Wild animal image compression and encryption transmission method driven by edge calculation

Abstract

The invention discloses an edge computing driven wild animal image compression and encryption transmission method, which relates to the technical field of image processing, and comprises the following steps of: constructing a cache access density curve based on task triggering frequency and cache calling times of edge computing equipment in unit time, recording a time sequence of the maximum available space and the current available space of the cache region, and extracting a cache fluctuation period and an instantaneous compression amplitude; and calculating a difference value between the cache allocation duration and the release delay, and constructing a cache response lag index. According to the method, the multi-dimensional cache state indexes are constructed, and the cache resource pressure bearing coefficients are generated, so that the quantitative perception and prediction of the cache risk are realized. When the cache resource pressure bearing coefficient exceeds the overvoltage risk threshold value, the system automatically executes image frame self-adaptive deletion, the load is relieved, the event continuity is kept, and the stability and the resource utilization efficiency of edge equipment in a high-density image processing scene are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and particularly to a method for compressing and encrypting wildlife images driven by edge computing and transmitting them. Background Art

[0002] The compression and encrypted transmission of wildlife images driven by edge computing refers to a technical solution in which intelligent monitoring terminals deployed in the wild (such as edge computing devices or AI cameras) perform local compression and encryption processing on the captured wildlife image data and then perform remote transmission. Specifically, the system integrates edge nodes with computing capabilities on the terminal device side, can identify animal targets in the image based on object detection algorithms, and uses adaptive image compression technology (such as ROI-based local fidelity compression) to retain high resolution for key regions and compress and downsample background regions, thereby significantly reducing the image volume. At the same time, combined with lightweight encryption algorithms (such as AES or a hybrid mechanism of symmetric encryption and blockchain hashing), data encryption processing is completed locally to ensure the privacy security and anti-tampering ability of the image during transmission. This method not only reduces the bandwidth resources required for image transmission, improves the real-time performance and energy efficiency ratio of the system, but also realizes the efficient and reliable processing of sensitive ecological monitoring data without relying on cloud computing, and is particularly suitable for ecological monitoring scenarios in the wild without public networks and with low power consumption.

[0003] The existing technologies have the following deficiencies: In the existing technologies, image compression and encryption processing usually rely on edge computing nodes to perform serial task scheduling in a time-division multiplexing manner to reduce device resource consumption and adapt to low-power environments. However, in the wild monitoring scenario, when multiple high-speed image acquisition events are continuously triggered (such as wildlife group migrations, cluster activities, etc.), it is easy to cause the edge device cache resources to be under pressure, and the situation where the compression task and the encryption task overlap on the time axis occurs. Since the buffer management mechanism does not have the ability of dynamic concurrent scheduling, problems such as the compression frame being called by the encryption module before it is completed or part of the encrypted data being overwritten by a new task are likely to occur, resulting in the destruction of the image compression structure or the incompleteness of the encrypted content. Eventually, the generated image file cannot be restored during the decoding stage, causing batch invalidation of the image data. Such problems are particularly prominent in high-density event perception scenarios, seriously affecting the logical continuity of the image data and the stability of the system operation.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The objective of the present invention is to provide a method for compressing and encrypting the transmission of wildlife images driven by edge computing. By constructing a cache access density curve, a cache response lag index, and a load increase rate, a comprehensive perception of the dynamic usage status of cache resources is achieved, and a normalized cache resource pressure coefficient is generated through fusion, which is used to quantitatively determine the tightness of cache resources and their potential failure risks. When the cache resource pressure coefficient exceeds the preset risk threshold, the system automatically enters a high-risk processing state, and according to the current resource pressure level, an adaptive uniform deletion strategy is executed on the set of image frames that are frequently triggered, which not only effectively alleviates the cache load but also retains the temporal continuity of events and the integrity of image expression as much as possible, realizing the transformation from the traditional "passive response to cache overflow" to "active prediction and regulation of cache risks", so as to solve the problems in the above background technology.

[0006] To achieve the above objective, the present invention provides the following technical solution: A method for compressing and encrypting the transmission of wildlife images driven by edge computing, including the following steps: S101. Based on the task trigger frequency and cache call times of the edge computing device within a unit time, construct a cache access density curve, record the time series of the maximum available space and the current available space in the cache area, and extract the cache fluctuation period and the instantaneous compression amplitude to reflect the dynamic usage characteristics of the cache resources; S102. During each image compression and encryption process, calculate the difference between the cache allocation duration and the release delay, and construct a cache response lag index; weight and fuse the cache response lag index with the cache access density curve to generate an intermediate feature vector representing the continuity of cache resource occupation and the imbalance of release; S103. Collect the intensity change trend of the image trigger event and the average stacking time of the corresponding processing tasks, calculate the load increase rate of the cache processing, and introduce a task stacking saturation function, where the task stacking saturation function is the Sigmoid function, to perform a non-linear amplification mapping on the cache resource pressure under sudden events and identify the cache limit approaching behavior; S104. Fuse the cache access density curve, the intermediate feature vector, and the load increase rate to construct a multi-dimensional pressure state vector; through normalization processing, map the pressure state vector to a cache resource pressure coefficient ranging from 0 to 1, which is used to characterize the tightness of cache resource usage and the potential failure risk of the edge computing device in the high-frequency image processing scenario; S105. Compare the cache resource pressure coefficient with a preset overpressure risk threshold. When the pressure index exceeds the overpressure risk threshold, it is determined that the edge device enters the high-risk state of cache resources. In the high-risk state of cache resources, based on the current cache resource pressure degree of the edge device, adaptively and uniformly delete the set of image frames of the same continuous event continuously triggered within a short period of time to reduce the cache pressure and prevent the image data from being batch-invalidated due to cache conflicts.

