Mountain collapse prediction method and collapse prediction system

Abstract

The invention provides a mountain collapse prediction method and a mountain collapse prediction system, and relates to the technical field of mountain collapse prediction. The method comprises the following steps: acquiring animal data of a target mountain area, screening out target resident animals, determining a monitoring site according to a daily activity range of the target resident animals, and performing monitoring. Historical and real-time daily activity video data of target resident animals are collected, behavior features are extracted and compared to obtain distinguishing features, and the distinguishing features are matched to obtain matching scores; when the matching score exceeds a set threshold value, a geological sensor is controlled to collect geological data, if the data is abnormal, it is determined that mountain collapse is about to occur, and early warning information is generated in combination with the distinguishing features. Early warning is realized by utilizing the sensitivity of animals to geological changes, and accurate judgment is performed through geological data.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mountain collapse prediction, and in particular to a mountain collapse prediction method and a collapse prediction system. BACKGROUND

[0002] At present, mountain collapse poses a serious threat to human life and property and the ecological environment, and efficient and accurate prediction technology has become the key to disaster prevention and mitigation.

[0003] At present, the most common existing technology for predicting mountain collapse is to deploy geological sensors in mountainous areas to monitor physical parameters such as displacement, stress, and crack width of the mountain in real time, and to issue a mountain collapse warning when the monitoring data exceeds the threshold.

[0004] However, this existing technology has a significant technical problem, i.e., the change in monitoring data usually occurs in the approach phase of mountain collapse, and the warning time window is too short to meet the needs of early evacuation and prevention, making it difficult to effectively implement disaster prevention and mitigation measures and fundamentally reduce the losses caused by mountain collapse. SUMMARY

[0005] The present application provides a mountain collapse prediction method and a collapse prediction system for accurately identifying potential risk areas of mountain collapse by comprehensively analyzing changes in the behavior of target resident animals and geological structure characteristics of the mountain, obtaining key geological data in a timely manner, and achieving early warning of mountain collapse.

[0006] In a first aspect, the present application provides a mountain collapse prediction method, which comprises: obtaining animal data of a target mountainous area, and screening the animal data to determine a plurality of different target resident animals, the animal data at least including the daily activity range of the animals; determining monitoring points based on the daily activity range corresponding to each target resident animal, and monitoring each target resident animal based on the monitoring points; obtaining historical daily activity video data of each target resident animal within a set time, and extracting the historical behavior characteristics of each target resident animal corresponding thereto, obtaining real-time daily activity video data of each target resident animal, and extracting real-time behavior characteristics corresponding to each target resident animal; comparing the real-time behavior characteristics of each target resident animal with the historical behavior characteristics to determine distinguishing characteristics; matching the distinguishing characteristics of each target resident animal to obtain a matching score; if the matching score is greater than a set matching score threshold, controlling a plurality of pre-set geological sensors to collect geological data; if the geological data is abnormal, determining that a mountain collapse is about to occur, and generating abnormal animal behavior warning information according to the distinguishing characteristics.

[0007] By adopting the technical scheme, the target mountain region animal data is acquired and the target resident animals are screened, the monitoring points are determined according to the animal daily activity range, the historical and real-time behavior characteristics are compared, and the animal abnormal behavior can be captured in advance. When the matching score exceeds the threshold, the geological sensor is started, and the early warning information is generated when the abnormality occurs. The process builds a double monitoring system of "animal behavior preliminary screening-geological data verification", and realizes early warning by using the sensitivity of animals to geological changes, and accurately judges by geological data. Compared with a single monitoring method, the timeliness and accuracy of mountain collapse prediction are significantly improved, and the disaster risk is effectively reduced.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the animal data of the target mountain region is acquired, and the animal data is screened to determine a plurality of different target resident animals, including: acquiring the infrasound wave perception sensitivity of each animal in the animal data and the activity range curve of each animal; calculating the perception weight of the infrasound wave perception sensitivity and the habitat stability weight of the activity range curve; and selecting the top N animals as the target resident animals in combination with the perception weight and the stability weight.

[0009] By adopting the technical scheme, in the screening of the target resident animals, the infrasound wave perception sensitivity and the activity range curve are comprehensively considered, the perception weight and the stability weight are calculated, and the animals are selected. Animals with high infrasound wave perception sensitivity can detect infrasound waves generated by geological activities earlier, and animals with stable activity range can avoid interference in monitoring due to migration and other factors. In combination with the two weights for screening, it can be ensured that the selected animals are sensitive to geological changes and can provide stable monitoring data for a long time, thereby improving the reliability and effectiveness of the monitoring data from the source, and laying a solid foundation for accurately predicting mountain collapse through animal behavior subsequently.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the different characteristics of each target resident animal are matched to obtain a matching score, including: acquiring a time sequence and a space sequence of each different characteristic; for any two animals in each target resident animal, performing the following operations: based on the time sequence, calculating a time interval of a first target resident animal and a second target resident animal, and based on the space sequence, constructing a first activity track of the first target resident animal and a second activity track of the second target resident animal; comparing the time interval with a preset time window threshold to determine a time synchronization score; calculating the overlap area and the spatial position deviation of the first activity track and the second activity track to determine a spatial consistency score; based on the time synchronization score and the spatial consistency score, calculating a behavior correlation score of the first target resident animal and the second target resident animal; and integrating the behavior correlation scores between all pairs of animals in each target resident animal to obtain the matching score.

[0011] By adopting the technical scheme, in the calculation of the matching score, the time and space sequences of the distinguishing features are obtained, the time synchronization score and the space consistency score of the animals are analyzed, and the matching score is integrated. The time synchronization score determines whether the abnormal behaviors of the animals occur simultaneously, and the space consistency score measures the correlation degree of the activity trajectories. The behavior correlation is evaluated by comprehensively considering the two, which can effectively exclude the accidental behaviors of the individual animals, accurately determine whether the abnormal behaviors of the animals are caused by the geological activities through cross verification of the behaviors of the animals, avoid misjudgment, and greatly improve the accuracy and reliability of the mountain collapse warning.

[0012] In combination with some embodiments of the first aspect, in some embodiments, before the step of controlling the plurality of pre-set geological sensors to collect geological data, the method further comprises: obtaining mountain geological structure data of the mountain area; performing density clustering analysis on the spatial sequence to construct an aggregation area of the abnormal behaviors of the animals; determining an abnormal behavior area of the animals based on a spatial boundary of the aggregation area; performing spatial superposition analysis on the abnormal behavior area of the animals and the mountain geological structure data to determine a geological structure weak zone corresponding to the abnormal behavior area of the animals; and determining position information of a collapse risk area according to the geological structure weak zone.

[0013] By adopting the technical scheme, before starting the geological sensors, the mountain geological structure data is obtained, the clustering analysis on the spatial sequence of the abnormal behaviors of the animals is performed to determine the aggregation area and the abnormal area, and the superposition analysis on the geological structure is performed to determine the weak zone and the collapse risk area. The geological structure data provides inherent risk information of the mountain, the abnormal behaviors of the animals reflect real-time changes, and the combination of the two can accurately locate the area where the geological structure is weak and the abnormal behaviors of the animals occur, reduce the monitoring range, make the subsequent geological sensor monitoring more targeted, and improve the monitoring efficiency and prediction accuracy.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the position information of the collapse risk area according to the geological structure weak zone, the method further comprises: obtaining two target resident animals with the highest scores in the behavior correlation scores; obtaining activity trajectories corresponding to the two target resident animals, and determining a danger source area according to the activity trajectories; determining an overlapping area of the position information and the danger source area; and controlling a geological sensor located in the overlapping area to collect geological data.

[0015] By adopting the technical scheme, after the position information of the collapse risk area is determined, the two animals with the highest behavior correlation scores are selected, the danger source area is determined according to the activity trajectories of the two animals, the overlapping area with the position information is found out, and the sensor in the area is controlled to collect data. The behaviors of the animals with high scores are more representative, the danger source area determined by backtracking of the activity trajectories points to a potential collapse starting point, and the key monitoring area is accurately locked after the overlapping with the risk area, so that the geological sensor focuses on the key position to collect data, reduces invalid monitoring, and improves the data effectiveness and prediction efficiency.

