A method and system for adaptive compression and breakpoint resume scheduling of wildlife images

Abstract

This invention relates to the field of image processing and data transmission technology, specifically disclosing a method and system for adaptive compression and breakpoint resume scheduling of wildlife images. This invention acquires multi-dimensional data of wildlife images in real time, constructs a quantitative evaluation system, and dynamically optimizes data transmission and processing. First, raw data is collected from the wildlife image acquisition terminal, covering core information such as terminal identification, network status, and image content. Then, value assessment coefficients and breakpoint resume scheduling coefficients are calculated sequentially, and an image compression strategy is obtained accordingly. Adaptive compression and breakpoint resume transmission are then executed. Finally, verification, retransmission, and image enhancement optimization are performed on the server side, achieving closed-loop optimization across the entire link from terminal to server. This ensures high-definition transmission and reliable delivery of rare species images even in weak network environments, improving the efficiency and reliability of the monitoring system.

Description

Technical Field

[0001] This invention relates to the field of image processing and data transmission technology, and in particular to an adaptive compression and breakpoint resume scheduling method and system for wildlife images. Background Technology

[0002] In the field of wildlife monitoring, particularly in wildlife ecological research in nature reserves, wetland parks, and migratory bird flyways, the deployment of large-scale image acquisition terminals (such as solar-powered 4G infrared cameras and high-definition network PTZ cameras) has become an important technical means for wide-area, continuous, and automated monitoring of wildlife. Researchers hope to acquire a large number of high-resolution wildlife images to analyze wildlife population sizes, habitat ranges, migration routes, and behavioral patterns. However, when deploying large-scale wildlife image monitoring systems in remote forest areas and wetlands, monitoring points typically rely on 4G / 5G or high-throughput satellite communication. These communication transmissions often experience drastic bandwidth fluctuations (ranging from a few KB / s to a few MB / s) and high data costs. Furthermore, the existing "transmit first, process later" model suffers from severe congestion and latency due to large image transmissions. Servers are also prone to overload when handling tens of thousands of concurrent terminal connections due to processing large amounts of invalid data (such as aerial images without wildlife). Secondly, existing technologies often employ uniform and fixed compression or transmission strategies. In actual scientific research, high-resolution images of rare species (such as the Chinese Merganser and Cabot's Tragopan) are far more valuable than those of common species. Existing solutions cannot dynamically allocate transmission resources based on image content, leading to the potential loss of identification features due to over-compression in weak network conditions or loss due to untimely retransmission. Therefore, an adaptive compression and breakpoint resumption scheduling method for wildlife images is needed to address these issues. Summary of the Invention

[0003] The purpose of this invention is to provide an adaptive compression and breakpoint resume scheduling method for wildlife images, comprising: The system acquires image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal, wherein the image acquisition information includes terminal identification information and terminal network transmission status information. Basic information about wild animals is obtained based on the content information of the wild animal images, and a value assessment coefficient is obtained based on the basic information about wild animals. The network bandwidth data and image transmission priority are obtained based on the terminal network transmission status information, and the breakpoint resume scheduling coefficient is obtained based on the network bandwidth data and the image transmission priority. The image compression strategy is obtained based on the value assessment coefficient and the breakpoint resume scheduling coefficient. Based on the image compression strategy and the breakpoint resume scheduling coefficient, the wildlife images acquired by the wildlife image acquisition terminal are adaptively compressed and resumable from breakpoints to obtain wildlife images after adaptive compression and resumable from breakpoints. The identification server is obtained based on the terminal identification information. The wildlife image is received and verified through the identification server to obtain the verification result. A coordination optimization strategy is obtained based on the verification result and the image compression strategy. The wildlife image is then coordinated and optimized based on the coordination optimization strategy to obtain the wildlife image optimization result.

