Wild animal monitoring system and method based on multi-dimensional information fusion

Abstract

The invention discloses a wildlife monitoring system and method based on multi-dimensional information fusion, and relates to the technical field of wildlife monitoring, and the method comprises the steps: collecting environment data and animal activity information, processing the collected data, carrying out the data fusion of the processed data, carrying out the risk assessment of the fused target feature information, and carrying out the risk assessment of the target feature information. The method comprises the following steps: calculating a comprehensive risk index, evaluating behavior risks of wild animals, carrying out risk index grading, setting a risk threshold value, carrying out risk early warning, dynamically optimizing an early warning threshold value according to historical data and environmental parameters, generating a control instruction set according to a risk grade, and carrying out response according to a fusion decision result. According to the method, the improved D-S evidence theory and the dynamic weight factor are introduced, the data fusion process is optimized, and more scientific and effective technical support is provided for wild animal protection and management.

Description

Technical Field

[0001] The present invention relates to the technical field of wildlife monitoring, and particularly to a wildlife monitoring system and method based on multi-dimensional information fusion. Background Art

[0002] With the enhancement of ecological environment protection awareness and the improvement of wildlife protection requirements, wildlife monitoring technology has received extensive attention and development. Traditional wildlife monitoring methods mainly rely on manual observation and simple sensor technologies, and these methods have certain limitations in terms of the accuracy, real-time nature, and comprehensiveness of data collection. In recent years, with the progress of sensor technology, data processing technology, and artificial intelligence algorithms, multi-dimensional information fusion technology has been applied in the field of wildlife monitoring, improving the intelligence and automation levels of monitoring systems.

[0003] However, there are still deficiencies in the multi-dimensional information fusion of existing wildlife monitoring systems. First of all, existing systems often only focus on single or a few types of sensors in data collection, lacking the comprehensive application of multi-source heterogeneous data, resulting in incomplete monitoring results. Secondly, data processing and fusion algorithms are often relatively simple and difficult to effectively process multi-modal data in complex environments, thus affecting the accuracy and reliability of monitoring. Moreover, existing risk assessment and early warning systems lack dynamics and adaptability and are difficult to quickly respond to the real-time changes in wildlife behavior. Therefore, it is particularly important to develop a wildlife monitoring system that can efficiently fuse multi-source information, accurately assess risks, and implement hierarchical early warnings. Summary of the Invention

[0004] In view of the above existing problems, the present invention provides a wildlife monitoring system and method based on multi-dimensional information fusion to solve the problems in the prior art that the monitoring results are not comprehensive enough, it is difficult to effectively process multi-modal data in complex environments, and it is difficult to quickly respond to the real-time changes in wildlife behavior.

[0005] To solve the above technical problems, a wildlife monitoring system based on multi-dimensional information fusion is proposed, including a data collection and processing module, a data fusion module, a risk assessment and early warning module, and a strategy matching and decision-making module; The data collection and processing module is used to collect environmental data and animal activity information, and process the collected data; the data fusion module is used to construct a spatiotemporal feature tensor, fuse multimodal data using the improved DS evidence theory, and optimize the fusion result by introducing a dynamic weight factor and a confidence correction mechanism; the risk assessment and early warning module is used to calculate a comprehensive risk index, assess the behavioral risk of wild animals, and classify the risk index through a nonlinear risk grading model, dynamically set the risk threshold, and perform graded early warnings; the strategy matching and decision-making module is used to match the response strategy according to the early warning level, generate a control instruction set, and execute a response operation according to the fusion decision result.

[0006] As a preferred solution of the wildlife monitoring system based on multi-dimensional information fusion described in the present invention, the data acquisition and processing module includes an infrared sensing unit, a voiceprint acquisition unit, an optical imaging unit and an environmental monitoring unit; The infrared sensing unit includes using a dual-band infrared detector to synchronously collect thermal radiation data of the target area and dynamically correct the thermal sensitivity threshold according to the ambient humidity; The voiceprint collection unit includes real-time collection of voiceprint data of the target area, focusing on the target sound source using directional sound pickup technology, and identifying low-frequency vibration signals through an infrasound detection unit to determine animal activities; The optical imaging unit includes a motion compensation exposure technology, dynamically adjusting shutter parameters according to the target moving speed, and performing multi-spectral synchronous capture; The environmental monitoring unit includes real-time collection of microenvironment parameters, correcting the thermal sensitivity threshold of the infrared sensing unit through temperature and humidity data, and compensating the sound wave propagation error of the voiceprint collection unit using barometer data.