[0016] In some embodiments of the first aspect, in some embodiments, the determining the dangerous source area according to the activity trajectory comprises: performing time synchronization and spatial coordinate calibration on the activity trajectory data to generate a three-dimensional motion path with timestamps; performing trajectory backtracking in the reverse direction of the motion path to construct a set of reverse motion vectors; calculating a convergence area of each reverse motion vector in a three-dimensional space, the convergence area being an area where the spatial distribution density of the intersection points of the reverse vectors exceeds a preset distribution density threshold; and determining the convergence area as the dangerous source area.

[0017] By adopting the above technical solution, when determining the dangerous source area, the activity trajectory is time-synchronized and spatially calibrated to generate a three-dimensional path, and then the vector set is constructed by backtracking in the reverse direction, and the convergence area is calculated. The time synchronization and spatial calibration ensure the accuracy of the trajectory data, the reverse backtracking simulates the escape direction of the animal, and the convergence area is the source of the animal's escape, reflecting the potential collapse starting position. This method accurately locates the dangerous source based on animal behavior logic, provides accurate basis for subsequent targeted monitoring and early warning, and enhances the reliability of prediction.

[0018] In some embodiments of the first aspect, in some embodiments, if the geological data is abnormal, it is determined that a mountain collapse will occur, comprising: obtaining a set of geological data features of historical mountain collapses; matching the geological data with the set of geological data features to obtain a mountain collapse matching score; and if the mountain collapse matching score is greater than a set mountain collapse matching score threshold, it is determined that the geological data is abnormal.

[0019] By adopting the above technical solution, when determining whether the geological data is abnormal, the real-time data is matched with the set of geological data features of historical mountain collapses to obtain a matching score, and if the matching score exceeds the threshold, it is determined to be abnormal. The historical feature set accumulates a large amount of collapse data rules, and the matching of new data with the historical feature set can quickly determine whether the current geological condition conforms to the collapse feature mode, avoiding subjective judgment. Through quantitative analysis of data, the objectivity and accuracy of the geological data abnormality judgment are improved, and the scientificity and reliability of the mountain collapse prediction result are further ensured.

[0020] In a second aspect, the present application provides a collapse prediction system, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors invoke the computer instructions to make the collapse prediction system execute the method as described in the first aspect and any possible implementation manner of the first aspect.

[0021] In a third aspect, the present application provides a computer-readable storage medium comprising instructions that, when executed on a collapse prediction system, cause the collapse prediction system to perform the method as described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer program product that, when executed on a collapse prediction system, causes the collapse prediction system to perform the method as described in the first aspect and any possible implementation of the first aspect.

[0023] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Since the target resident animals are screened based on the daily activity range of animals, the real-time behavior characteristics of animals are compared with historical characteristics, the distinguishing characteristics of multiple animals are matched, and the geological sensor verification is linked, the technical problems of short warning time window and high monitoring cost in the prior art are effectively solved, and early and accurate early warning of mountain collapse is realized, thereby reducing the monitoring cost and disaster risk.

[0024] 2. Since the time sequence and spatial sequence based on distinguishing characteristics are adopted, the time synchronization score and spatial consistency score between animals are calculated, and the matching score is integrated, the technical problem that it is difficult to distinguish whether the abnormal behavior of animals is caused by geological activity or accidental factors in the prior art, resulting in high false alarm rate, is effectively solved, and the abnormality caused by geological activity is accurately judged through cross verification of multiple animal behaviors, thereby greatly improving the credibility and accuracy of mountain collapse warning.

[0025] 3. Since the representative animals are screened based on the behavior correlation score, the activity track is traced back to determine the dangerous source area, and the dangerous source area is overlapped with the collapse risk area to accurately control the sensor data collection, the technical problem that the monitoring range of the geological sensor is large, the data effectiveness is low, and it is difficult to locate the key area of collapse in the prior art is effectively solved, and the key monitoring area is focused, the invalid monitoring is reduced, and the data effectiveness and prediction efficiency are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a flowchart of a mountain collapse prediction method in the embodiments of the present application; Figure 2 is a flowchart of a mountain collapse prediction method in the embodiments of the present application; Figure 3 is a flowchart of a mountain collapse prediction method in the embodiments of the present application; Figure 4 is another flowchart of a mountain collapse prediction method in the embodiments of the present application; Figure 5 is a schematic diagram of an entity device structure of a collapse prediction system in the embodiments of the present application. DETAILED DESCRIPTION

[0027] The terms used in the following embodiments of the present application are only for the purpose of describing the specific embodiments and are not intended to be limiting of the present application. As used in the specification and the appended claims of the present application, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "and / or," as used in the present application, refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0028] Hereinafter, the terms "first" and "second" are only for the purpose of description and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, the meaning of "a plurality of" is two or more, unless otherwise specified.

[0029] For the convenience of understanding, the method provided by the present embodiment is described in the following flow. Please refer to Figure 1 is a flowchart of a mountain collapse prediction method in the embodiments of the present application.

[0030] S101, obtaining animal data of a target mountain area, and screening the animal data to determine a plurality of different target resident animals, the animal data at least including the daily activity range of the animals; Among them, "target mountain area" means a specific mountain range selected as the research object of mountain collapse prediction. "Animal data" refers to a collection of various information about animals in the target mountain area, including the species, number, living habits, physiological characteristics, daily activity range, etc. of the animals. "Target resident animals" refers to a plurality of different animals that are determined to survive and move stably in the target mountain area after screening, which will be used as important biological indicators for monitoring mountain collapse. "Daily activity range" is used to represent the geographical area where the animals carry out foraging, drinking, resting, socializing, etc. in normal state every day, for example, wild boars will forage and move in a mountainous area within a few kilometers, and this area is their daily activity range.

[0031] Specifically, when landslide prediction begins, the system's primary task is to gather data on the animals in the target mountain area. The system uses a variety of methods to gather this data. First, it deploys specialized field survey teams equipped with telescopes, infrared cameras, GPS trackers, and other equipment to deeply penetrate the target mountain area. Through long-term observation, investigators use telescopes to record animal species and behaviors; infrared cameras capture images of nocturnal animals; and GPS trackers are attached to some animals to track their movements in real time, thereby obtaining information on their daily ranges. Furthermore, the system integrates existing data resources, including animal census records from local forestry departments, animal research results from universities and research institutions in the area, and monitoring data from wildlife conservation organizations. After collecting animal data, the system enters a screening phase to identify target resident animals. During this screening process, the system not only considers the animals' daily ranges to ensure that their activities are primarily concentrated within the target mountain area, but also assesses their ability to perceive environmental changes. Many animals are highly sensitive to infrasound, tiny vibrations, and other precursor signals of landslides. The system will obtain data on the sensitivity of different animals to infrasound by consulting animal physiological research materials. At the same time, it will analyze the animal's activity trajectories and draw an activity range curve to assess the stability of the animal's habitat. For example, migratory birds that have a wide range of activities and frequently leave the target mountain area are not suitable as target resident animals because their behavior is greatly disturbed by external factors. Animals such as blue sheep that live in the steep rock walls of the target mountain for a long time and are sensitive to environmental changes will be given priority. By comprehensively analyzing various indicators and applying scientific algorithms and evaluation models, the system will eventually select a number of different target resident animals from a large number of animals. These animals will subsequently become important biological indicators for monitoring landslides.

[0032] S102, determining a monitoring point based on the daily activity range corresponding to each target resident animal, and monitoring each target resident animal based on the monitoring point; "Monitoring points" refer to specific locations within the daily activity range of target resident animals, scientifically planned and precisely selected, where various monitoring devices are installed to collect animal behavior data. Examples include monitoring points in the center of areas where animals frequently forage, near nests, and near watering points. "Monitoring" refers to the use of various technical means, such as high-definition cameras, infrared sensors, sound collectors, and radar monitoring equipment, to continuously or periodically observe, record, and collect data on the behavioral dynamics, activity trajectories, and vocal changes of target resident animals, thereby capturing the behavioral characteristics of animals at different times and in different scenarios.