[0004] Preferably, the step of obtaining basic information about wild animals based on the content information of the wild animal images, and obtaining a value assessment coefficient based on the basic information about wild animals, includes: Based on the content information of the wildlife images, the wildlife recognition results in the images are obtained through a preset wildlife image recognition model, and the corresponding basic wildlife information is obtained from a preset wildlife information database based on the wildlife recognition results. The basic wildlife information includes the wildlife name, rarity level, and protection level. Based on the rarity level and the protection level, the rarity score of wild animals is obtained through a preset rarity score mapping table; Based on the content information of the wildlife images, a preset wildlife feature extraction model is used to obtain the feature clarity score and behavioral salience score of the wildlife in the images, and the wildlife feature salience index is obtained based on the feature clarity score and behavioral salience score. Initial value assessment factors are obtained based on the wildlife rarity score and the wildlife characteristic significance index; Obtain a preset historical image value benchmark threshold, and obtain a value assessment coefficient by ratio between the initial value assessment factor and the historical image value benchmark threshold.

[0005] Preferably, the step of obtaining network bandwidth data and image transmission priority based on the terminal network transmission status information, and obtaining the breakpoint resumption scheduling coefficient based on the network bandwidth data and image transmission priority, includes: Based on the terminal network transmission status information, multiple real-time uplink bandwidth values ​​of the wildlife image acquisition terminal within a preset time window are obtained, and a network transmission stability index is obtained based on the multiple real-time uplink bandwidth values. The device type and task priority of the wildlife image acquisition terminal are obtained based on the terminal identification information, and the image transmission priority is obtained through a preset priority decision tree based on the device type and the task priority. Obtain the amount of data already transmitted and the total amount of data for the wildlife images currently to be transmitted, and obtain the image transmission progress based on the amount of data already transmitted and the total amount of data. The initial scheduling factor is obtained based on the network transmission stability index, the image transmission priority, and the image transmission progress. Obtain the system's preset scheduling baseline coefficient, and obtain the breakpoint resume scheduling coefficient by ratio of the initial scheduling factor and the scheduling baseline coefficient.

[0006] Preferably, the step of obtaining the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient includes: The system obtains the current remaining battery power and storage space of the wildlife image acquisition terminal, and determines the terminal resource stress based on the current remaining battery power and storage space. Based on the value assessment coefficient, the breakpoint resume scheduling coefficient, and the terminal resource stress, the initial compression algorithm and the initial compression rate are determined through a preset compression strategy decision matrix. A preset image quality loss threshold is obtained, and the wildlife image is simulated and compressed according to the initial compression algorithm and the initial compression ratio to obtain the simulated compressed image quality loss value. Determine whether the image quality loss value is greater than the image quality loss threshold; If the image quality loss value is not greater than the image quality loss threshold, then the initial compression algorithm and the initial compression ratio are used as the image compression strategy. If the image quality loss value is greater than the image quality loss threshold, the initial compression ratio is adjusted by gradient descent until the image quality loss value corresponding to the adjusted compression ratio is not greater than the image quality loss threshold, and an image compression strategy is obtained based on the adjusted compression ratio and the initial compression algorithm.

[0007] Preferably, the step of adaptively compressing and resuming interrupted transmission of the wildlife images acquired by the wildlife image acquisition terminal according to the image compression strategy and the interrupted transmission resumption scheduling coefficient, to obtain the wildlife images after adaptive compression and interrupted transmission resumption scheduling, includes: According to the image compression strategy, the original wildlife images acquired by the wildlife image acquisition terminal are compressed to obtain compressed wildlife image data packets, and the compression parameters are recorded. Based on the breakpoint resume scheduling coefficient, determine whether the compressed wildlife image data packet needs to enable the breakpoint resume mode. If the breakpoint resume scheduling coefficient is less than the preset resume scheduling threshold, then the single transmission mode is adopted to send the compressed wildlife image data packet to the identification server all at once. If the breakpoint resume scheduling coefficient is greater than or equal to the preset resume scheduling threshold, the breakpoint resume mode is enabled, the compressed wildlife image data packet is divided into multiple data blocks, and sent sequentially according to the preset transmission order, while recording the transmission status of each data block and the information of the data blocks that have been confirmed to be received. When the connection is restored after a transmission interruption, the transmission continues from the first unacknowledged data block according to the confirmed data block information, until all data blocks are sent, resulting in a wildlife image after adaptive compression and breakpoint resume scheduling.

