius, and the accuracy attenuation factor; If the number of satellites is not less than a preset number threshold, the horizontal positioning error radius is not greater than a preset error threshold, and the accuracy attenuation factor is not greater than a preset attenuation threshold, then the signal quality is determined to meet the preset signal requirements; otherwise, it is determined not to meet the requirements.

3. The park inspection method based on gait prediction and route matching according to claim 1, characterized in that, The method of recursively predicting multiple candidate locations of the current trajectory point based on real-time step size and real-time heading angle specifically includes: Check if the candidate location point set generated in the previous time step exists in the local cache; If it does not exist, the most recently output trajectory point before the current time is used as the reference position. Based on the preset Gaussian distribution rule, several initial candidate position points are generated within a preset range of the reference position, and each initial candidate position point is assigned an equal initial weight to obtain the candidate position point set at the current time. If they exist, for each candidate position point in the candidate position point set of the previous time step, the new coordinates of each candidate position point after moving are calculated with the real-time step size as the moving distance and the real-time heading angle as the moving direction. All the new coordinates after moving are used as the predicted candidate position points of the current time step, thus obtaining the candidate position point set of the current time step.

4. The park inspection method based on gait prediction and route matching according to claim 1, characterized in that, The step of assigning corresponding position weights to each candidate position point based on the vertical deviation distance, and generating new candidate position points for the estimated trajectory point based on the position weights, specifically includes: Based on the vertical deviation distance, a position weight is calculated for each candidate position point using a preset distance-weight mapping function; the distance-weight mapping function is in the form of exponential decay, where the smaller the vertical deviation distance, the higher the weight, and the larger the vertical deviation distance, the weight approaches zero exponentially. The position weights of all candidate positions are normalized, and the candidate positions are resampled according to the normalized weight distribution. The range with higher weights generates more candidate positions, and the range with lower weights generates fewer hypothetical positions. All the new candidate locations obtained after resampling are used as the candidate location set at the current time.

5. The park inspection method based on gait prediction and route matching according to claim 1, characterized in that, The step of calculating the confidence level of the current trajectory point based on the current positioning mode and the current operating state specifically includes: Determine the current positioning mode and current operating status; the positioning mode includes satellite positioning mode and candidate location point estimation mode, and the operating status includes foreground operating status and background operating status; When the current positioning mode is satellite positioning mode and the current running state is foreground running state, the confidence level of the previous moment is restored upward by one step value to obtain the confidence level of the current moment; When the positioning mode is candidate location point estimation mode and the running state is foreground running state, the confidence level of the previous moment is modified based on the attenuation factor to obtain the confidence level of the current moment. When the running state is the background running state, based on the confidence level determined in the corresponding foreground mode, additional attenuation is performed according to a preset acceleration attenuation ratio to obtain the confidence level at the current moment.

6. The park inspection method based on gait prediction and route matching according to claim 5, characterized in that, After marking the confidence status of the current trajectory point based on the comparison result, the method further includes: Adjust the terminal sampling frequency according to the confidence level value; When the confidence level is higher than the first confidence threshold, the sampling interval of the satellite positioning data is set to the first sampling period, and the sampling frequency of the inertial measurement data is set to the first frequency. When the confidence level is lower than the second confidence threshold, the sampling interval of the satellite positioning data is switched to the second sampling period, and the sampling frequency of the inertial measurement data is switched to the second frequency, wherein the second sampling period is less than the first sampling period, and the second frequency is higher than the first frequency; When the confidence level is between the second confidence threshold and the first confidence threshold, the current sampling frequency remains unchanged.

7. A park inspection method based on gait prediction and route matching according to claim 6, characterized in that, After adjusting the terminal sampling frequency based on the confidence level, the method further includes: The coordinates, confidence level, and corresponding timestamp of the current trajectory point are encapsulated into a local data packet and stored in a local cache queue. When network connectivity is detected, the local data packets in the local cache queue are uploaded to the management terminal in batches according to time sequence; After the management terminal receives local data units from multiple mobile terminals, it merges multiple local data packets within the same time window and calculates the fusion confidence score based on the confidence score corresponding to each local data packet. If the fusion confidence is higher than the preset fusion threshold, then the trajectory points corresponding to each local data unit within the corresponding time window will be merged into a high-confidence positioning point. If the fusion confidence level is lower than the preset fusion threshold, it is marked as a low-confidence segment.

8. The park inspection method based on gait prediction and route matching according to claim 1, characterized in that, The calculation of the coverage of key road sections in the preset inspection route by the travel process based on all marked as reliable trajectory points specifically includes: Obtain a set of pre-marked key road segments in the preset inspection route, and configure a tolerance radius for each key road segment; For each critical road segment, traverse all track points that have been marked as trustworthy and calculate the shortest distance from each trustworthy track point to the critical road segment; If there exists at least one reliable trajectory point whose shortest distance to the critical road segment does not exceed the tolerance radius, then the critical road segment is determined to be covered. The coverage rate is obtained by calculating the ratio of the number of covered key road segments to the total number of key road segments, and an inspection completion record is generated based on the coverage rate.

9. A park inspection device based on gait prediction and route matching, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform a park inspection method based on gait prediction and route matching as described in any one of claims 1 to 8.

10. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are configured to execute a park inspection method based on gait prediction and route matching as described in any one of claims 1 to 8.

Images