[0033] Specifically, after the target resident animals and their daily activity ranges are determined by the collapse prediction system, the determination of monitoring points begins. This process requires the system to consider multiple factors to ensure that the behavior information of target resident animals can be captured comprehensively and accurately. First, the system deeply analyzes the activity habits and behavior patterns of each target resident animal. For animals with relatively fixed and concentrated activity ranges, such as sheep that feed and rest in fixed areas, the system sets monitoring points in key locations such as the main feeding grassland of the sheep and the resting hillside, ensuring that the feeding, moving, and resting behaviors of the sheep can be clearly recorded. For animals with wide activity ranges and complex activity patterns, such as squirrels that move throughout the entire forest, the system adopts a combination of zoned and key point placement. First, the activity range of the squirrel is divided into multiple sub-regions, and in each sub-region, key locations such as tree-dense areas and nut-growing areas where the squirrel frequently moves are selected to set monitoring points. At the same time, monitoring points are also set at the connecting channels between different sub-regions, forming a comprehensive monitoring network. When determining monitoring points, the system also fully considers the influence of topography, environmental conditions on monitoring effectiveness. In areas with complex terrain and obstructed vision, high-definition cameras with wide-angle shooting capabilities are selected and installed at reasonable angles to expand the monitoring range; in areas with insufficient night light, infrared night vision cameras and infrared sensors are equipped to ensure effective monitoring of animal activity in dark environments; for some sound signal-rich animals, such as birds, sound collectors are set up in their activity areas to record their calls, flying sounds, and other sound information. After the monitoring points are determined, the system installs corresponding monitoring equipment at these points and conducts strict debugging and calibration work. For example, the focal length, shooting angle, and picture clarity of the camera are adjusted; the sensing range and sensitivity of the infrared sensor are tested; the volume and sound quality of the sound collector are optimized, etc., to ensure that all monitoring equipment can operate stably and accurately. After the equipment installation and debugging are completed, the system monitors each target resident animal based on these monitoring points, collects animal behavior data at set time intervals or in real time, and transmits the obtained video, image, sound, and other data to the data processing center, providing detailed basic materials for subsequent analysis of the relationship between animal behavior and mountain collapse.

[0034] In some embodiments, in the scenario of real-time monitoring of the monitoring area in daily life, the monitoring device can not be turned on all the time, but can be turned on only when it is determined that there is really abnormal activity of the animal, so as to save resources. Specifically, the system detects whether there is an animal heat source signal in the preset surrounding environment through a plurality of infrared sensors arranged in advance at the monitoring point position in real time. The infrared sensor continuously scans the surrounding environment, and when an infrared radiation conforming to the characteristics of the animal heat source is sensed, the subsequent sound collection process is triggered. When the animal heat source signal is detected, the system controls the sound sensor to collect target sound data. The sound sensor starts to work under the trigger of the signal detected by the infrared sensor, and collects the sound of the surrounding environment. The collected target sound data is transmitted to the analysis module of the system, and is compared with the sound frequency characteristics of the target resident animal. If the target sound data conforms to the sound frequency characteristics of the target resident animal, it means that the target resident animal may appear in the monitoring area, and the system triggers the monitoring device corresponding to the monitoring point to collect video data of the target resident animal. After the monitoring device is started, the system identifies the video data collected within the set starting time to determine whether the animal corresponding to the animal heat source signal is the target resident animal. This identification process uses image recognition technology to analyze and judge the animal shape and behavior characteristics in the video. If the identification result is yes, it is confirmed that it is the target resident animal, and the system continues to control the monitoring device to collect video data for further observation of the behavior change. If the identification result is no, it means that the animal corresponding to the detected heat source signal is not the target resident animal, and in order to save resources, the system controls the monitoring device to stop working. In addition, when the target resident animal leaves the set monitoring range or no animal heat source signal is detected within the preset waiting time, the system also controls the monitoring device to stop working. This is because in these cases, it is unnecessary to continue monitoring the area, and stopping the device can effectively reduce the energy consumption and resource occupation of the system. The infrared sensor represents a device capable of sensing infrared radiation and converting it into an electrical signal, which is used to detect changes in heat in the surrounding environment to discover possible animal heat source signals; the animal heat source signal refers to a specific infrared radiation pattern formed by the heat generated by the animal body; the sound sensor is a device for collecting sound signals and converting them into electrical signals, which can be set in advance near the monitoring point. Implementing the above can not only obtain animal abnormal behavior data in time to provide a basis for mountain collapse prediction, but also avoid resource waste, significantly improve the monitoring efficiency and accuracy of the system, and build a data collection defense line for geological disaster warning.

[0035] S103, acquiring historical daily activity video data of each target resident animal within a set time, and extracting corresponding historical behavior characteristics of each target resident animal, acquiring real-time daily activity video data of each target resident animal, and extracting real-time behavior characteristics corresponding to each target resident animal; wherein, "set time" represents a time period artificially determined in advance, used to define the time range of collecting historical daily activity video data, which can be flexibly set according to actual monitoring needs, for example, it can be one month, half a year or one year, etc., the purpose is to obtain the regular behavior pattern data of the target resident animal in a relatively stable period. "Historical daily activity video data" refers to the video information of animal daily activity recorded by the equipment (such as high-definition camera, infrared camera, etc.) deployed on the monitoring point in the daily activity range of the target resident animal within the set time. These videos record the animal's activities in different time periods and different scenes in detail, such as recording the video clips of wild boars foraging, resting and moving in the mountains in the past half year. "Historical behavior characteristics" are used to represent a series of characteristic information that can reflect the regular behavior pattern of the target resident animal within the set time, which is extracted from the historical daily activity video data through specific analysis techniques and algorithms, including the activity rule of the animal (such as activity time, frequency), moving track, behavior type (foraging, resting, socializing, etc.), behavior duration, etc. "Real-time daily activity video data" refers to the real-time video information of the target resident animal's daily activity being performed, which is collected by the monitoring equipment at the current moment, for example, the real-time picture of the squirrel jumping and foraging in the forest taken by the camera at this moment. "Real-time behavior characteristics" refers to the characteristic information reflecting the current behavior state of the target resident animal, which is extracted from the real-time daily activity video data. These characteristic information reflects the animal's current behavior performance.

[0036] Specifically, when the collapse prediction system completes the setting of the monitoring points and ensures the normal operation of the equipment, it begins to continuously collect the daily activity video data of the target resident animals. For the acquisition of historical daily activity video data, the system regularly organizes and filters the video data stored in the database according to the pre-set time period. For example, if the set time period is three months, the system will automatically aggregate all the video data of the target resident animals captured by each monitoring point within the period at the end of every three months. Then, the system uses advanced image recognition, computer vision, and deep learning algorithms to analyze these historical video data in depth. Taking a deer herd as an example, the system determines the daily activity route of the deer herd by analyzing the movement trajectory of the deer herd in the video using trajectory tracking algorithms; by counting the duration and frequency of behaviors such as feeding, drinking, and resting of the deer herd, it constructs the behavior pattern characteristics of the deer herd within that time period. For the acquisition of real-time daily activity video data, the system uses real-time video streaming technology to transmit the video data captured by the monitoring equipment to the data processing center in the shortest delay. Once the real-time video data is received, the system immediately starts the real-time behavior feature extraction program. Based on deep learning models such as convolutional neural networks (CNN), the system identifies and analyzes animal behavior in real time. For example, when a bird is detected, the system can quickly identify the bird's flight posture, foraging actions, etc., and extract real-time behavior features such as the bird's current flight speed, foraging frequency, flight height, etc. The system continuously acquires and extracts historical and real-time behavior features to provide dynamic and comprehensive data support for subsequent analysis of animal behavior changes, so as to timely detect animal behavior abnormalities and provide important basis for mountain collapse prediction.

[0037] S104, comparing the real-time behavior features of each target resident animal with the historical behavior features to determine the distinguishing features; Among them, "distinguishing features" are used to represent feature information that exists between real-time behavior features and historical behavior features. These differences may exist in multiple dimensions such as time distribution of behavior, behavior mode, activity range, behavior frequency, etc. For example, an animal that originally forages in a fixed area suddenly expands its activity range, or the animal's activity time changes significantly, these changes are all distinguishing features.