[0008] Preferably, the steps of receiving the wildlife image through the identification server and verifying it to obtain a verification result, obtaining a coordination optimization strategy based on the verification result and the image compression strategy, and performing coordination optimization on the wildlife image according to the coordination optimization strategy to obtain an optimized wildlife image include: The identification server receives the wildlife images after adaptive compression and breakpoint resume scheduling, and decompresses and verifies the integrity of the received wildlife image data according to the compression parameters in the image compression strategy, to obtain the decompressed wildlife images and integrity verification results. Based on the decompressed wildlife image and the integrity verification result, an image quality verification result is obtained, wherein the image quality verification result includes whether the image is complete, whether there is a decompression error, and whether the image resolution meets a preset standard; If any of the image quality verification results fails, the verification result is determined to require optimization, and a retransmission request is sent to the wildlife image acquisition terminal to request the retransmission of the corresponding original wildlife image or compressed wildlife image data packet. If all the image quality verification results pass, then based on the decompressed wildlife image and the basic information of the wildlife, an image optimization parameter set is obtained through a preset image enhancement model, wherein the image optimization parameter set includes contrast adjustment parameters, sharpness adjustment parameters and color correction parameters; The decompressed wildlife image is enhanced and optimized according to the image optimization parameter set to obtain the wildlife image optimization result. The image compression strategy, the integrity verification result, and the image optimization parameter set are stored as a coordinated optimization strategy for use as a reference for optimizing subsequently received wildlife images.

[0009] This application also provides an adaptive compression and breakpoint resume scheduling system for wildlife images, including: The first acquisition module is used to acquire image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal, wherein the image acquisition information includes terminal identification information and terminal network transmission status information; The second acquisition module is used to acquire basic information about wild animals based on the content information of the wild animal images, and to acquire a value assessment coefficient based on the basic information about wild animals. The third acquisition module is used to acquire network bandwidth data and image transmission priority based on the terminal network transmission status information, and to acquire breakpoint resume scheduling coefficient based on the network bandwidth data and the image transmission priority. The fourth acquisition module is used to acquire the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient. The fifth acquisition module is used to adaptively compress and schedule breakpoint resume transmission of the wildlife images acquired by the wildlife image acquisition terminal according to the image compression strategy and the breakpoint resume transmission scheduling coefficient, so as to obtain the wildlife images after adaptive compression and breakpoint resume transmission scheduling. The sixth acquisition module is used to acquire an identification server based on the terminal identification information, receive the wildlife image through the identification server and verify it to obtain a verification result, acquire a coordination optimization strategy based on the verification result and the image compression strategy, and perform coordination optimization on the wildlife image based on the coordination optimization strategy to obtain an optimized wildlife image result.

[0010] Preferably, the second acquisition module includes: The first acquisition unit is used to acquire the wild animal recognition result in the image based on the wild animal image content information through a preset wild animal image recognition model, and to acquire the corresponding wild animal basic information from a preset wild animal information database based on the wild animal recognition result, wherein the wild animal basic information includes the wild animal name, rarity level and protection level; The second acquisition unit is used to acquire the rarity score of wild animals according to the rarity level and the protection level through a preset rarity score mapping table; The third acquisition unit is used to acquire the feature clarity score and behavior salience score of the wild animals in the image based on the content information of the wild animal image, through a preset wild animal feature extraction model, and to acquire the wild animal feature salience index based on the feature clarity score and behavior salience score. The fourth acquisition unit is used to acquire initial value assessment factors based on the wildlife rarity score and the wildlife characteristic salience index; The fifth acquisition unit is used to acquire a preset historical image value benchmark threshold, and to acquire a value assessment coefficient by ratio based on the initial value assessment factor and the historical image value benchmark threshold.