[0007] As a preferred solution of the wildlife monitoring system based on multidimensional information fusion described in the present invention, wherein: the data fusion module includes a spatiotemporal feature tensor construction unit, an improved DS evidence theory fusion unit, a multimodal data association analysis unit and a fusion result optimization unit; The spatiotemporal feature tensor construction unit includes integrating the multimodal data collected by the infrared sensing unit, the voiceprint collection unit, the optical imaging unit and the environmental monitoring unit into a spatiotemporal feature tensor according to time series and spatial distribution, and using a time alignment algorithm to eliminate the time deviation of the multi-source data; The improved DS evidence theory fusion unit includes synthesizing evidence of multimodal data in the spatiotemporal feature tensor, optimizing the evidence synthesis rule by introducing a dynamic weight factor and a confidence correction mechanism, and generating fused target feature information; The multimodal data correlation analysis unit includes constructing a multimodal data correlation matrix, analyzing the correlation between different modal data, and dynamically adjusting the correlation weight by introducing a time decay factor to generate the correlation features of multimodal data. The fusion result optimization unit includes optimizing the fusion result by introducing a correction factor based on environmental parameters to generate the final target feature information.

[0008] As a preferred solution of the wildlife monitoring system based on multi-dimensional information fusion according to the present invention, wherein: the risk assessment and early warning module includes a comprehensive risk index calculation unit, a risk grading and threshold setting unit, and an early warning trigger unit. The comprehensive risk index calculation unit includes calculating a comprehensive risk index according to the fused target feature information generated by the data fusion module and in combination with a preset non-linear risk assessment model, wherein the non-linear risk assessment model dynamically adjusts the risk weight by introducing environmental parameters and historical data. The risk grading and threshold setting unit includes grading the wildlife behavior risk according to the comprehensive risk index and dynamically setting a risk threshold by means of an adaptive threshold algorithm. The early warning trigger unit includes triggering a risk warning signal when the comprehensive risk index exceeds the set risk threshold and generating warning information of different levels through a multi-level early warning mechanism. The policy matching and decision-making module includes matching a response policy according to the early warning level, generating a control instruction set, and performing a response operation according to the fusion decision result.

[0009] In addition, the present invention also proposes a wildlife monitoring method based on multi-dimensional information fusion, which includes collecting environmental data and animal activity information and processing the collected data; performing data fusion on the processed data, optimizing the fusion result, performing risk assessment on the fused target feature information, calculating a comprehensive risk index, and assessing the behavior risk of wild animals; performing risk index grading, setting a risk threshold, performing risk early warning, dynamically optimizing the early warning threshold according to historical data and environmental parameters, generating a control instruction set according to the risk level, and performing a response according to the fusion decision result.

[0010] As a preferred solution of the wildlife monitoring method based on multi-dimensional information fusion according to the present invention, wherein: the collecting environmental data and animal activity information includes collecting thermal radiation data, voiceprint data, image data and environmental parameters of a target area and preprocessing the collected data.

[0011] As a preferred solution of the wild animal monitoring method based on multi-dimensional information fusion according to the present invention, wherein: the data fusion includes integrating multi-modal data collected by an infrared sensing unit, a voiceprint acquisition unit, an optical imaging unit, and an environmental monitoring unit into a spatio-temporal feature tensor according to time series and spatial distribution, eliminating the time deviation of multi-source data through a time alignment algorithm, and performing evidence synthesis on the multi-modal data in the spatio-temporal feature tensor by using an improved D-S evidence theory. By introducing a dynamic weight factor and a confidence correction mechanism, the evidence synthesis rule is optimized to generate fused target feature information; The improved D-S evidence theory includes calculating a dynamic weight factor of multi-modal data according to micro-environment parameters collected by the environmental monitoring unit, introducing a confidence correction factor to correct the evidence synthesis result, and performing evidence synthesis on the multi-modal data according to the dynamic weight factor and the confidence correction factor to generate fused target feature information; The formula of the dynamic weight factor is expressed as: , wherein, is the dynamic weight factor of the i-th modal data, is the change amount of the environmental parameter of the i-th modal data, is the base of the natural logarithm, is the change amount of the environmental parameter of the j-th modal data, n is the total number of modal data, and i, j, and k are variable indexes, is the environmental sensitivity coefficient, is the k-th environmental parameter value of the i-th modal data, is the k-th average value of the current environmental parameter, and m is the number of environmental parameters; The dynamic adjustment of the correlation weight includes calculating the correlation between multi-modal data by using Pearson correlation, constructing a multi-modal data correlation matrix, calculating a time decay factor according to the time series characteristics of the multi-modal data, dynamically adjusting the correlation weight in the correlation matrix, and optimizing the fusion weight of the multi-modal data; The formula for calculating the time decay factor is expressed as: , wherein, is the time decay factor between the i-th modal data and the j-th modal data, is the time difference between the i-th modal data and the j-th modal data, is the time decay coefficient, and i and j are variable indexes, is the base of the natural logarithm; The adjusted correlation weight is expressed as: , wherein, Adjusted correlation weight, is the correlation coefficient between the i-th modality data and the j-th modality data, is the time decay factor between the i-th modality data and the j-th modality data, where i and j are variable indices; The formula for calculating the adjusted fusion weight is: , where, is the adjusted fusion weight, is the fusion weight before adjustment, is the learning rate, and A is the fusion error function.