[0038] Specifically, after obtaining the real-time and historical behavioral characteristics of a target resident animal, the landslide prediction system conducts a detailed comparative analysis of each characteristic. Regarding activity time characteristics, the system compares the animal's historical daily activity start and end times with the current real-time activity times. For example, if a monkey typically began activity around 6:00 a.m. for the past six months, but recently delayed its activity until 8:00 a.m. for several consecutive days, this significant change in activity time will be flagged as a distinguishing characteristic by the system. Regarding activity range characteristics, the system compares the area covered by the animal's historical activity patterns with its real-time activity patterns. For example, if a wolf's previous activity range was primarily concentrated in a valley, but real-time monitoring reveals that it frequently appears on hillsides farther out from the valley, this expansion of its activity range will also be identified as a distinguishing characteristic. Regarding behavioral patterns, the system analyzes whether there have been changes in the animal's movement patterns or social behavior. For example, if real-time monitoring reveals that some individuals in a herd of antelope, which previously moved in groups, begin to act independently, or if there are abnormalities in the antelope's running posture or jumping frequency, these behavioral changes will also be recorded as distinguishing characteristics. During the comparison process, the system applies statistical methods and machine learning algorithms to quantitatively assess the degree of difference in various features. For subtle changes, the system combines multiple related features for comprehensive judgment, avoiding misjudgments due to minor fluctuations in a single feature. By comprehensively and in-depth comparing real-time and historical behavioral features, the system accurately identifies meaningful distinguishing features. These distinguishing features will serve as important clues for subsequent assessments of landslide risk, providing a key basis for further analysis and decision-making.

[0039] S105, matching the distinguishing features of each target resident animal to obtain a matching score; This step is performed after the system has determined the distinguishing characteristics of each target resident animal. It further assesses the coordination and correlation of abnormal animal behaviors, thereby determining the likelihood of landslide risk. This step occurs during the data processing and analysis phase of the landslide prediction system, where the system comprehensively analyzes the distinguishing characteristics of animal behavior. This step will be explained in detail in steps S401-S406 and will not be repeated here.

[0040] S106, if the matching score is greater than a set matching score threshold, controlling a plurality of pre-set geological sensors to collect geological data; The "matching score" is a quantitative score obtained by matching analysis of the distinguishing features of each target resident animal, which reflects the degree of association and synergy between different target resident animal abnormal behaviors. The "set matching score threshold" is a reference score value set in advance, which is used as a standard for determining whether to start the geological sensor to collect geological data. This threshold is determined based on a large amount of experimental data, historical cases, and expert experience. Alternatively, the matching score threshold can be determined as follows: collect the matching scores of animal abnormal behaviors before past mountain collapse events in the target mountain area and similar geological environments; collect the matching scores of animal abnormal behaviors (such as individual accidental behaviors) that do not occur in collapse but occur, and form a matching score dataset of "collapse-non-collapse" two types of samples. Draw the ROC curve, calculate the true positive rate (correctly identify abnormal behaviors before collapse) and false positive rate (mistakenly determine accidental behaviors as precursors of collapse) under different thresholds, and select the score that has the highest true positive rate and the lowest false positive rate as the initial threshold. For example, if statistics show that 90% of the matching scores before the collapse event exceed 80 points, and 95% of the matching scores of non-collapse cases are less than 80 points, the initial threshold can be set to 80 points. Invite geology disaster experts and animal behavior experts to modify the initial threshold value in combination with regional geological characteristics (such as rock stability, fault distribution) and animal habits (such as sensory sensitivity, activity regularity). For example, for areas with extremely unstable geological structures, the threshold can be appropriately lowered to improve the warning sensitivity; for areas where animal behavior is easily affected by external interference (such as frequent tourist activities), the threshold can be appropriately increased to reduce false positives.

[0041] Specifically, after calculating the matching score, the collapse prediction system compares the matching score with a pre-set matching score threshold. If the matching score is less than or equal to the set matching score threshold, it indicates that the correlation between the abnormal behaviors of the target resident animals is weak, which may be caused by some accidental factors or individual differences. In this case, the system judges that the current mountain collapse risk is low, and does not start the geological sensor to collect geological data, but continues to monitor the animal behavior and waits for the next round of data collection and analysis. However, if the matching score is greater than the set matching score threshold, it indicates that the abnormal behaviors of multiple target resident animals have high coordination and correlation in time and space, which is likely to be caused by changes in the mountain (such as stress changes, rock loosening, etc.) that cause animals to perceive abnormalities and change their behavior. In this case, the system judges that the mountain has a high risk of collapse. In this case, the system will immediately issue an instruction to control multiple geological sensors pre-set in the target mountain area to start collecting geological data. These geological sensors will collect various geological parameters of the mountain in real time according to the pre-set sampling frequency and data transmission method after receiving the system instruction. For example, displacement sensors will monitor the small displacement changes on the surface of the mountain, pressure sensors will detect the stress changes inside the mountain, vibration sensors will capture the vibration signals of the mountain, humidity sensors will measure the soil humidity, etc. The collected data will be transmitted in real time to the data processing center through wireless or wired communication network, so that the system can further analyze these geological data to determine whether the mountain is really about to collapse, thereby providing accurate geological basis for subsequent warning and decision-making.

[0042] In some embodiments, after obtaining the matching score of the target resident animal behavior and the matching score is greater than the set matching score threshold, the scene is the process of accurately positioning the collapse risk area when judging that the mountain has a potential collapse risk. First, the system obtains the mountain geological structure data of the mountain area from geological exploration data, satellite remote sensing data, geographic information system (GIS), etc. These data may include the hardness of the rock, the inclination angle of the rock layer, the underground water distribution, the location of the fault zone, etc. At the same time, the system retrieves the spatial sequence data of the animal abnormal behavior obtained before. Then, the spatial sequence is processed using a density clustering analysis algorithm. Taking the DBSCAN (Density-Based Spatial Clustering Application) algorithm as an example, it will divide the points with connected density in space into a cluster. When processing the spatial sequence of animal abnormal behavior, the algorithm will scan all the spatial position points. When the number of animal abnormal behavior position points in a certain area reaches or exceeds the set density threshold, and the distance between these points is within the set neighborhood radius, these points will be classified into an animal abnormal behavior cluster area. For example, in a certain forest area, multiple target resident animals appear abnormal running, gathering and other behaviors in a short time. After density clustering analysis of the spatial position points of these behaviors, an aggregation area is formed.

[0043] Then, according to the spatial boundaries of the aggregation area, the system expands or integrates adjacent aggregation areas, determining a larger range of animal abnormal behavior area that covers all spatial ranges where animal abnormal behavior is apparent. After that, the system performs spatial overlay analysis of the animal abnormal behavior area and the mountain geological structure data on the GIS platform. By comparing the spatial distribution of the two, the overlapping area of the animal abnormal behavior area and the weak part of the mountain geological structure is found, which is the geological structure weak belt. For example, if the animal abnormal behavior area overlaps with the fault area in the mountain or the weathered rock area, these overlapping parts are identified as the geological structure weak belt. Finally, according to the determined geological structure weak belt, the system combines the topography and other information of the mountain to further clarify the location information of the collapse risk area, accurately to the specific coordinate range or geographical area description, providing accurate basis for subsequent targeted monitoring and early warning. Throughout the process, the accuracy of the data and the rationality of the algorithm are crucial. For mountain geological structure data, it needs to be updated and verified regularly to ensure that it can truly reflect the geological conditions of the mountain; for the density clustering algorithm, the density threshold and neighborhood radius parameters need to be set reasonably according to the actual situation to ensure accurate identification of animal abnormal behavior aggregation area. At the same time, when performing spatial overlay analysis, the coordinate system conversion and precision matching of different data should be considered to ensure the reliability of the analysis results.