[0011] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0012] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0013] The beneficial effects of this application are as follows: This invention acquires multi-dimensional data of wildlife images in real time, constructs a quantitative evaluation system, and dynamically optimizes data transmission and processing. First, it collects raw data from wildlife image acquisition terminals, covering core information such as terminal identification, network status, and image content. Then, through wildlife image recognition and feature extraction models, it sequentially calculates value assessment coefficients (integrating wildlife rarity and image feature saliency) and breakpoint resumption scheduling coefficients (combining network stability, transmission priority, and transmission progress). Based on this, it obtains an image compression strategy (comprehensively considering value, network status, and terminal resources). Then, based on the compression strategy and scheduling coefficients, it performs adaptive compression and breakpoint resumption transmission. Finally, it performs verification, retransmission, and image enhancement optimization on the identification server, forming a closed-loop feedback. This invention enables dynamic allocation of transmission resources according to the value of image content in weak network environments, ensuring high-definition transmission and reliable delivery of rare species images, while reducing the occupation of network and server resources by invalid data, and improving the efficiency and reliability of the overall monitoring system. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of a method flow according to an embodiment of this application.

[0015] Figure 2 This is a schematic diagram of the system structure according to an embodiment of this application.

[0016] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0018] like Figure 1 As shown, this application provides an adaptive compression and breakpoint resume scheduling method for wildlife images, including: S1. Obtain image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal, wherein the image acquisition information includes terminal identification information and terminal network transmission status information; S2. Obtain basic information about wild animals based on the content information of the wild animal images, and obtain a value assessment coefficient based on the basic information about wild animals; S3. Obtain network bandwidth data and image transmission priority based on the terminal network transmission status information, and obtain the breakpoint resume scheduling coefficient based on the network bandwidth data and the image transmission priority. S4. Obtain the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient. S5. Based on the image compression strategy and the breakpoint resume scheduling coefficient, adaptive compression and breakpoint resume scheduling are performed on the wildlife images acquired by the wildlife image acquisition terminal to obtain wildlife images after adaptive compression and breakpoint resume scheduling. S6. Obtain an identification server based on the terminal identification information, receive the wildlife image through the identification server and verify it to obtain a verification result, obtain a coordination optimization strategy based on the verification result and the image compression strategy, and perform coordination optimization on the wildlife image based on the coordination optimization strategy to obtain a wildlife image optimization result.