[0012] As a preferred solution of the wild animal monitoring method based on multi-dimensional information fusion according to the present invention, wherein: optimizing the fusion result includes calculating an environmental parameter correction factor according to the micro-environmental parameters collected by the environmental monitoring unit, and optimizing the fusion result according to the environmental parameter correction factor to generate the final target feature information; Evaluating the behavioral risk of wild animals includes calculating a comprehensive risk index according to the fused target feature information, in combination with a preset non-linear risk assessment model, and setting a risk index threshold to classify the behavioral risk of wild animals; The non-linear risk assessment model is expressed as: , where, is the comprehensive risk index, , and are dynamically adjusted environmental weight factors, is the weight of the a-th feature, is the non-linear function of the a-th feature, is the specific value of the a-th feature, is the influence function of the b-th environmental parameter, is the specific value of the b-th feature, where a and b are variable indices, o is the total number of features, p is the total number of environmental parameters, is the historical correction function, and c is the historical data.

[0013] As a preferred solution of the wild animal monitoring method based on multi-dimensional information fusion according to the present invention, wherein: performing risk index classification includes, based on historical data, determining the distribution of the comprehensive risk index R, setting an initial threshold, dynamically setting a risk threshold through an adaptive threshold algorithm, and classifying the behavioral risk of wild animals into a low risk level, a medium risk level, and a high risk level; The setting of the initial threshold includes setting the low - risk threshold as the 25th percentile of the comprehensive risk index R, setting the medium - risk threshold as the 50th percentile of the comprehensive risk index R, and setting the high - risk threshold as the 75th percentile of the comprehensive risk index R; The adaptive threshold algorithm is expressed as: , where, is the adjusted risk threshold, is the initial risk threshold, is the environmental sensitivity coefficient, is the change in environmental parameters, v is the variable index, and v = 1 or 2; When , it is determined that the risk of wild animal behavior is at a low - risk level. When , it is determined that the risk of wild animal behavior is at a medium - risk level. When , it is determined that the risk of wild animal behavior is at a high - risk level.

[0014] As a preferred solution of the wild animal monitoring method based on multi - dimensional information fusion described in the present invention, wherein: the execution of the response according to the fusion decision result includes dynamically matching a preset response strategy according to the risk level generated by the risk assessment; The preset response strategies include issuing a general risk warning at a low - risk level, increasing the data collection frequency, starting the low - power mode, recording abnormal behavior data and generating a preliminary analysis report; issuing an intermediate - risk warning at a medium - risk level, simultaneously starting the infrared sensing unit, the voiceprint collection unit, the optical imaging unit and the environmental monitoring unit, expanding the monitoring range from the target area to the surrounding area, generating a detailed analysis report based on multi - modal data and sending it to the monitoring center, and starting the warning preparation mechanism; issuing a high - risk warning at a high - risk level, triggering the sound and light alarm system, notifying relevant personnel to take emergency measures, starting the drone to patrol the target area, generating an emergency report based on multi - modal data and uploading it to the cloud, and starting the emergency response mechanism; The strategy matching formula is: , where, S is the finally selected optimal strategy, is the strategy set, is a single strategy, is the matching probability of strategy s under the comprehensive risk index R, R is the comprehensive risk index, is the priority weight of strategy s, is the initial weight of strategy s, is the time decay coefficient, is the execution time of strategy s, is the remaining strategies in the strategy set, is the base of the natural logarithm, is the strategy 's initial weight, is the strategy 's execution time.

[0015] A computer device includes a memory and a processor. The memory stores a computer program. It is characterized in that when the processor executes the computer program, it implements the method steps of a wildlife monitoring based on multi-dimensional information fusion.

[0016] A computer-readable storage medium stores a computer program. It is characterized in that when the computer program is executed by a processor, it implements the method steps of a wildlife monitoring based on multi-dimensional information fusion.

[0017] Advantages of the present invention: Through the data acquisition and processing module including an infrared sensing unit, a voiceprint acquisition unit, an optical imaging unit and an environmental monitoring unit, the present invention realizes the comprehensive acquisition of environmental data and animal activity information, and through the application of the temperature gradient compensation algorithm of the dual-band infrared detector and the directional sound pickup technology, improves the accuracy and pertinence of data acquisition, ensures the comprehensiveness and reliability of monitoring data, and improves the response speed and accuracy of the monitoring system; By constructing a spatio-temporal feature tensor and using the improved D-S evidence theory to fuse multi-modal data, the efficient fusion of data is realized, and a dynamic weight factor and a confidence correction mechanism are introduced to optimize the fusion result, making the fused data more accurate, improving the intelligent level of data processing, and enhancing the ability of the monitoring system to identify wildlife behaviors; By calculating the comprehensive risk index and adopting a non-linear risk grading model, the risk threshold can be dynamically set and graded early warning can be carried out, realizing the real-time assessment of the wildlife behavior risk, improving the early warning ability and response speed of the monitoring system, matching the response strategy according to the early warning level, and generating a control instruction set to execute the response operation, ensuring that the monitoring system can take corresponding measures according to different risk levels, and improving the automation and intelligent management level of the monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings, where: Figure 1 is the overall flowchart of a wildlife monitoring method based on multi-dimensional information fusion provided by an embodiment of the present invention.