[0044] In some embodiments, after determining the location information of the collapse risk area, the system filters out the highest scoring group of animals from all pairwise behavioral correlation scores of target resident animals. For example, assuming there are 5 target resident animals, the pairwise behavioral correlation scores are as follows: goat-squirrel 90 points, goat-hare 85 points, squirrel-hare 78 points, etc. Then, the system retrieves the activity trajectory data of these two animals, which are constructed based on their spatial sequences and may include multiple three-dimensional coordinate points corresponding to time stamps. Then, the system determines the dangerous source area based on the activity trajectory. The specific process may include: time synchronization and spatial coordinate calibration of activity trajectory data to generate time-stamped three-dimensional motion paths, ensuring the time and spatial accuracy of the trajectory; trajectory backtracking in the opposite direction of the motion path to construct a set of reverse motion vectors, for example, an animal moves from point A to point B and then to point C, and reverse backtracking is from point C to point B and then to point A; calculating the convergence area of each reverse motion vector in three-dimensional space, i.e., the area where the spatial distribution density of the reverse vector intersection point exceeds the pre-set distribution density threshold, which is determined as the dangerous source area because the animal may have escaped from this area and exhibited abnormal behavior. For example, the activity trajectories of the goat and the squirrel both show that they quickly move from a certain area on the hillside to the foot of the hill. After reverse backtracking, it is found that their motion vectors converge in a small area on the hillside, which is the dangerous source area. Then, the system spatially compares the location information of the determined collapse risk area with the dangerous source area to find the overlapping area. This overlapping area is both a weak geological structure belt and a risk source that may trigger abnormal animal behavior, so the collapse risk is extremely high. Finally, the system controls the pre-deployed geological sensors in the overlapping area to start data collection. These sensors may include displacement sensors, stress sensors, crack monitors, soil moisture sensors, etc., which will collect real-time geological parameters such as the small displacement of the mountain surface, the stress change inside the rock, the expansion of the crack, and the soil moisture. During execution, the following details need to be noted: when obtaining the behavioral correlation score, ensure the integrity and accuracy of the data to avoid misjudgment of the highest scoring animal due to individual abnormal data; when time synchronizing and spatially calibrating the activity trajectory, consider the time deviation and coordinate system consistency of different monitoring devices, which can be achieved through GPS clock synchronization and coordinate conversion algorithm; when determining the dangerous source area, the pre-set distribution density threshold should be reasonably set based on historical data and animal behavior characteristics, for example, the threshold for social animals may be different from that for solitary animals; when controlling the geological sensors to collect data, ensure that the sensor is in normal working state and the communication link is smooth to ensure the real-time and reliability of the collected data.In addition, if there are multiple groups of animal combinations with similar behavior correlation scores, the system can consider factors such as the sensitivity of animals to infrasound waves and habitat stability to select more valuable animal pairs for analysis, ensuring more accurate positioning of the danger source area.

[0045] S107, if the geological data is abnormal, it is determined that a landslide will occur, and abnormal animal behavior warning information is generated according to the distinguishing features.

[0046] In the formula, "geological data" represents various types of data related to the geological conditions of the mountain collected by the geological sensors pre-installed in the mountain area. These data cover displacement, stress, strain, groundwater level, soil moisture, and other aspects of the mountain, reflecting the current geological state of the mountain. "Abnormal animal behavior warning information" is used to represent warning messages generated according to animal abnormal behavior characteristics, reminding relevant personnel that a landslide may occur. The content generally includes animal abnormal behavior, possible collapse area, and predicted collapse time. For example, in a certain mountain area, the displacement data collected by the geological sensor increased significantly in a short period of time, which is obviously beyond the normal fluctuation range, which is an abnormal geological data. If the animals are restless and migrate in large numbers at this time, the message generated according to these behaviors to remind nearby personnel to evacuate is abnormal animal behavior warning information.

[0047] Specifically, the timing of this step is after analyzing the behavior characteristics of the target resident animals and finding that their behavior is abnormal (i.e., the matching score is greater than the set matching score threshold), and then controlling the geological sensor to collect geological data. The scene is mainly for the target mountain area where a landslide may occur, in order to discover the danger in advance and protect people's lives and property safety and ecological environment stability. In actual operation, first, the historical geological data feature set of mountain collapse is needed, which is obtained by analyzing, summarizing and concluding the geological data collected in the past numerous mountain collapse cases. It contains the typical feature patterns of geological data before different types of mountain collapse.

[0048] Then, the current collected geological data is matched with the geological data feature set. In the matching process, a series of professional data analysis algorithms and models are used. For example, for mountain displacement data, the similarity between the current data and the displacement data in the feature set is calculated, and the similarity is evaluated by comparing the amplitude, speed, trend and other dimensions of the displacement change; for stress data, the distribution of stress, the change rule and the difference degree with the stress data in the feature set are analyzed. Through the comprehensive analysis of various geological data, a mountain collapse matching score is obtained, which can reflect the matching degree of the current geological data and the historical mountain collapse geological data feature set. If the mountain collapse matching score is greater than the set mountain collapse matching score threshold, it can be determined that the current geological data is abnormal, which means that the mountain is in an unstable state and is about to collapse. When it is determined that the mountain is about to collapse, the system will generate abnormal animal behavior warning information according to the distinguished features of the target resident animals obtained before. In generating the warning information, it will specify which abnormal behaviors the animals have, such as sudden collective migration of animals originally active in a fixed area, or some animals showing restlessness and frequent running; at the same time, it will clearly point out the area where the collapse may occur, which is determined by combining the aggregation area of animal abnormal behaviors and the mountain geological structure data; it will also preliminarily estimate the time of the expected collapse, although it is difficult to accurately predict the collapse time, but according to the trend and severity of the geological data change, a rough time range can be given. Finally, these key information is integrated together, and through various channels such as SMS, broadcast, alarm system, etc., timely and accurate information is sent to the personnel in the affected area, so that they can quickly take measures to evacuate to a safe area, and minimize the loss caused by mountain collapse.

[0049] In the embodiments of the present application, a series of technical means are adopted, such as obtaining target mountain area animal data to screen target resident animals, determining monitoring points based on the activity range and monitoring, determining distinguished features by comparing animal real-time and historical behavior characteristics, matching the distinguished features to obtain a matching score, controlling the geological sensor to collect geological data when the matching score exceeds the threshold, and judging the mountain collapse risk according to the geological data, so that the sensitivity of animals to geological changes and the accurate monitoring of geological sensors to the state of the mountain can be fully utilized, the problem that the traditional mountain collapse prediction method is difficult to find mountain abnormalities in early stage and lacks prediction accuracy and timeliness is effectively solved, and the early warning of mountain collapse is realized, which guarantees the safety of life and property and the stability of ecological environment of the personnel around the mountain, and provides strong support for disaster prevention and mitigation work.

[0050] In the following, in combination with the above content, please refer to Figure 2 The geological structure weak belt in the above animal abnormal behavior area is described in detail. Figure 2Fig. 1 is a schematic diagram of an application scenario of the mountain collapse prediction method in the embodiments of the present application.

[0051] Suppose that the target resident animals in the target mountain area are rabbits and birds, and the geological structure data of the mountain area has been acquired in advance, including the information of weak geological structure areas such as fault zones and weathered rock layers. The behaviors of the rabbits and birds are monitored in real time through the monitoring points set in advance. If a geological disaster is about to occur, the rabbits and birds will exhibit obvious abnormal behaviors before the disaster: the rabbits will flee in different directions, but all of them will run away from the place where the disaster is likely to occur, and the birds will fly away in a cluster in the direction of the arrow. The abnormal behaviors of the animals are acquired, and the spatial sequence is formed through the spatial position data of the abnormal behaviors. After the density clustering analysis of the spatial sequence, it is found that the abnormal activities of the rabbits and birds are concentrated in a certain area, forming an animal abnormal behavior aggregation area. Based on the spatial boundary of the aggregation area, the animal abnormal behavior area covering the abnormal activities is determined. Then, the animal abnormal behavior area is superimposed with the geological structure data of the mountain, and finally, in the superimposed result, the area where the animal abnormal behavior area and the weak parts such as fault zones and weathered rock layers in the geological structure data coincide is the weak geological structure zone of the mountain.

[0052] Based on the above Figure 2 , the source of the hazard can be further accurately determined, so as to facilitate more accurate data collection in the future. Please refer to Figure 3 for the determination of the source area of the weak geological structure zone. Figure 3 Fig. 2 is another schematic diagram of an application scenario of the mountain collapse prediction method in the embodiments of the present application.