[0019] As described in steps S1-S6 above, when deploying wildlife image monitoring systems in remote areas, communication link bandwidth fluctuates drastically and traffic costs are high. Furthermore, the research value of images of rare species far exceeds that of common species. Existing technologies cannot dynamically allocate transmission resources based on the value of image content, leading to the potential loss of identification features due to excessive compression in weak networks or loss due to untimely retransmission. This invention first acquires image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal. The image acquisition information includes terminal identification information and terminal network transmission status information. By simultaneously acquiring three types of data—terminal identification, network status, and image content—it ensures that no key variables affecting compression and transmission are overlooked, avoiding strategy deviations due to incomplete information and providing a basis for subsequent decision-making. Among them, the parameter acquisition steps are as follows: image acquisition information is collected and reported in real time by wildlife image acquisition terminals such as solar-powered 4G infrared cameras and high-definition network PTZ cameras deployed in the monitoring area, and the terminal identification information is a unique identification code for each acquisition terminal, which can be obtained directly from the terminal's device configuration information. Next, basic information about wild animals is obtained based on the content information of the wild animal images, and a value assessment coefficient is obtained based on the basic information of wild animals. In this way, by taking the value assessment coefficient, it is possible to achieve value-based hierarchical scheduling (automatically matching the protection level and giving higher weight to high-value images such as Cabot's Tragopan, South China Tiger, and Chinese Merganser), thereby avoiding the excessive compression of rare species and allowing resources to be prioritized for high-value scientific research data. Due to the instability of networks in remote protected areas, it is necessary to assess network stability, transmission priority, and transmission progress to determine whether to enable breakpoint resumption and avoid packet loss and duplicate transmission. Therefore, network bandwidth data and image transmission priority are obtained based on the terminal network transmission status information, and then breakpoint resumption scheduling coefficient is obtained based on the network bandwidth data and image transmission priority. In this way, breakpoint resumption can be automatically enabled in weak networks, and reconnection after network failure will not start from the beginning. At the same time, high-priority terminals can occupy bandwidth first. Thus, by using the breakpoint resumption scheduling coefficient as the scheduling basis, retransmission traffic and latency can be significantly reduced, and the traffic costs caused by photo retransmission can also be significantly reduced. Secondly, since it is necessary to balance between size and image quality and protect the features of high-value images during the image uploading process, an image compression strategy is obtained based on the value assessment coefficient and the breakpoint resume scheduling coefficient. In this way, the image compression strategy can achieve low compression of high-value images while preserving features, while high compression of ordinary images saves data. At the same time, the compression rate is automatically increased when the terminal has low battery power to extend battery life. Next, based on the image compression strategy and the breakpoint resume scheduling coefficient, the wildlife images acquired by the wildlife image acquisition terminal are adaptively compressed and resumable, resulting in wildlife images after adaptive compression and resumable scheduling. This enables integrated compression and transmission, with the strategy being implemented immediately after generation and running fully automatically on the end side. Furthermore, it solves the problem that large files are almost impossible to transmit on weak networks, and resumable transmission avoids the need to retransmit the entire image due to interruption, thus preventing wasted bandwidth. Finally, an identification server is obtained based on the terminal identification information. The identification server receives and verifies the wildlife images to obtain the verification results. Based on the verification results and the image compression strategy, a coordinated optimization strategy is obtained, thus forming a closed-loop system: the server verifies the integrity and quality, and retransmits unqualified images; if qualified, the image quality is enhanced, and the strategy is stored for subsequent optimization. Then, the wildlife images are coordinated and optimized according to the coordinated optimization strategy to obtain the optimized wildlife images. In this way, the quality control of the transmitted images is achieved through professional verification on the server side, retransmission requests are initiated for unqualified images to ensure data integrity, and targeted enhancement and optimization are performed on qualified images to improve image quality. At the same time, the strategies and parameters in the optimization process are stored to provide a reference for subsequent compression and transmission strategy optimization, forming a closed-loop system of "acquisition-compression-transmission-verification-optimization-iteration". The specific implementation steps are as follows: First, based on the terminal identification information, obtain the identification server corresponding to the terminal from the system's server scheduling database. Each data collection terminal corresponds to a dedicated identification server, thereby achieving the directionality of data transmission and the balancing of server load. After receiving the wildlife images after adaptive compression and breakpoint resume scheduling, the identification server calls the corresponding decompression algorithm to decompress the image data packets according to the compression parameters in the image compression strategy. At the same time, it performs integrity verification on the decompressed image data through a data verification algorithm. The verification content includes whether data blocks are missing, whether data has been tampered with, and whether the format is correct, etc., to obtain the decompressed wildlife images and integrity verification results. Next, combining the decompressed wildlife images and the integrity verification results, image quality verification was carried out. The verification included whether the images were complete, whether there were decompression errors, and whether the image resolution met the preset standards. These preset standards were formulated based on the image usage requirements of wildlife ecological research. The overall image quality verification results were obtained. If any item in the image quality verification result fails, the verification result is determined to require optimization. The identification server will send a retransmission request to the corresponding wildlife image acquisition terminal, requesting the retransmission of the original wildlife image or the compressed wildlife image data packet according to the actual situation, to ensure that the finally acquired image data is complete and usable. If all image quality verification results pass, it means that the transmission and compression of the image meet the requirements. At this time, the decompressed wildlife image and basic wildlife information are input into the preset image enhancement model. This model is built based on deep learning and is trained for the features and image attributes of different wildlife. It can output the corresponding image optimization parameter set, which includes contrast adjustment parameters, sharpness adjustment parameters and color correction parameters, which are used to optimize the visual effect and feature recognition of the image, respectively. Subsequently, the decompressed wildlife images are enhanced and optimized based on the image optimization parameter set. The contrast, sharpness, and color of the images are adjusted to improve the image quality and the recognizability of wildlife features, resulting in optimized wildlife images. At the same time, the image compression strategy, integrity verification results, and image optimization parameter set are stored in the system's strategy database as a coordinated optimization strategy. The data in this database will serve as a reference for subsequent wildlife image compression, transmission, and optimization, enabling iterative optimization of the strategy and improving processing efficiency.