[0007] Preferably, the steps of constructing a cache access density curve and extracting dynamic cache resource usage characteristics further include: Based on the number of image trigger requests received by the image processing module per second and the actual number of compression or encryption processes started, establish the corresponding relationship between the task trigger frequency and the cache call frequency, and use a sliding time window to calculate the cache call density per unit time to form a preliminary cache access density curve; During the use of cache resources, continuously monitor the maximum available space and the current remaining available space in the cache area, generate time series data, and extract the periodic fluctuation characteristics of the cache space change through Fourier transform or wavelet analysis to judge the usage trend and sudden change mode of cache resources; Combine the change slope of the cache call density and the fluctuation gradient of the cache space to calculate the instantaneous compression amplitude of the cache resources per unit time, and judge the degree of sudden cache contraction based on the compression amplitude, so as to obtain the dynamic usage characteristics of cache resources in real time in the image processing link and improve the warning ability of task scheduling and the system stability.

[0008] Preferably, the steps of constructing a cache response latency index and generating an intermediate feature vector are: In each image compression and encryption process, record the time point when the current task applies for cache resources and the time point when the cache resources corresponding to the task are released and completed respectively, calculate the time difference between the two to obtain the cache response latency corresponding to the task; Normalize the cache response latency and the cache access density curve within the time period to which the task belongs, and perform weighted combination according to a preset fusion weight factor to reasonably amplify the weight influence of the latency behavior in the processing scenario with a high task density; Based on the cache response latency index generated during the continuous processing of multiple tasks, statistically obtain three-dimensional indicators of the latency peak, average latency level, and fluctuation amplitude as the basic parameters for measuring the continuity of cache usage and the resource recovery efficiency; Fuse and construct the latency statistical features of the three dimensions into a multi-dimensional intermediate feature vector to characterize the balance of cache resource usage and the release latency distribution characteristics of the edge computing device in the image-intensive processing state, and enhance the accuracy and robustness of the cache pressure state assessment.

[0009] Preferably, the steps of calculating the cache load increase rate and extracting the cache limit approximation behavior include: Continuously monitor the intensity change trend of the image acquisition trigger event, construct an event intensity curve based on the number and frequency of trigger events per unit time, and synchronously record the average waiting time of each image processing task from entering the processing queue to being scheduled for execution, which is used to dynamically characterize the current task backlog degree and processing queuing pressure; Calculate the load increase rate of the cache processing load based on the product of the change gradient of the event intensity curve and the growth rate of the average task backlog time. This load increase rate is used to reflect the rising speed of the cache scheduling pressure of the edge computing device in the high-frequency image input scenario; Construct a task backlog saturation function containing non-linear mapping features, and use a Sigmoid function, exponential curve or other monotonically increasing function to perform weighted amplification processing on the load increase rate, so that when burst image events are frequently triggered, the cache pressure increases exponentially, thereby improving the system's sensitive recognition ability to high-intensity event impacts; Use the processed non-linear load increase value as an input feature, combine it with the usage boundary of the cache resources within the time window, and identify the limit approximation behavior that the cache resources are about to reach the full-load state, which is used in the subsequent pressure-bearing state modeling and risk prediction mechanism to achieve a high-precision response of the edge device to the sudden image processing pressure.

[0010] Preferably, the steps of constructing a multi-dimensional pressure-bearing state vector and generating a cache resource pressure-bearing coefficient include: Perform feature alignment and time window synchronization processing on the cache access density curve representing the cache access frequency, the intermediate feature vector characterizing the cache response lag and release imbalance degree, and the load increase rate reflecting the dynamic growth trend of the image processing pressure, and construct a multi-dimensional feature data set under a unified time scale to ensure time consistency and statistical stability when different types of indicators are fused; Use weighted linear combination, principal component analysis (PCA) or other multi-dimensional dimensionality reduction and fusion methods to perform feature fusion on the above feature data set, generate a pressure-bearing state vector that comprehensively reflects the tightness of the cache resource usage, and combine it with a preset weight dynamic adjustment mechanism, so that in different operating scenarios, the feature dimensions that are more sensitive to the system have higher calculation weights, thereby enhancing the system's ability to identify early signs of cache failure; Normalize the pressure-bearing state vector and map it to a cache resource pressure-bearing coefficient ranging from 0 to 1. The closer the value of the cache resource pressure-bearing coefficient is to 1, the tighter the cache resource usage is, and the closer the edge computing device is to the upper critical state of the cache processing capacity; Take the bearing coefficient of the cached resources as a reference index for risk quantification in the system operation, provide a judgment basis for the subsequent implementation of the image deletion mechanism, and realize the real-time perception and active response to the cache risk in the image processing link.