[0053] In (a) in Figure 3 , the spatial distribution and trajectory distribution of the animal abnormal behaviors in the target mountain area are shown. Through the recorded video, the system can obtain multiple abnormal activity trajectory points (such as L3, L4, A point, C point, etc.) of the target resident animals in a certain time period: the rabbits in the time period escape from the A point to the B point through the path L1 to the safe area, and another batch of rabbits escape from the E point to the F point through the path L2 to the safe area; similarly, the birds escape from the C point to the D point through the path L3 to the safe area, and another batch of birds escape from the G point to the H point through the path L4 to the safe area. The abnormal activities of these animals all occur in Figure 2 the weak geological structure zone obtained, but if the weak geological structure zone is too wide, it is also not conducive to subsequent more accurate geological detection. Therefore, the more accurate source area of the hazard in the weak geological structure zone can be determined according to the method shown in (b) in Figure 3 .

[0054] In Figure 3 In (b) of FIG. 3, the paths of L1, L2, L3 and L4 are normalized to the same coordinate system, which is a two-dimensional plane with x-axis and y-axis as horizontal spatial coordinates, and the z-axis as the vertical height, to form a three-dimensional spatial coordinate system. The reverse motion vectors are extended along the paths respectively, and these lines point to the dangerous source area from different directions. Then, the convergence area of each reverse motion vector in the three-dimensional space is calculated. The definition of the convergence area is the area where the spatial distribution density of the intersection points of the reverse vectors exceeds the preset distribution density threshold. In FIG. 3(b), intersection point 1, intersection point 2 and intersection point 3 are points formed by the intersection of the extended reverse motion vectors. By statistically analyzing a large number of intersection points, the spatial distribution density of the intersection points is calculated. If the distance between two adjacent intersection points of the three intersection points is less than the set distance threshold, it can be determined that the preset range area of the intersection line is the convergence area, or when the intersection point density of a certain area exceeds the preset density threshold, it can be determined that the area is the convergence area. Finally, the convergence area is determined as the dangerous source area. In FIG. 3(b), the area marked by the gray ellipse is the dangerous source area determined by the above analysis. This area concentrates a large number of intersection points of the reverse motion vectors, indicating that the abnormal behavior of the animal is likely to be caused by the geological activity in this area. Comparing this area with the known geological structure weak belt, it is found that the two areas overlap highly, which further verifies the rationality of the area as the dangerous source area.

[0055] After combining the above content, the method provided by the present embodiment is further described in more detail. Please refer to Figure 4 , which is another flowchart of the mountain collapse prediction method in the present embodiment.

[0056] S401, obtain the time sequence and the space sequence of each distinguished feature; The time sequence refers to the sequence formed by arranging the distinguished features in the order of occurrence time, which is used to record the distribution of the distinguished features in the time dimension, such as the abnormal calling of an animal at 8 am and the frequent movement at 9 am. These events are arranged in time sequence to form a time sequence; the space sequence refers to the distribution sequence of the distinguished features in the spatial position, which is used to reflect the specific position and movement path of the abnormal behavior of the animal, such as the position change sequence formed by the process of the animal moving from habitat A to northwest to area B.

[0057] Specifically, the timing of this step is after the comparison of the real-time behavior characteristics and historical behavior characteristics of each target resident animal is completed and the distinguishing features are determined. The scene is the process of analyzing and processing animal behavior data by the collapse prediction system. The system needs to sort each distinguishing feature in time and space dimensions. First, the system will extract the occurrence time point corresponding to each distinguishing feature from the real-time video data obtained by the monitoring device, such as the time of the animal's sudden running, the time of abnormal gathering, etc., and then arrange these time points in chronological order into a time sequence, so that the chronological order and distribution of different distinguishing features in time can be clearly seen. At the same time, for each distinguishing feature, the system will determine the specific location where the feature occurs by combining the spatial coordinate information of the monitoring point, for example, a certain deer appears abnormal wandering at coordinates (X1, Y1) and then moves to (X2, Y2) to stay. These location information connected in time order forms a spatial sequence. In this process, the system needs to ensure the accuracy of the time data, which may involve calibrating the time of the monitoring device to avoid sequence confusion caused by time errors. For the acquisition of spatial position, it may need to rely on GPS positioning technology or other spatial positioning means to ensure the accuracy of the spatial coordinates of each distinguishing feature. In addition, when there are multiple distinguishing features, the system needs to construct the time and space sequences for each feature, for example, for the distinguishing feature 1 and distinguishing feature 2 of the target resident animal A, respectively generate the corresponding time sequence and space sequence, in order to analyze different features. Moreover, when processing the time sequence, the system will also consider the accuracy of the time, whether it is accurate to seconds, minutes or hours, which depends on the collection frequency of the monitoring data and the needs of subsequent analysis. For the spatial sequence, it may involve the conversion of spatial coordinates to unify the coordinates obtained by different monitoring devices to the same coordinate system, to ensure the consistency and comparability of the spatial sequence.

[0058] For any two animals in the target resident animals, the following operations are performed: S402, based on the time sequence, calculate the time interval of the first target resident animal and the second target resident animal, and based on the space sequence, construct a first activity trajectory of the first target resident animal and a second activity trajectory of the second target resident animal; The first target resident animal and the second target resident animal represent two different animals selected from a plurality of target resident animals for pairwise behavior correlation analysis, such as selecting goats and squirrels as analysis objects; the time interval refers to the difference between the time when the distinguishing feature of the first target resident animal occurs and the time when the distinguishing feature of the second target resident animal occurs, for measuring the time sequence of the behaviors of the two, for example, the goat abnormally moves at 8:00, the squirrel abnormally chirps at 8:05, and the time interval is 5 minutes; the first activity track is a movement path constructed based on the spatial sequence of the distinguishing feature of the first target resident animal, which is a direct representation of the abnormal behavior of the first target resident animal in space, such as the path connecting A, B and C points on the hillside; the second activity track is the same, which is a movement path constructed based on the spatial sequence of the distinguishing feature of the second target resident animal.

[0059] Specifically, the timing of this step is after obtaining the time series and spatial series of distinguishing features, and during the process of the system analyzing the pairwise behavioral correlation of target resident animals. The system needs to operate for each pair of target resident animals. First, for the selected first target resident animal and the second target resident animal, the system extracts the occurrence time points of distinguishing features from their respective time series. For example, assume that the first target resident animal has three distinguishing features in the time series, with occurrence times t1, t2, t3, and the second target resident animal has two distinguishing features, with occurrence times t4, t5. The system calculates the time intervals between each time point of the first target resident animal and each time point of the second target resident animal, i.e. |t1-t4|, |t1-t5|, |t2-t4|, |t2-t5|, |t3-t4|, |t3-t5|. When calculating the time intervals, attention should be paid to the uniformity of time, and different sources of time data may need to be converted to a unified standard time format to avoid calculation errors due to time zone, clock error, etc. At the same time, the system records these time intervals for subsequent comparison with the preset time window threshold. Next, based on the spatial series, the activity trajectory is constructed. For the first target resident animal, the system connects the coordinate points in its spatial series in chronological order to form the first activity trajectory. For example, the coordinate points in the spatial series are (x1, y1, z1), (x2, y2, z2), (x3, y3, z3), and the first activity trajectory is the curve or polyline formed by connecting these three points in order. When constructing the trajectory, it may be necessary to smooth the spatial coordinates to eliminate trajectory jitter caused by positioning errors, making the trajectory more accurately reflect the animal's true movement path. For the second target resident animal, the same method is used to construct the second activity trajectory. In addition, when constructing the activity trajectory, the time dimension also needs to be considered to ensure that each point on the trajectory corresponds to an accurate timestamp for subsequent analysis of the time variation of the trajectory. When the activity trajectory of the target resident animal involves three-dimensional space, the system needs to have the ability to model three-dimensional space, including the height information (such as the z-axis) in the spatial coordinates into the trajectory construction, to more comprehensively reflect the animal's activity range. Moreover, for different types of animals, their activity trajectories may have different characteristics, for example, the activity trajectory of a bird may change greatly in the vertical direction, while the trajectory of a ground animal is mainly in the horizontal plane, and the system needs to perform corresponding trajectory processing and analysis according to the characteristics of the animal species.