[0019] Figure 2System flowchart of a wildlife monitoring system based on multi-dimensional information fusion provided by an embodiment of the present invention. Detailed implementation manners

[0020] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed implementation manners of the present invention with reference to the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0022] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that can be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they embodiments that are mutually exclusive individually or selectively with other embodiments.

[0023] The present invention is described in detail with reference to schematic diagrams. When detailing the embodiments of the present invention, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally not in accordance with the general scale, and the schematic diagrams are only examples and should not limit the protection scope of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0024] At the same time, in the description of the present invention, it should be noted that the orientation or positional relationships indicated by terms such as "upper, lower, inner, and outer" are based on the orientation or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0025] Unless otherwise clearly defined and limited in the present invention, the terms "installation, connection, and coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can also be a mechanical connection, an electrical connection, or a direct connection, and can also be indirectly connected through an intermediate medium, or can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0026] Example 1, referring to Figure 1 , which is the first embodiment of the present invention. This embodiment provides a wildlife monitoring method based on multi-dimensional information fusion, including: S1: Collect environmental data and animal activity information, and process the collected data.

[0027] Use a dual-band infrared detector to synchronously collect the thermal radiation data of the target area, integrate the temperature gradient compensation algorithm, dynamically correct the thermal sensitivity threshold according to the environmental humidity. When the detected temperature difference is greater than or equal to 2°C, automatically increase the sampling rate to 30 Hz, mark the target area, and optimize the energy consumption through a pulsed scanning strategy, reducing the sampling rate to 10 Hz in the non-target area; Real-time collect the voiceprint data of the target area, and use directional sound pickup technology to focus on the target sound source. When the energy mutation of the frequency band exceeds the baseline value by 15 dB, trigger high-frequency sampling and record the voiceprint characteristics, and identify low-frequency vibration signals through the infrasound detection unit to judge animal activities; Adopt motion compensation exposure technology, dynamically adjust the shutter parameters according to the target movement speed, perform multi-spectral synchronous capture. When the infrared sensing unit detects the target, automatically trigger the optical unit to collect images, and enhance the image contrast through adaptive histogram equalization; Real-time collect the micro-environment parameters, correct the thermal sensitivity threshold of the infrared sensing unit through the temperature and humidity data, compensate the acoustic wave propagation error of the voiceprint collection unit using the barometer data, and detect the environmental vibration through a three-dimensional accelerometer to assist the voiceprint unit in distinguishing the target from the environmental noise.

[0028] It should be noted that through the data acquisition and processing modules of the infrared sensing unit, voiceprint collection unit, optical imaging unit and environmental monitoring unit, the comprehensive collection of environmental data and animal activity information is realized. In particular, through the application of the temperature gradient compensation algorithm of the dual-band infrared detector and the directional sound pickup technology, the accuracy and pertinence of data collection are improved.

[0029] S2: Perform data fusion on the processed data, optimize the fusion result, perform risk assessment on the fused target feature information, calculate the comprehensive risk index, and evaluate the behavior risk of wild animals.

[0030] Further, the data fusion includes integrating multi-modal data collected by the infrared sensing unit, voiceprint acquisition unit, optical imaging unit, and environmental monitoring unit into a spatio-temporal feature tensor according to the time series and spatial distribution, eliminating the time deviation of multi-source data through the time alignment algorithm, and performing evidence synthesis on the multi-modal data in the spatio-temporal feature tensor using the improved D-S evidence theory. By introducing a dynamic weight factor and a confidence correction mechanism, the evidence synthesis rule is optimized to generate the fused target feature information; Furthermore, the improved D-S evidence theory includes calculating the dynamic weight factor of multi-modal data according to the micro-environment parameters collected by the environmental monitoring unit, introducing a confidence correction factor to correct the evidence synthesis result, and performing evidence synthesis on the multi-modal data according to the dynamic weight factor and the confidence correction factor to generate the fused target feature information; The formula for the dynamic weight factor is expressed as: , where, is the dynamic weight factor of the i-th modal data, is the change amount of the environmental parameter of the i-th modal data, is the base of the natural logarithm, is the change amount of the environmental parameter of the j-th modal data, n is the total number of modal data, and i, j, and k are variable indices, is the environmental sensitivity coefficient, is the k-th environmental parameter value of the i-th modal data, is the k-th average value of the current environmental parameter, and m is the number of environmental parameters; The formula for the confidence correction factor is: , where, is the confidence correction factor of the i-th modal data, is the standard deviation of the i-th modal data, is the maximum value of the standard deviation of the modal data, and i is the variable index; The dynamic adjustment of the correlation weight includes calculating the correlation between multi-modal data using Pearson correlation, constructing a multi-modal data correlation matrix, calculating the time decay factor according to the time series characteristics of the multi-modal data, dynamically adjusting the correlation weight in the correlation matrix, and optimizing the fusion weight of the multi-modal data according to the adjusted correlation weight; The formula for calculating the time decay factor is expressed as: , where, is the time decay factor between the i-th modal data and the j-th modal data, is the time difference between the i-th modal data and the j-th modal data, is the time decay coefficient, and i and j are variable indices; The adjusted correlation weight is expressed as: , where, the adjusted correlation weight, is the correlation coefficient between the i-th modal data and the j-th modal data, is the time decay factor between the i-th modal data and the j-th modal data, and i and j are variable indices; The formula for calculating the adjusted fusion weight is: , where, is the adjusted fusion weight, is the fusion weight before adjustment, is the learning rate, and A is the fusion error function.