[0011] Preferably, in the high-risk state of the cached resources, adaptively and uniformly delete the set of image frames of the same continuous event continuously triggered within a short period of time. The specific steps are as follows: After the cached resources enter the high-risk state, calculate the image deletion proportion coefficient that needs to be deleted in the set of image frames according to the current bearing coefficient of the cached resources and the preset overpressure risk threshold. The calculation formula is as follows: , where: is the bearing coefficient of the cached resources, indicating the current cache pressure intensity. The larger the value, the more tense the resources are. ; is the overpressure risk threshold of the cached resources. When , the deletion mechanism is triggered; is the deletion sensitivity adjustment factor, which is used to regulate the non-linear response speed of the deletion intensity. ; is the image deletion proportion coefficient, indicating the proportion that needs to be deleted from the set of image frames. The larger the value, the more deletion. Suppose the set of image frames of the same event continuously triggered within a short period of time is , where N is the original number of frames. According to the image deletion proportion coefficient calculated in step one, determine the number of frames to be retained. The determination expression of the number of frames to be retained is: , where: K is the number of image frames to be retained, calculated from the original number of frames N according to the deletion proportion. indicates rounding up; Subsequently, retain K frames of images in a uniform interval manner to form the set of image frames retained after deletion (the retained frames of the set of image frames are extracted at equal intervals to cover the integrity of the original event as much as possible). , where: , where: is the -th frame of image retained after deletion; represents the index of the image frame retained after deletion. .

[0012] In the above technical solution, the technical effects and advantages provided by the present invention are: The present invention realizes a comprehensive perception of the dynamic usage status of cache resources by constructing a cache access density curve, a cache response lag index, and a load increase rate, and integrates and generates a normalized cache resource pressure coefficient, which is used to quantitatively determine the tension of cache resources and their potential failure risks. When the cache resource pressure coefficient exceeds the preset risk threshold, the system automatically enters a high-risk processing state, and according to the current resource pressure level, performs an adaptive uniform deletion strategy on the high-frequency triggered image frame set, which effectively alleviates the cache load and preserves the temporal continuity of events and the integrity of image expression as much as possible, realizing the transformation from the traditional "passive response to cache overflow" to "active prediction and regulation of cache risks", significantly improving the system stability, image data integrity and cache resource utilization efficiency of edge devices in high-density image processing scenarios, and has good engineering adaptability and promotion and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0014] Figure 1 This is a flow chart of the method for compression and encryption transmission of wildlife images driven by edge computing of the present invention. DETAILED DESCRIPTION

[0015] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0016] The present invention provides Figure 1 The edge computing-driven wildlife image compression and encryption transmission method shown includes the following steps: S101. Construct a cache access density curve based on the task trigger frequency and cache call count per unit time of the edge computing device. Record the time series of the maximum available space and the current available space in the cache area. Extract the cache fluctuation period and instantaneous compression amplitude to reflect the dynamic usage characteristics of the cache resources. The steps of constructing a cache access density curve and extracting cache resource dynamic usage characteristics further include: Based on the number of image trigger requests received by the image processing module per second and the actual number of compression or encryption processes started, establish the corresponding relationship between the task trigger frequency and the cache call frequency, and use a sliding time window to calculate the cache call density per unit time to form a preliminary cache access density curve; During the use of cache resources, continuously monitor the maximum available space and the current remaining available space in the cache area to generate time series data, and extract the periodic fluctuation characteristics of the cache space change through Fourier transform or wavelet analysis to judge the usage trend and sudden change pattern of the cache resources; Combine the slope of the cache call density change and the gradient of the cache space fluctuation to calculate the instantaneous compression amplitude of the cache resources per unit time, and judge the degree of sudden cache contraction based on the compression amplitude, so as to obtain the dynamic usage characteristics of the cache resources in real time in the image processing link, and improve the warning ability and system stability of task scheduling.

[0017] By performing real-time quantification and dynamic modeling on the usage status of cache resources in edge computing devices, construct a "cache access density curve" that reflects the relationship between the image processing task density and the cache call intensity, and cooperate with the time series change of the cache space to comprehensively describe the actual usage status and pressure trend of the cache resources in the device per unit time. Traditional image processing systems often use static thresholds or fixed priority strategies to schedule compression and encryption tasks, lacking accurate perception of the dynamic fluctuations of cache resources, and are prone to problems such as scheduling imbalance, data conflict, or cache overflow in high-frequency trigger scenarios. Through this step, not only can the periodic fluctuations caused by task accumulation in the cache area during continuous image processing be captured, but also the drastic compression behavior (i.e., instantaneous compression amplitude) of the cache space in a short period of time can be quantified, thereby forming an early warning indicator for system load. This provides a decision-making basis for subsequent resource scheduling optimization and task deletion determination, enabling edge devices to have stronger task scheduling flexibility and fault tolerance in the monitoring environment of low power consumption and high data density in the wild, and fundamentally improving the stability of the system and the integrity of image data processing.