[0060] S403, comparing the time interval with the preset time window threshold to determine the time synchronization score; The time interval represents the difference between the occurrence times of the distinguishing features of the first target resident animal and the second target resident animal. The preset time window threshold is a time range criterion for judging whether the behaviors of the two animals have time synchronization, for example, set to 10 minutes. When the time interval is within the range, it is considered that the behaviors of the two animals have a certain time synchronization. The time synchronization score is a score determined according to the comparison result of the time interval and the preset time window threshold, which is used to quantify the degree of association of the behaviors of the two animals in the time dimension. The higher the score, the stronger the time synchronization.

[0061] Specifically, the timing of this step is after calculating the time interval of the first target resident animal and the second target resident animal, and during the process of analyzing the time association of the behaviors of the animals by the system. The system needs to evaluate each calculated time interval. First, the system will call the preset time window threshold, which may be obtained from historical data statistical analysis. For example, through the study of a large number of animal behavior time intervals under normal circumstances, it is determined that when the time interval is less than or equal to 10 minutes, the animal behavior may have a certain association. Then, compare each time interval with the threshold. For example, if the time interval is 5 minutes, which is less than the threshold of 10 minutes, it means that the occurrence times of the distinguishing features of the two animals are relatively close, and they have synchronization in time. If the time interval is 15 minutes, which is greater than the threshold, it means that the time synchronization is poor. When determining the time synchronization score, the system may use different scoring rules. One common way is linear scoring, that is, when the time interval is less than or equal to the threshold, the score increases as the time interval decreases. For example, the threshold is 10 minutes, the time interval is 0 minutes, the score is 100 points, the time interval is 10 minutes, the score is 60 points, and the time interval exceeds 10 minutes, the score is 0 points. Another way is segmented scoring, which divides the time interval into different intervals, each interval corresponds to a different score, for example, 0-5 minutes gets 80-100 points, 5-10 minutes gets 60-80 points, and more than 10 minutes gets 0 points.

[0062] The system may also consider the direction of the time interval, that is, whether the behavior of the first target resident animal precedes or follows that of the second target resident animal. In some cases, different time sequences may reflect different behavior association patterns, so different weights may be given when scoring. In addition, when there are multiple time intervals, the system needs to determine how to integrate the scores of these intervals to obtain the final time synchronization score.

[0063] S404, calculate the overlap area and spatial position deviation of the first activity trajectory and the second activity trajectory, and determine a spatial consistency score; wherein the overlap area refers to the size of the region that is jointly covered by the first and second activity trajectories in space, used to measure the degree of coincidence of the two trajectories in spatial position, for example, both trajectories pass through a certain valley region, which is the overlapping part; the spatial position deviation refers to the average distance or deviation between the two trajectories, reflecting their separation degree in space. The distance between two points is a measure of spatial position deviation; the spatial consistency score is a score calculated according to the overlap area and the spatial position deviation, used to evaluate the similarity and relevance of the two trajectories in the spatial dimension, the higher the score, the stronger the spatial consistency.

[0064] Specifically, the timing of this step is after the first and second activity trajectories are constructed, in the process of the system analyzing the spatial behavior relevance of the target resident animals. The system needs to calculate and evaluate the activity trajectories of each pair of target resident animals in detail. First, calculate the overlap area. The system will project the first and second activity trajectories onto the same plane (usually the horizontal ground), and then determine their coverage areas on the plane. This may involve discretizing the trajectories into a series of points, and then determining the coverage range by a polygon area recognition algorithm. For example, for a continuous activity trajectory, it can be divided into multiple line segments, each connecting two adjacent coordinate points, and the area enclosed by these line segments is the coverage range of the trajectory.

[0065] Then, the system finds the intersection part of the two coverage ranges, and calculates the area of the intersection as the overlap area. In the calculation process, the time dimension of the trajectory may need to be considered, for example, only the overlapping part of the two trajectories in the same time period is calculated, because the activity of the animal may change over time. Next, calculate the spatial position deviation. The system will select a series of corresponding points on the two trajectories, usually in time order. Then, calculate the distance between each pair of corresponding points, such as Euclidean distance. Finally, take the average or weighted average of all these distances to get the spatial position deviation. When selecting corresponding points, attention should be paid to time synchronization. If the time sequences of the two trajectories are not completely consistent, time interpolation or sampling may be needed to ensure that reasonable corresponding points can be found. When determining the spatial consistency score, the system will consider the overlap area and the spatial position deviation. A common method is to take the overlap area as a positive indicator and the spatial position deviation as a negative indicator, and calculate the score by weighted summation.

[0066] Another method is to normalize the overlap area and spatial position deviation respectively, and then take the weighted average of the two. In addition, the system may also consider the shape similarity of the trajectory, such as using a shape matching algorithm to evaluate whether the overall shape of the two trajectories is similar, and also include the shape similarity in the calculation of the spatial consistency score. For different types of animal activity trajectories, the system may use different calculation methods. For example, for three-dimensional trajectories of flying animals, it may be necessary to calculate the overlap volume and spatial position deviation in three-dimensional space; for the trajectories of ground animals, the calculation is mainly carried out in a two-dimensional plane. At the same time, the system also needs to consider the density and complexity of the trajectory, and for complex trajectories, it may need to be simplified to improve the efficiency and accuracy of the calculation.

[0067] S405, based on the time synchronization score and the spatial consistency score, calculating the behavior correlation score of the first target resident animal and the second target resident animal; The behavior correlation score is the final score obtained by considering the time synchronization and spatial consistency, which is used to comprehensively evaluate the correlation strength between the behaviors of the two animals.

[0068] Specifically, the timing of this step is after determining the time synchronization score and the spatial consistency score, and in the process of the system comprehensively evaluating the behavior correlation between the target resident animals. The system needs to integrate the scores of these two dimensions. First, the system will determine the weights of the time synchronization score and the spatial consistency score. The setting of these two weights depends on the specific application scenario and research purpose. For example, in some cases, time synchronization may be more important, because the simultaneous occurrence of animal behavior may better reflect their response to the same external stimulus, so the weight of the time synchronization score can be set higher; while in other cases, spatial consistency may be more critical, because the activities of animals in similar spatial areas may indicate that they are affected by the same environmental factors, so the weight of the spatial consistency score can be set higher. A common method to determine the weight is to analyze based on expert knowledge or historical data. For example, through the study of animal behavior before the mountain collapse event in the past, it is found that the contribution of time synchronization and spatial consistency to the prediction of collapse is 6:4, then the weight of the time synchronization score can be set to 0.6, and the weight of the spatial consistency score can be set to 0.4. Then, the system will calculate the behavior correlation score by weighted summation.

[0069] S406, integrating the behavior correlation scores between all pairs of animals in each target resident animal to obtain a matching score.

[0070] Specifically, the timing of this step is after calculating the behavior correlation scores between all pairs of target resident animals, during the process of the system's comprehensive assessment of the behavior of the entire target resident animal group. The system needs to integrate these scores. First, the system will collect the behavior correlation scores between all pairs of animals. Then, the system will integrate these scores using appropriate methods. One common method is to calculate the average. Add the behavior correlation scores between all pairs of animals, then divide by the total number of combinations to get the average behavior correlation score as the matching score. Another method is to calculate the weighted average. According to the importance or reliability of different animal pairs, each behavior correlation score is assigned a different weight. For example, for animal pairs with higher infrasound perception sensitivity, a higher weight may be given because their behavior changes may better reflect potential changes in the mountain. The system may also consider the distribution of behavior correlation scores. For example, calculate the median, mode, or variance of the scores to better understand the distribution characteristics of the scores. If the variance of the scores is small, it means that the behavior correlation between animal pairs is consistent, and the matching score may be higher; conversely, if the variance is large, it means that the behavior correlation varies greatly, and the matching score may be lower.