[0031] Furthermore, the present invention realizes the efficient fusion of data by constructing a spatio-temporal feature tensor and using an improved D-S evidence theory to fuse multi-modal data, and introduces a dynamic weight factor and a confidence correction mechanism to optimize the fusion result, making the fused data more accurate.

[0032] It should be noted that optimizing the fusion result includes calculating an environmental parameter correction factor according to the micro-environmental parameters collected by the environmental monitoring unit, and optimizing the fusion result according to the environmental parameter correction factor to generate the final target feature information; The formula for calculating the environmental parameter correction factor is: , where, is the environmental parameter correction factor, is the current environmental parameter, is the environmental parameter threshold, is the environmental sensitivity coefficient, is the base of the natural logarithm; Evaluating the behavior risk of wild animals includes calculating a comprehensive risk index according to the fused target feature information, combining a preset non-linear risk assessment model, setting a risk index threshold, classifying the behavior risk of wild animals, and giving a classification warning; The non-linear risk assessment model is expressed as: , where, is the comprehensive risk index, , and is the environmental weight factor for dynamic adjustment, is the weight of the a-th feature, is the non-linear function of the a-th feature, is the specific value of the a-th feature, is the influence function of the b-th environmental parameter, is the specific value of the b-th feature. a and b are variable indices, o is the total number of features, and p is the total number of environmental parameters, is the historical correction function, and c is the historical data.

[0033] By calculating the comprehensive risk index and adopting the non-linear risk grading model, the present invention can dynamically set the risk threshold and conduct grading early warning, realizing the real-time assessment of the wild animal behavior risk and improving the early warning ability and response speed of the monitoring system.

[0034] S3: Conduct risk index grading, set the risk threshold, conduct risk early warning, and dynamically optimize the early warning threshold according to the historical data and environmental parameters, and generate a control instruction set according to the risk level and execute the response according to the fusion decision result.

[0035] Further, grading the wild animal behavior risk includes determining the distribution of the comprehensive risk index R based on the historical data, setting the initial threshold, dynamically setting the risk threshold through the adaptive threshold algorithm, and classifying the wild animal behavior risk into low risk level, medium risk level and high risk level; The setting of the initial threshold includes setting the low risk threshold as the 25th percentile of the comprehensive risk index R, setting the medium risk threshold as the 50th percentile of the comprehensive risk index R, and setting the high risk threshold as the 75th percentile of the comprehensive risk index R; The adaptive threshold algorithm is expressed as: , where, is the adjusted risk threshold, is the initial risk threshold, is the environmental sensitivity coefficient, is the change amount of the environmental parameter, v is the variable index, and v = 1 or 2; When , it is determined that the wild animal behavior risk is at the low risk level. When , it is determined that the wild animal behavior risk is at the medium risk level. When , it is determined that the wild animal behavior risk is at the high risk level.

[0036] Furthermore, the matching response strategy includes dynamically matching the preset response strategy according to the risk level generated by the risk assessment; The preset response strategies include issuing ordinary risk warnings for low-risk levels, increasing the data collection frequency, activating the low-power mode, recording abnormal behavior data and generating preliminary analysis reports; issuing medium-risk warnings for medium-risk levels, simultaneously activating the infrared sensing unit, voiceprint collection unit, optical imaging unit and environmental monitoring unit, expanding the monitoring range from the target area to the surrounding area, generating detailed analysis reports based on multi-modal data and sending them to the monitoring center, and activating the warning preparation mechanism; issuing high-risk warnings for high-risk levels, triggering the audible and visual alarm system, notifying relevant personnel to take emergency measures, activating the drone to conduct inspections on the target area, generating emergency reports based on multi-modal data and uploading them to the cloud, and activating the emergency response mechanism; The strategy matching formula is: , where S is the best strategy finally selected, is the strategy set, is a single strategy, is the matching probability of strategy s under the comprehensive risk index R, and R is the comprehensive risk index, is the priority weight of strategy s, is the initial weight of strategy s, is the time decay coefficient, is the execution time of strategy s, is the remaining strategies in the strategy set, is the base of the natural logarithm, is the strategy of the initial weight, is the strategy of the execution time.