[0018] S102. During each image compression and encryption process, calculate the difference between the cache allocation duration and the release delay to construct a cache response lag index; weight and fuse the cache response lag index with the cache access density curve to generate an intermediate feature vector representing the continuity of cache resource occupation and the degree of uneven release; The steps to construct the cache response lag index and generate the intermediate feature vector are as follows: During each image compression and encryption process, record the time point when the current task requests cache resources and the time point when the cache resources corresponding to the task are released and completed respectively, calculate the time difference between the two to obtain the cache response lag corresponding to the task; Normalize the cache response latency with the cache access density curve within the time period to which the task belongs, and perform weighted combination according to the preset fusion weight factor to reasonably amplify the weight influence of the latency behavior in the processing scenario with high task density; Based on the cache response latency index generated during the continuous processing of multiple tasks, three-dimensional indexes including the latency peak value, average latency level, and fluctuation amplitude are statistically obtained as the basic parameters for measuring the continuity of cache usage and resource recycling efficiency; Fuse the latency statistical features in three dimensions to construct a multi-dimensional intermediate feature vector to characterize the balance of cache resource usage and the release latency distribution characteristics of the edge computing device in the image-intensive processing state, enhancing the accuracy and robustness of the cache pressure state assessment.

[0019] By quantifying the timeliness of cache resource usage of image processing tasks in edge devices, establish a "cache response latency index" with dynamic perception ability, and fuse it with the current cache access density curve of the system, so as to accurately evaluate the continuous occupancy characteristics of cache resources and the imbalance of release delay. Traditional image processing systems usually only focus on whether the task is successfully completed, ignoring the actual occupancy efficiency of cache resources during the task execution process. Especially in the edge computing scenario where tasks are triggered frequently and at high density, if the cache is not released for a long time or the release efficiency is low after cache allocation, it will cause subsequent tasks to wait or even be overwritten when calling the cache, seriously affecting the continuity and reliability of the entire processing link.

[0020] By recording the "resource allocation time" and "resource release completion time" of each image compression and encryption task at the cache level, and calculating the duration difference between the two, the occupied duration of the cache in this task can be quantified, thus constructing an index reflecting the resource occupancy efficiency of a single task, that is, the cache response latency. This index itself already has the ability to identify latency tasks and monitor the system congestion trend. However, in a high-concurrency scenario, the meaning of a single latency value will be diluted. Therefore, this method further performs weighted fusion of the latency with the cache access density curve, so that when tasks are severely piled up and cache calls are frequent, the weight of the latency behavior is amplified, thus more sensitively reflecting the system risks that may be brought by untimely resource release.

[0021] Finally, the "intermediate feature vector" generated through this fusion process can comprehensively represent the continuity, latency degree, and release balance of cache resource usage among multiple tasks, serving as the key input variable for the subsequent system to judge the cache pressure state, trigger scheduling adjustments, or image deletion strategies. This method not only improves the resolution of the cache resource state assessment but also provides the edge device with the ability of active adjustment based on behavior feature feedback, significantly enhancing the stability and robustness of the system during the peak period of image processing.

[0022] S103. Collect the intensity change trend of the image acquisition trigger event and the average queuing time of the corresponding processing task, calculate the load increase rate of the cache processing, and introduce a task queuing saturation function to perform a non-linear amplification mapping on the cache resource pressure under emergency events to identify the cache limit approaching behavior; The steps of calculating the cache load increase rate and extracting the cache limit approaching behavior include: Continuously monitor the intensity change trend of the image acquisition trigger event, construct an event intensity curve based on the number and frequency of trigger events per unit time, and synchronously record the average waiting time of each image processing task from entering the processing queue to being scheduled for execution, which is used to dynamically characterize the current task queuing degree and processing queuing pressure; Calculate the load increase rate of the cache processing based on the product of the change gradient of the event intensity curve and the growth rate of the average task queuing time. This load increase rate is used to reflect the rising speed of the cache scheduling pressure of the edge computing device in the high-frequency image input scenario; Construct a task queuing saturation function with non-linear mapping characteristics, and use a Sigmoid function, exponential curve or other monotonically increasing function to perform weighted amplification processing on the load increase rate, so that when emergency image events are frequently triggered, the cache pressure increases exponentially, thereby improving the system's sensitive recognition ability to high-intensity event impacts; Use the processed non-linear load increase value as an input feature, combined with the usage boundary of the cache resources within the time window, to identify the limit approaching behavior that the cache resources are about to reach the full load state, which is used in the subsequent pressure-bearing state modeling and risk prediction mechanism to achieve a high-precision response of the edge device to the emergency image processing pressure.

[0023] In order to identify the "limit approaching behavior" of cache resources when they are about to reach full load, first, the load increase rate after being processed by a non-linear function is used as the core input feature, which can already represent the pressure growth rate of the current system under the accumulation of image processing tasks. Then, the usage boundary data of cache resources are collected within consecutive time windows, including indicators such as the space already used in the cache per unit time, the remaining available space, and the gradient of the usage rate change. By constructing a joint change surface of cache usage rate and time, it is judged whether the cache resources show a continuous trend of approaching the upper limit. Specifically, when the non-linear load increase value is at a high level (such as above 0.8), and the cache usage rate shows a monotonic increase within the sliding time window, and the growth slope exceeds the set threshold, it can be considered that the system has entered the cache resource limit approaching interval. In addition, by setting multiple overpressure critical bands (such as the warning area, the alarm area, and the critical area), and combining the comparison result of the cache release speed and the new task entry frequency, the accuracy of the judgment can be further improved. When the above comprehensive conditions are met, the system determines that the cache is in the "limit approaching behavior" state, thereby triggering strategies such as image task deletion, cache reallocation, or processing throttling to achieve early intervention in the cache overflow risk. This mechanism has high dynamicity and predictive ability and is an important technical link to ensure the stability of the edge device cache.