[0071] In the embodiments of the present application, by obtaining the time series and spatial series of each distinguishing feature, calculating the time interval between animals, constructing activity trajectories, determining the time synchronization score and spatial consistency score, and then calculating the behavior correlation score and integrating the matching score, the correlation and consistency of the behavior of the target resident animal group can be comprehensively and accurately analyzed, effectively solving the problem of accurately determining potential risks through animal behavior in traditional mountain collapse prediction, and achieving the technical effect of early warning of mountain collapse using abnormal animal behavior.

[0072] The collapse prediction system in the embodiments of the present application will be described from the perspective of hardware processing. Please refer to Figure 5 , which is a schematic diagram of an entity device structure of the collapse prediction system in the embodiments of the present application.

[0073] It should be noted that, Figure 5 The structure of the collapse prediction system shown is only an example and should not limit the function and use range of the embodiments of the present application.

[0074] As Figure 5As shown, the collapse prediction system includes a Central Processing Unit (CPU) 501 which can perform various appropriate actions and processes, such as executing the methods described in the above embodiments, according to programs stored in a Read-Only Memory (ROM) 502 or programs loaded from a storage section 508 into a Random Access Memory (RAM) 503. Various programs and data required for operation of the system are also stored in the RAM 503. The CPU 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An Input / Output (I / O) interface 505 is also connected to the bus 504.

[0075] Connected to the I / O interface 505 are an input section 506 including an audio input device, a push button switch, and the like; an output section 507 including a Liquid Crystal Display (LCD), an audio output device, an indicator, and the like; a storage section 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as necessary. A removable media 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is mounted on the drive 510 as necessary, so that a computer program read therefrom is installed in the storage section 508 as necessary.

[0076] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing a computer program for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication section 509, and / or installed from the removable media 511. When the computer program is executed by the Central Processing Unit (CPU) 501, various functions defined in the present application are performed.

[0077] Note that specific examples of computer-readable storage media can include but are not limited to an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, computer-readable storage media can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

[0078] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functional processes, and operational processes, according to various embodiments of the present application. It will be understood that each block of the flow diagrams and / or block diagrams, and combinations of blocks in the flow diagrams and / or the block diagrams, can be implemented by computer readable program instructions such as program code. Such computer readable program instructions can be provided to a processor of a computer, 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, create means for implementing the functions / acts specified in the flow diagrams and / or block diagrams. These computer readable program instructions can also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and the other

[0079] In particular, the collapse prediction system of the embodiment includes a processor and a memory, and the memory stores a computer program. When the computer program is executed by the processor, the mountain collapse prediction method provided in the above embodiment is implemented.

[0080] As another aspect, the present application also provides a computer readable storage medium. The storage medium can be included in the collapse prediction system described in the above embodiments, or can exist independently without being assembled into the collapse prediction system. The storage medium carries one or more computer programs. When the one or more computer programs are executed by a processor of the collapse prediction system, the collapse prediction system implements the mountain collapse prediction method provided in the above embodiments.

[0081] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application; even though the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

[0082] In the above embodiments, the term "when" can be interpreted to mean "if" or "after" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "on determining" or "if detecting (a stated condition or event)" can be interpreted to mean "if determining" or "in response to determining" or "on detecting (a stated condition or event)" or "in response to detecting (a stated condition or event)" depending on the context.

[0083] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by a computer program instructing the relevant hardware to complete, and the program can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. The aforementioned storage medium includes ROM or random storage memory RAM, magnetic disk or optical disk and various storage program codes.

Claims

1. A landslide prediction method, characterized in that: The method comprises: Obtaining animal data in the target mountain area and screening the animal data to determine a plurality of different target resident animals, wherein the animal data at least includes the daily activity range of the animals; Determining monitoring points based on the daily activity range corresponding to each target resident animal, and monitoring each target resident animal based on the monitoring points; Obtaining historical daily activity video data of each target resident animal within a set time, and extracting corresponding historical behavioral characteristics of each target resident animal, obtaining real-time daily activity video data of each target resident animal, and extracting corresponding real-time behavioral characteristics of each target resident animal; comparing the real-time behavioral characteristics of each target resident animal with the historical behavioral characteristics to determine distinguishing characteristics; Matching the distinguishing features of each target resident animal to obtain a matching score; If the matching score is greater than a set matching score threshold, controlling a plurality of pre-set geological sensors to collect geological data; If the geological data is abnormal, it is determined that a landslide is about to occur, and abnormal animal behavior warning information is generated based on the distinguishing features.

2. The method according to claim 1, characterized in that The method of obtaining animal data of the target mountain area and screening the animal data to determine a plurality of different target resident animals includes: Obtaining the infrasound perception sensitivity of each animal in the animal data and the activity range curve of each animal; Calculate the perception weight of the infrasound perception sensitivity and the habitat stability weight of the activity range curve; The top N animals are selected as target resident animals based on the perception weight and the stability weight.

3. The method according to claim 1, characterized in that The matching of the distinguishing features of each target resident animal to obtain a matching score includes: Obtaining a time series and a spatial series of each of the distinguishing features; Perform the following operations on any two of the target resident animals: Based on the time sequence, calculating the time interval between a first target resident animal and a second target resident animal, and constructing a first activity trajectory of the first target resident animal and a second activity trajectory of the second target resident animal based on the spatial sequence; Comparing the time interval with a preset time window threshold to determine a time synchronization score; calculating an overlapping area and a spatial position deviation between the first activity trajectory and the second activity trajectory to determine a spatial consistency score; calculating a behavioral correlation score between the first target resident animal and the second target resident animal based on the temporal synchrony score and the spatial consistency score; The behavioral correlation scores between all pairs of animals in each target resident animal are integrated to obtain the matching score.

4. The method according to claim 1 or 3, characterized in that Before the step of controlling a plurality of pre-set geological sensors to collect geological data, the method further includes: Acquiring mountain geological structure data of the mountain area; Performing density cluster analysis on the spatial sequence to construct clusters of abnormal animal behaviors; Determining areas of abnormal animal behavior based on the spatial boundaries of the gathering area; Performing spatial overlay analysis on the abnormal animal behavior area and the mountain geological structure data to determine the weak geological structure zone corresponding to the abnormal animal behavior area; The location information of the collapse risk area is determined based on the geological structural weak zone.

5. The method according to claim 4, characterized in that After the step of determining the location information of the collapse risk area according to the geological structural weak zone, the method further includes: Obtaining the two target resident animals with the highest scores among the behavioral correlation scores; Obtaining the activity trajectories corresponding to the two target resident animals, and determining the danger source area based on the activity trajectories; Determining an overlapping area between the location information and the danger source area; The geological sensors in the overlapping area are controlled to collect geological data.

6. The method according to claim 5, characterized in that Determining the danger source area according to the activity trajectory includes: Performing time synchronization and spatial coordinate calibration on the activity trajectory data to generate a three-dimensional motion path with a time stamp; Tracing back the trajectory in the opposite direction of the motion path to construct a reverse motion vector set; Calculating the convergence area of ​​each reverse motion vector in three-dimensional space, wherein the convergence area is an area where the spatial distribution density of the reverse vector intersections exceeds a preset distribution density threshold; The convergence area is determined as a danger source area.

7. The method according to claim 1, characterized in that The monitoring of each of the target resident animals based on the monitoring points includes: A plurality of infrared sensors pre-arranged at the monitoring point are used to detect in real time whether there are animal heat source signals in the surrounding preset area environment; When the animal heat source signal is detected, controlling the sound sensor to collect target sound data; If the target sound data matches the sound frequency characteristics of the target resident animal, triggering the monitoring device corresponding to the monitoring point to collect video data of the target resident animal; Identifying the collected video data within a set startup time after the monitoring device is started to determine whether the animal corresponding to the animal heat source signal is the target resident animal; If yes, continue to collect video data; If not, controlling the monitoring device to stop working; When the target resident animal leaves the set monitoring range, or the animal heat source signal is not detected within the preset waiting time, the monitoring device is controlled to stop working.

8. A collapse prediction system, characterized in that: The collapse prediction system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the collapse prediction system to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a collapse prediction system, the collapse prediction system is caused to perform the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a collapse prediction system, the collapse prediction system is caused to perform the method according to any one of claims 1 to 7.

Images