[0037] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

[0038] Example 2, referring to Figure 2 , which is the second embodiment of the present invention. This embodiment provides a wildlife monitoring system based on multi-dimensional information fusion, including a data collection and processing module, a data fusion module, a risk assessment and warning module, and a strategy matching and decision-making module; The data collection and processing module includes an infrared sensing unit, a voiceprint collection unit, an optical imaging unit and an environmental monitoring unit; The infrared sensing unit includes using a dual-band infrared detector to synchronously collect the thermal radiation data of the target area, integrating a temperature gradient compensation algorithm, dynamically correcting the thermal sensitivity threshold according to the environmental humidity. When the detected temperature difference is greater than or equal to 2°C, the sampling rate is automatically increased to 30 Hz, and the target area is marked. The energy consumption is optimized through a pulse scanning strategy, and the sampling rate is reduced to 10 Hz in the area without a target.

[0039] The voiceprint acquisition unit includes real-time collecting the voiceprint data of the target area, and using directional sound pickup technology to focus on the target sound source. When the sudden change in frequency band energy exceeds the baseline value by 15 dB, high-frequency sampling is triggered and the voiceprint features are recorded. The infrasonic wave detection unit is used to identify low-frequency vibration signals to judge animal activities.

[0040] The optical imaging unit includes adopting motion compensation exposure technology, dynamically adjusting the shutter parameters according to the target movement speed, and performing multi-spectral synchronous capture. When the infrared sensing unit detects a target, the optical unit is automatically triggered to perform image acquisition, and the image contrast is enhanced through adaptive histogram equalization.

[0041] The environmental monitoring unit includes real-time collecting the micro-environment parameters, correcting the thermal sensitivity threshold of the infrared sensing unit through temperature and humidity data, compensating the sound wave propagation error of the voiceprint acquisition unit by using barometer data, and detecting environmental vibrations through a three-dimensional accelerometer to assist the voiceprint unit in distinguishing the target from environmental noise.

[0042] Furthermore, the data fusion module includes a spatio-temporal feature tensor construction unit, an improved D-S evidence theory fusion unit, a multi-modal data correlation analysis unit, and a fusion result optimization unit.

[0043] The spatio-temporal feature tensor construction unit includes integrating the multi-modal data collected by the infrared sensing unit, the voiceprint acquisition unit, the optical imaging unit, and the environmental monitoring unit into a spatio-temporal feature tensor according to the time series and spatial distribution, and using a time alignment algorithm to eliminate the time deviation of multi-source data.

[0044] The improved D-S evidence theory fusion unit includes performing evidence synthesis on the multi-modal data in the spatio-temporal feature tensor, and optimizing the evidence synthesis rule by introducing a dynamic weight factor and a confidence correction mechanism to generate the fused target feature information.

[0045] The multi-modal data correlation analysis unit includes analyzing the correlation between different modal data by constructing a multi-modal data correlation matrix, and dynamically adjusting the correlation weight by introducing a time decay factor to generate the correlation features of multi-modal data.

[0046] The fusion result optimization unit includes optimizing the fusion result by introducing a correction factor based on environmental parameters to generate the final target feature information.

[0047] Furthermore, the risk assessment and early warning module includes a comprehensive risk index calculation unit, a risk grading and threshold setting unit, and an early warning trigger unit.

[0048] The comprehensive risk index calculation unit calculates the comprehensive risk index according to the fused target feature information generated by the data fusion module and in combination with a preset non-linear risk assessment model, wherein the non-linear risk assessment model dynamically adjusts the risk weight by introducing environmental parameters and historical data.

[0049] The risk grading and threshold setting unit grades the wild animal behavior risk according to the comprehensive risk index and dynamically sets the risk threshold through an adaptive threshold algorithm.

[0050] The early warning trigger unit triggers a risk early warning signal when the comprehensive risk index exceeds the set risk threshold and generates early warning information of different levels through a multi-level early warning mechanism.

[0051] The policy matching and decision-making module matches response policies according to the early warning level, generates a control instruction set, and executes response operations according to the fused decision result.

[0052] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

[0053] Embodiment 3, the third embodiment of the present invention, which is different from the previous two embodiments in that: If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the essence of the technical solution of the present invention, or the part that contributes to the prior art, or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0054] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered as a definable sequence of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. As used in this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with the instruction execution system, apparatus, or device.

[0055] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, as the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0056] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

Claims

1. A wildlife monitoring system based on multi-dimensional information fusion, characterized in that: It includes data collection and processing module, data fusion module, risk assessment and early warning module and strategy matching and decision-making module; The data collection and processing module is used to collect environmental data and animal activity information, and process the collected data; The data fusion module is used to construct a spatiotemporal feature tensor, fuse multimodal data using an improved DS evidence theory, and optimize the fusion result by introducing a dynamic weight factor and a confidence correction mechanism; The risk assessment and early warning module is used to calculate the comprehensive risk index, assess the behavioral risk of wild animals, and classify the risk index through a nonlinear risk classification model, dynamically set the risk threshold, and conduct graded early warning; The strategy matching and decision module is used to match the response strategy according to the warning level, generate a control instruction set, and execute a response operation according to the fusion decision result.