[0024] The role of this step is to provide an edge computing device with a cache load dynamic evaluation mechanism based on the collaborative modeling of image trigger event intensity and task accumulation characteristics, so as to achieve non-linear enhanced recognition of cache resource pressure changes and early prediction of the "limit approaching state". The image monitoring system deployed in the wild environment faces highly uncertain trigger event intensities, such as situations like mass animal migrations and sudden cluster activities. These events often generate a large amount of image data in a short period of time, causing the edge device to bear a sudden high-density task load. If the current cache resource pressure of the system cannot be evaluated in real time and its state approaching the collapse boundary cannot be recognized, it may lead to serious consequences such as image data processing failure, cache overflow, and batch data loss.

[0025] By continuously collecting the intensity change trend of image trigger events and combining with the average waiting time after the image processing task enters the processing queue, the backlog degree of tasks in the current stage of the system is reflected. Further, the "cache load increase rate" is calculated jointly using the change slope of the event intensity and the growth rate of the backlog time. This indicator depicts the growth trend of the system processing pressure per unit time. However, in actual scenarios, the impact of pressure often does not develop linearly. Especially in sudden high-intensity events, the load critical state often approaches the resource bottleneck rapidly in an exponential or saturation curve manner. Therefore, on this basis, the "task backlog saturation function" is introduced, and a non-linear amplification mapping of the load increase rate is performed using functions such as the Sigmoid function or the exponential growth model, so as to more sensitively identify the "limit approaching behavior" where the cache resources are about to reach the bearing limit. This identification mechanism can significantly improve the perception accuracy of the cache failure risk in high-pressure scenarios, enabling the device to take countermeasures (such as task deletion, delayed scheduling) in advance before the load truly gets out of control, and has important engineering practical value and system robustness guarantee ability, and is one of the key links to ensure the continuity of the image processing chain and data integrity.

[0026] S104. Integrate the cache access density curve, the intermediate feature vector, and the load increase rate to construct a multi-dimensional pressure-bearing state vector; map this pressure-bearing state vector to a cache resource pressure-bearing coefficient ranging from 0 to 1 through normalization processing, which is used to characterize the tightness of cache resource usage and the potential failure risk of the edge computing device in the high-frequency image processing scenario. The steps of constructing a multi-dimensional pressure-bearing state vector and generating a cache resource pressure-bearing coefficient include: Perform feature alignment and time window synchronization processing on the cache access density curve representing the cache access frequency, the intermediate feature vector characterizing the cache response lag and release imbalance degree, and the load increase rate reflecting the dynamic growth trend of the image processing pressure, and construct a multi-dimensional feature data set under a unified time scale to ensure time consistency and statistical stability when fusing different types of indicators. Adopt weighted linear combination, principal component analysis (PCA) or other multi-dimensional dimensionality reduction and fusion methods to perform feature fusion on the above feature data set, generate a pressure-bearing state vector comprehensively reflecting the tightness of cache resource usage, and combine a preset weight dynamic adjustment mechanism to make the feature dimensions more sensitive to the system have higher calculation weights in different operating scenarios, so as to enhance the system's ability to identify early signs of cache failure. When fusing the multi-dimensional feature data set representing the cache resource state (including the cache access density curve, the intermediate feature vector, and the cache load increase rate), using principal component analysis (PCA) is an efficient and interpretable dimensionality reduction method, which can compress redundant feature dimensions while retaining the main information, and improve the calculation efficiency and response speed in the edge computing environment. The specific process is as follows: First, preprocess the original feature dataset. Since the feature dimensions have different units, for example, the cache access density is the number of accesses per second, the intermediate feature vector is a dimensionless statistic, and the load increase rate is a ratio - type time - varying index, it is necessary to unify the standard. Usually, Z - score standardization is adopted, that is, subtract the mean of each feature vector and then divide by its standard deviation, so that all features are transformed into a form with zero mean and unit variance, thus eliminating the interference of dimensional differences on the principal component analysis results.

[0027] Second, construct the covariance matrix based on the standardized feature matrix. The covariance matrix describes the linear correlation relationship between different features. Essentially, it measures whether there is a consistent increasing or decreasing trend between any two - dimensional features in the sample data. For example, if the cache load increase rate is highly positively correlated with the intermediate lag feature, it means that when the system pressure rises, resource release is also likely to lag.

[0028] Next, perform eigenvalue decomposition on the covariance matrix to obtain several sets of eigenvalues and their corresponding eigenvectors. Each eigenvalue represents the proportion of the variance explained by the principal component (i.e., the eigenvector direction) in the original data, and the eigenvector itself represents a new principal component direction. After sorting these eigenvalues from large to small, select the top k principal components according to the cumulative contribution rate (such as achieving 90% information retention) to retain the most information with the least number of dimensions.

[0029] Then, use the selected principal component vectors to linearly project the original data, that is, map the original high - dimensional features into the newly selected k - dimensional subspace to generate a new vector, called the pressure - bearing state vector. This vector comprehensively expresses the most important change trends among the original features and can retain key system operation state information while significantly reducing the dimensions.

[0030] Finally, the obtained pressure - bearing state vector, as a compact expression of the cache resource usage status, can be further normalized or mapped through a risk function to generate a cache pressure - bearing index between 0 and 1. This index not only reflects the current resource pressure degree of the system but also can be used as the core criterion for subsequent modules such as image deletion and adaptive scheduling. By introducing PCA, the system can maintain an efficient, stable, and risk - self - aware operating state while processing large - scale image trigger tasks.

[0031] Normalize the pressure - bearing state vector and map it to a cache resource pressure - bearing coefficient ranging from 0 to 1. The closer the value of the cache resource pressure - bearing coefficient is to 1, the more tense the cache resource usage is, and the closer the edge computing device is to the upper - limit critical state of the cache processing capacity; Take the cache resource bearing coefficient as a risk quantification reference index in system operation, provide a judgment basis for subsequent implementation of the image deletion mechanism, and realize real-time perception and active response to cache risks in the image processing link.

[0032] Through the fusion modeling of multi-source heterogeneous indicators, construct a cache resource bearing coefficient that can dynamically quantify and uniformly measure the system cache pressure, so as to provide accurate operation status evaluation and risk identification basis for edge computing devices in high-frequency image processing scenarios. During the process of edge devices executing image compression and encryption tasks, the occupancy of cache resources is a dynamic process affected by multiple factors. A single indicator (such as the number of tasks or cache utilization rate) often cannot accurately capture the full picture of the device facing the risk of cache failure. Therefore, this step proposes to comprehensively model three types of key indicators. Among them, the three types of key indicators include the cache access density curve, the intermediate feature vector (reflecting the continuity of cache occupancy and the lag of release), and the load increase rate (reflecting the trend of sudden task accumulation).

[0033] First, by performing time alignment and feature space unification processing on these three types of features, construct a multi-dimensional bearing state vector under a unified time scale, which solves the problems of index asynchrony and scale non-uniformity; second, use methods such as weighted fusion and principal component analysis to aggregate high-dimensional feature data into a representative comprehensive vector, enabling the system to automatically adjust weights according to the actual risk sensitivity under different operating states and improve the ability to predict potential failures. Finally, through normalization processing, map this state vector to the cache resource bearing coefficient in the interval [0, 1]. The closer the cache resource bearing coefficient is to 1 numerically, the more tense the current cache resources are, and the closer the device is to the failure boundary.

[0034] The greatest value of the cache resource bearing coefficient is that it transforms multiple complex time-series features into an interpretable, traceable, and drivable quantification index, which can be used as the core trigger condition for internal task scheduling optimization, image deletion strategy initiation, risk notification and alarm mechanisms in the system, enabling edge computing devices to have highly intelligent load self-perception and risk active intervention capabilities in the wild complex environment, and significantly improving the stability, robustness, and task processing continuity of the system.

[0035] S105. Compare the cache resource bearing coefficient with the preset overpressure risk threshold. When the bearing index exceeds the overpressure risk threshold, it is determined that the edge device enters the high-risk state of cache resources. In the high-risk state of cache resources, based on the current cache resource pressure degree of the edge device, adaptively and uniformly delete the set of image frames of the same continuous event continuously triggered within a short time to reduce the cache pressure and prevent batch failure of image data due to cache conflicts; In the high-risk state of cache resources, adaptively and uniformly delete a set of image frames of the same continuous event continuously triggered within a short period of time. The specific steps are as follows: After the cache resources enter the high-risk state, calculate the image deletion proportion coefficient that needs to be deleted in the set of image frames according to the current cache resource pressure-bearing coefficient and the preset overpressure risk threshold. The calculation formula is as follows: , where: is the cache resource pressure-bearing coefficient, indicating the current cache pressure intensity. The larger the value, the more tense the resources are, ; is the overpressure risk threshold of the cache resources. When , the deletion mechanism is triggered; is the deletion sensitivity adjustment factor, used to regulate the non-linear response speed of the deletion intensity, ; is the image deletion proportion coefficient, indicating the proportion that needs to be deleted from the set of image frames. The larger the value, the more deletion; The function of this step is to convert the cache overpressure degree into a measurable deletion decision variable, so that the system can intelligently adjust the image processing load based on the current resource pressure, avoid adopting extreme strategies such as "full deletion" or "full retention", and improve the processing accuracy and cache release efficiency.

[0036] Let the set of image frames of the same event continuously triggered within a short period of time be , where N is the original number of frames. According to the image deletion proportion coefficient calculated in step one, determine the number of frames to be retained. The determination expression of the number of frames to be retained is: , where: K is the number of image frames to be retained, calculated from the original number of frames N according to the deletion proportion, indicates rounding up; Subsequently, retain the K frames of images in a uniform interval manner to form the set of image frames retained after deletion (the retained frames of the set of image frames are extracted at equal intervals to cover the integrity of the original event as much as possible), , where: , where: is the -th frame of image retained after deletion; represents the index of the image frame retained after deletion, ; The purpose of this formula is to select K frames at equal intervals from a large number of non-uniformly spaced and suddenly generated image frames to maximize the time coverage of the event.