2. The wildlife monitoring system based on multi-dimensional information fusion according to claim 1, characterized in that: The data acquisition and processing module includes an infrared sensing unit, a voiceprint acquisition unit, an optical imaging unit and an environment monitoring unit; The infrared sensing unit includes using a dual-band infrared detector to synchronously collect thermal radiation data of the target area and dynamically correct the thermal sensitivity threshold according to the ambient humidity; The voiceprint collection unit includes real-time collection of voiceprint data of the target area, focusing on the target sound source using directional sound pickup technology, and identifying low-frequency vibration signals through an infrasound detection unit to determine animal activities; The optical imaging unit includes a motion compensation exposure technology, dynamically adjusting shutter parameters according to the target moving speed, and performing multi-spectral synchronous capture; The environmental monitoring unit includes real-time collection of microenvironment parameters, correcting the thermal sensitivity threshold of the infrared sensing unit through temperature and humidity data, and compensating the sound wave propagation error of the voiceprint collection unit using barometer data.

3. The wildlife monitoring system based on multi-dimensional information fusion according to claim 2, wherein: The data fusion module includes: Spatiotemporal feature tensor construction unit, improved DS evidence theory fusion unit, multimodal data association analysis unit and fusion result optimization unit; The spatiotemporal feature tensor construction unit includes integrating the multimodal data collected by the infrared sensing unit, the voiceprint collection unit, the optical imaging unit and the environmental monitoring unit into a spatiotemporal feature tensor according to time series and spatial distribution, and using a time alignment algorithm to eliminate the time deviation of the multi-source data; The improved DS evidence theory fusion unit includes synthesizing evidence of multimodal data in the spatiotemporal feature tensor, optimizing the evidence synthesis rule by introducing a dynamic weight factor and a confidence correction mechanism, and generating fused target feature information; The multimodal data association analysis unit includes building a multimodal data association matrix, analyzing the correlation between different modal data, and dynamically adjusting the association weight by introducing a time decay factor to generate association features of the multimodal data; The fusion result optimization unit includes optimizing the fusion result by introducing a correction factor based on the environmental parameters to generate final target feature information.

4. The wildlife monitoring system based on multi-dimensional information fusion according to claim 3, characterized in that: The risk assessment and early warning module includes a comprehensive risk index calculation unit, a risk classification and threshold setting unit and an early warning triggering unit; The comprehensive risk index calculation unit calculates the comprehensive risk index based on the fused target feature information generated by the data fusion module and in combination with a preset non-linear risk assessment model, wherein the non-linear risk assessment model dynamically adjusts the risk weight by introducing environmental parameters and historical data; The risk grading and threshold setting unit grades the wild animal behavior risk according to the comprehensive risk index and dynamically sets the risk threshold through an adaptive threshold algorithm; The early warning trigger unit triggers a risk warning signal when the comprehensive risk index exceeds the set risk threshold and generates warning information of different levels through a multi-level early warning mechanism; The strategy matching and decision-making module matches response strategies according to the early warning level, generates a control instruction set, and executes response operations according to the fused decision result.

5. A wildlife monitoring method based on multi-dimensional information fusion, applied to a wildlife monitoring system based on multi-dimensional information fusion according to any one of claims 1-4, characterized in that: including, Collect environmental data and animal activity information and process the collected data; Fuse the processed data, optimize the fusion result, conduct risk assessment on the fused target feature information, calculate the comprehensive risk index, and assess the wild animal behavior risk; Conduct risk index grading, set risk thresholds, conduct risk early warnings, dynamically optimize the early warning thresholds according to historical data and environmental parameters, generate a control instruction set according to the risk level, and execute responses according to the fused decision result.

6. The wildlife monitoring method based on multi-dimensional information fusion according to claim 5, wherein: The collection of environmental data and animal activity information includes collecting thermal radiation data, voiceprint data, image data, and environmental parameters of the target area and preprocessing the collected data.