[0037] For example, if: There are a total of N=100 image frames that need to be retained K = 10 pictures, hope from the beginning to the end, every Keep one frame.

[0038] The expression This is the same proportional mapping extraction, but by "rounding up" to ensure that the frame number is a legal integer to prevent decimal frame numbers.

[0039] The purpose of this step is to apply the deletion ratio decision to the actual image sequence and ensure the temporal continuity and spatial representativeness of event expression by extracting frames at equal intervals. At the same time, it greatly reduces memory usage and scheduling pressure, alleviates cache resource bottlenecks while ensuring the validity of ecological monitoring data, and improves the overall stability and response elasticity of the system.

[0040] The purpose of this step is to build an edge image task load reduction mechanism with real-time feedback and adjustment capabilities by comparing the threshold of the cache resource pressure coefficient and linking it with the image adaptive deletion mechanism. This allows for precise intervention when the cache resources are at risk of overload, ensuring continuous and stable operation of the system. In edge computing environments, especially image acquisition systems used for wild animal monitoring, there is often a high concentration of image triggering events (such as animal group migration, gathering activities, etc.) within a certain period of time. This will cause a large number of compression and encryption tasks to be triggered simultaneously, causing cache resources to be consumed rapidly in a very short period of time. Once the cache exceeds its carrying capacity, not only will subsequent tasks be unable to execute, but it may also cause the early image data processing to be interrupted or overwritten, resulting in permanent loss of batch image data.

[0041] To avoid these issues, this step first compares the cache resource pressure coefficient (reflecting the system's current cache usage stress) with a preset overvoltage risk threshold to determine in real time whether the system has entered a "cache resource high-risk state." Once the cache resource pressure coefficient exceeds the overvoltage risk threshold, the device is deemed to be on the verge of overvoltage, and the system immediately initiates an adaptive response mechanism. This mechanism uses the current pressure level as an input parameter and prunes image frames (large numbers of images generated by the same event) that are triggered continuously within a short period of time. This pruning process uses evenly spaced frame extraction to ensure that the temporal continuity of the event and the image representativeness are preserved as much as possible while reducing cache load.

[0042] The core value of this measure lies in: on the one hand, it can prevent data processing crashes caused by a full cache; on the other hand, it avoids serious loss of image information under the "blind deletion" strategy. Through the linkage adjustment of the deletion ratio and cache pressure, the system realizes an active cropping mechanism for image processing driven by resource status perception, which has high intelligence, high practicality and high engineering deployability, and is one of the important guarantee means for the stable operation of the edge intelligent system.

[0043] Through the "Edge Computing Driven Wildlife Image Compression and Encryption Transmission Method" provided by the present invention, the problems of cache conflicts and data batch invalidation caused by serial scheduling of image compression and encryption tasks under resource-constrained conditions in the prior art are effectively solved. This method comprehensively perceives the dynamic usage status of cache resources by constructing a cache access density curve, a cache response lag index, and a load increase rate, and fuses them to generate a normalized cache resource pressure coefficient to realize the quantitative determination of the tightness of cache resources and the potential failure risk. When the cache resource pressure coefficient of the system exceeds the risk threshold, it automatically enters the high-risk processing state, and based on the current resource pressure level, an adaptive uniform deletion strategy is implemented for the set of image frames triggered by high-frequency events, which not only reduces the cache load but also retains the time continuity of events and the integrity of image expression. This solution realizes the transformation from "passive cache overflow response" to "active cache risk prediction and control", significantly improves the system stability, image data integrity and cache resource utilization efficiency of edge devices in high-density image processing scenarios, and has good engineering application prospects and promotion value.

[0044] The above formulas are all dimensionless and take their numerical calculations. The formula is a formula obtained by collecting a large amount of data for software simulation to approximate the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0045] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of the claims of the present invention.

[0046] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0047] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0048] In several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0049] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0050] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0051] If the 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 technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0052] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claimed rights.

Claims

1. Edge-computing-driven wildlife image compression and encrypted transmission method, characterized in that, It includes the following steps: S101. Construct a cache access density curve based on the task trigger frequency and cache call times of the edge computing device within a unit time, record the time series of the maximum available space and the current available space in the cache area, and extract the cache fluctuation period and the instantaneous compression amplitude; S102. Calculate the difference between the cache allocation duration and the release delay, construct a cache response lag index, and perform weighted fusion on the cache response lag index and the cache access density curve to generate an intermediate feature vector representing the continuity of cache resource occupation and the degree of uneven release; S103. Collect the intensity change trend of the image trigger event and the average stacking time of the corresponding processing tasks, calculate the load increase rate of cache processing, and introduce a task stacking saturation function, where the task stacking saturation function is a Sigmoid function, to perform a non-linear amplification mapping on the cache resource pressure under sudden events and identify the cache limit approaching behavior; S104. Integrate the cache access density curve, the intermediate feature vector, and the load increase rate to construct a multi-dimensional pressure-bearing state vector, and map the pressure-bearing state vector to a cache resource pressure coefficient ranging from 0 to 1 through normalization processing; S105. When the pressure index exceeds the overpressure risk threshold, it is determined that the edge device enters the high-risk state of cache resources. In the high-risk state of cache resources, based on the current cache resource pressure degree of the edge device, adaptively and uniformly delete the set of image frames of the same continuous event continuously triggered within a short time.

2. The edge-computing-driven wildlife image compression and encrypted transmission method according to claim 1, wherein, The steps of constructing the cache access density curve and extracting the dynamic usage characteristics of cache resources further include: Based on the number of image trigger requests received by the image processing module per second and the actual number of compression or encryption processes started, establish the corresponding relationship between the task trigger frequency and the cache call frequency, and use a sliding time window to calculate the cache call density within a unit time to form a preliminary cache access density curve; During the use of cache resources, continuously monitor the maximum available space and the current remaining available space in the cache area to generate time series data, and extract the periodic fluctuation characteristics of the cache space change through Fourier transform or wavelet analysis; Combine the change slope of the cache call density and the fluctuation gradient of the cache space to calculate the instantaneous compression amplitude of the cache resources within a unit time, and judge the degree of sudden cache contraction based on the compression amplitude.

3. The edge-computing-driven wildlife image compression and encrypted transmission method according to claim 1, wherein, The steps of constructing the cache response lag index and generating the intermediate feature vector are as follows: In each image compression and encryption processing process, record the time point when the current task requests cache resources and the time point when the cache resources corresponding to the task are released and completed respectively, calculate the time difference between the two to obtain the cache response lag corresponding to the task; Normalize the cache response lag and the cache access density curve within the time period to which the task belongs, and perform weighted combination according to the preset fusion weight factor; Based on the cache response lag index generated during the continuous processing of multiple tasks, statistically obtain three-dimensional indexes of the lag peak value, the average lag level, and the fluctuation amplitude as the basic parameters for measuring the continuity of cache use and the resource recovery efficiency; Fuse the lag statistical features in three dimensions to construct a multi-dimensional intermediate feature vector to characterize the balance of cache resource usage and the release lag distribution characteristics of edge computing devices in the state of intensive image processing.

4. The edge-computing-driven wildlife image compression and encrypted transmission method according to claim 1, wherein, The steps of calculating the cache load increase rate and extracting the cache limit approximation behavior include: Continuously monitor the intensity change trend of the image acquisition trigger event, construct an event intensity curve based on the number and frequency of trigger events per unit time, and synchronously record the average waiting time of each image processing task from entering the processing queue to being scheduled for execution; Calculate the load increase rate of the cache processing load based on the product of the change gradient of the event intensity curve and the growth rate of the average task stacking time; Construct a task stacking saturation function containing non-linear mapping features and perform weighted amplification processing on the load increase rate; Use the processed non-linear load increase value as an input feature, and combine the usage boundary of the cache resources within the time window to identify the limit approximation behavior that the cache resources are about to reach the full load state.

5. The edge-computing-driven wildlife image compression and encrypted transmission method according to claim 1, wherein, The steps of constructing a multi-dimensional pressure-bearing state vector and generating a cache resource pressure-bearing coefficient include: Perform feature alignment and time window synchronization processing on the cache access density curve representing the cache access frequency, the intermediate feature vector characterizing the cache response lag and release imbalance degree, and the load increase rate reflecting the dynamic growth trend of the image processing pressure, and construct a multi-dimensional feature data set under a unified time scale; Perform feature fusion on the feature data set to generate a pressure-bearing state vector comprehensively reflecting the tightness of cache resource usage, and combine a preset weight dynamic adjustment mechanism to set a higher calculation weight for the feature dimension that is more sensitive to the impact; Perform normalization processing on the pressure-bearing state vector, map it to a cache resource pressure-bearing coefficient ranging from 0 to 1, and use the cache resource pressure-bearing coefficient as a risk quantification reference index.

6. The edge computing-driven wildlife image compression and encrypted transmission method according to claim 1, characterized in that, In the high-risk state of cache resources, adaptively and uniformly delete the set of image frames of the same continuous event continuously triggered within a short period of time. The specific steps are as follows: After the cached resources enter the high-risk state, calculate the image deletion ratio coefficient that needs to be deleted from the set of image frames according to the current pressure-bearing coefficient of the cached resources and the preset overpressure risk threshold. The calculation formula is as follows: , where: is the pressure-bearing coefficient of the cached resources, indicating the current cache pressure intensity, ; is the overpressure risk threshold of the cached resources, and the deletion mechanism is triggered when ; is the deletion sensitivity adjustment factor, which is used to regulate the non-linear response speed of the deletion intensity, ; is the image deletion ratio coefficient, indicating the ratio that needs to be deleted from the set of image frames.

7. The edge-computing-driven wildlife image compression and encrypted transmission method according to claim 6, wherein Let the set of image frames of the same event continuously triggered within a short period be , where N is the original number of frames. Determine the number of frames to be retained according to the image deletion ratio coefficient . The determination expression for the number of frames to be retained is: , where: K is the number of image frames to be retained, calculated from the original number of frames N according to the deletion ratio, represents rounding up.

8. The edge computing-driven wildlife image compression and encrypted transmission method according to claim 7, wherein Retain in a uniform interval manner K frame images to form a set of retained image frames after deletion , , where: , where: is the th frame image retained after deletion; represents the index of the image frame retained after deletion, .

Images