7. The wildlife monitoring method based on multi-dimensional information fusion according to claim 6, wherein: The data fusion includes integrating multi-modal data collected by the infrared sensing unit, voiceprint collection unit, optical imaging unit, and environmental monitoring unit into a spatio-temporal feature tensor according to time series and spatial distribution, eliminating the time deviation of multi-source data through a time alignment algorithm, and performing evidence synthesis on the multi-modal data in the spatio-temporal feature tensor using an improved D-S evidence theory. Optimize the evidence synthesis rule by introducing a dynamic weight factor and a confidence correction mechanism to generate the fused target feature information; The improved D-S evidence theory includes calculating the dynamic weight factor of multi-modal data according to the micro-environmental parameters collected by the environmental monitoring unit, introducing a confidence correction factor to correct the evidence synthesis result, and performing evidence synthesis on the multi-modal data according to the dynamic weight factor and the confidence correction factor to generate the fused target feature information; The formula for the dynamic weight factor is expressed as: ， Among them, is the dynamic weight factor of the i-th modal data, is the change in environmental parameters of the i-th modal data, is the base of the natural logarithm, is the change in environmental parameters of the j-th modal data, n is the total number of modal data, and i, j, and k are variable indices, is the environment sensitivity coefficient, is the k-th environmental parameter value of the i-th modal data, is the k-th average value of the current environmental parameters, and m is the number of environmental parameters; Dynamically adjusting the correlation weight includes calculating the correlation between multi-modal data using Pearson correlation, constructing a multi-modal data correlation matrix, calculating a time decay factor according to the time series characteristics of the multi-modal data, dynamically adjusting the correlation weight in the correlation matrix, and optimizing the fusion weight of the multi-modal data; The formula for calculating the time decay factor is expressed as: ， wherein, is the time decay factor between the i-th modal data and the j-th modal data, is the time difference between the i-th modal data and the j-th modal data, is the time decay coefficient, i and j are variable indices, is the base of the natural logarithm; The adjusted correlation weight is expressed as: ， Among them, The adjusted correlation weight is the correlation coefficient between the i-th modal data and the j-th modal data, is the time decay factor between the i-th modal data and the j-th modal data, where i and j are variable indices; The formula for the adjusted fusion weight is: ， Among them, is the adjusted fusion weight, is the fusion weight before adjustment, is the learning rate, and A is the fusion error function.

8. The wildlife monitoring method based on multi-dimensional information fusion according to claim 7, characterized in that: The optimization of the fusion result includes calculating an environmental parameter correction factor according to the micro-environmental parameters collected by the environmental monitoring unit and optimizing the fusion result according to the environmental parameter correction factor to generate the final target feature information; The described behavior risk assessment of wild animals includes calculating a comprehensive risk index according to the fused target feature information, combining with a preset non-linear risk assessment model, setting a risk index threshold, and classifying the behavior risk of wild animals; The described non-linear risk assessment model is expressed as: ， Among them, is the comprehensive risk index, , and are dynamically adjusted environmental weight factors, is the weight of the a-th feature, is the non-linear function of the a-th feature, is the specific value of the a-th feature, is the influence function of the b-th environmental parameter, is the specific value of the b-th feature. a and b are variable indices, o is the total number of features, p is the total number of environmental parameters, is the historical correction function, and c is historical data.

9. The wildlife monitoring method based on multi-dimensional information fusion according to claim 8, characterized in that: The described risk index classification includes determining the distribution of the comprehensive risk index R based on historical data, setting an initial threshold, dynamically setting the risk threshold through an adaptive threshold algorithm, and classifying the behavior risk of wild animals into low-risk level, medium-risk level, and high-risk level; The described setting of the initial threshold includes setting the low-risk threshold as the 25th percentile of the comprehensive risk index R, setting the medium-risk threshold as the 50th percentile of the comprehensive risk index R, and setting the high-risk threshold as the 75th percentile of the comprehensive risk index R; The described adaptive threshold algorithm is expressed as: ， Among them, is the adjusted risk threshold, is the initial risk threshold, is the environmental sensitivity coefficient, is the change in environmental parameters, and v is the variable index, where v = 1 or 2; When it is the case, it is determined that the wild animal behavior risk is at a low risk level. When it is the case, it is determined that the wild animal behavior risk is at a medium risk level. When it is the case, it is determined that the wild animal behavior risk is at a high risk level.

10. The wildlife monitoring method based on multi-dimensional information fusion according to claim 9, characterized in that: The described execution of the response according to the fusion decision result includes dynamically matching a preset response strategy according to the risk level generated by the risk assessment; The preset response strategies include issuing a general risk warning for the low-risk level, increasing the data collection frequency, starting the low-power mode, recording abnormal behavior data, and generating a preliminary analysis report; For the medium-risk level, issue a medium-risk warning, simultaneously start the infrared sensing unit, voiceprint collection unit, optical imaging unit, and environmental monitoring unit, expand the monitoring range from the target area to the surrounding area, generate a detailed analysis report based on multi-modal data and send it to the monitoring center, and start the warning preparation mechanism; for the high-risk level, issue a high-risk warning, trigger the sound and light alarm system, notify relevant personnel to take emergency measures, start the drone to patrol the target area, generate an emergency report based on multi-modal data and upload it to the cloud, and start the emergency response mechanism; The strategy matching formula is: ， Among them, S is the finally selected optimal strategy, is the strategy set, is a single strategy, is the matching probability of strategy s under the comprehensive risk index R, and R is the comprehensive risk index, is the priority weight of strategy s, is the initial weight of strategy s, is the time decay coefficient, is the execution time of strategy s, is the remaining strategies in the strategy set, is the base of the natural logarithm, is the strategy 's initial weight, is the strategy 's execution time.

Images