[0020] In one embodiment, step S2, which involves obtaining basic information about wild animals based on the content information of the wild animal image and obtaining a value assessment coefficient based on the basic information about wild animals, includes: S21. Based on the content information of the wild animal image, obtain the wild animal recognition result in the image through a preset wild animal image recognition model, and obtain the corresponding basic information of the wild animal from a preset wild animal information database based on the wild animal recognition result, wherein the basic information of the wild animal includes the name of the wild animal, the rarity level and the protection level. S22. Based on the rarity level and the protection level, obtain the wildlife rarity score through a preset rarity score mapping table; S23. Based on the content information of the wild animal image, obtain the feature clarity score and behavior salience score of the wild animal in the image through a preset wild animal feature extraction model, and obtain the wild animal feature salience index based on the feature clarity score and behavior salience score. S24. Obtain initial value assessment factors based on the wildlife rarity score and the wildlife characteristic significance index; S25. Obtain a preset historical image value benchmark threshold, and obtain a value assessment coefficient by ratio based on the initial value assessment factor and the historical image value benchmark threshold.

[0021] As described in steps S21-S25 above, the present invention obtains the wildlife identification results in the image through a preset wildlife image recognition model based on the wildlife image content information, and obtains the corresponding basic wildlife information from a preset wildlife information database based on the wildlife identification results. The basic wildlife information includes the wildlife name, rarity level, and protection level. This can automatically complete species identification without manual annotation and can establish a correlation between the image and the species protection level, providing an objective standard for subsequent rarity scoring. For example, input a bird image → identify it as a Chinese Merganser → database query: Name: Chinese Merganser, Rarity level: Extremely high, Protection level: National Class I protected animal; Secondly, based on the rarity level and the protection level, the rarity score of wild animals is obtained through a preset rarity score mapping table. In this way, the qualitative rarity level and protection level in the basic information of wild animals are transformed into a standardized quantitative score in the range of 0-1, realizing the objective quantification of the rarity value of wild animals themselves, making the rarity attributes of different wild animals comparable, and providing a unified quantitative indicator for subsequent value integration. In this step, the preset rarity rating mapping table is a standardized mapping rule formulated in combination with wildlife protection regulations and wildlife ecological research rarity evaluation standards. The table combines the rarity level of wild animals (such as critically endangered, endangered, vulnerable, near endangered, and least endangered) and protection level (such as national first-class, national second-class, local protection, and no protection) and matches a corresponding rarity rating in the 0-1 range for each combination. The scoring system is based on the degree of endangerment and the importance of conservation of wild animals, with a tiered system. For example, critically endangered wild animals under first-class national protection have a score of 0.9-1.0, endangered wild animals under second-class national protection have a score of 0.7-0.8, and wild animals without protection have a score of 0.1-0.2. By substituting the rarity level and protection level of wild animals into this mapping table, the corresponding rarity score of wild animals can be directly obtained. This score realizes the standardized quantification of the rarity value of wild animals themselves, avoiding the problem that qualitative descriptions cannot be directly used for calculation.

[0022] For example, the Chinese Merganser is a critically endangered wild animal under first-class national protection. After being substituted into the rarity rating mapping table, it can be matched with a wild animal rarity rating of 0.95, while the sparrow is a wild animal of least concern without protection, and is matched with a wild animal rarity rating of 0.15. The rarity value of the two forms a clear comparison through quantitative scoring, providing a clear quantitative basis for subsequent value assessment. Next, based on the wildlife image content information, the feature clarity score and behavioral salience score of the wildlife in the image are obtained through a preset wildlife feature extraction model, which can quantify the wildlife image content information. This step involves quantifying the quality and behavioral value of wildlife images. Its physical significance lies in quantifying the recognizability of wildlife features and the scientific research value of behavior in wildlife images from a visual perspective. This compensates for the one-sidedness of evaluating value solely based on the rarity of wildlife, making the evaluation of image value more aligned with actual scientific research needs. Even images of rare wildlife will have significantly reduced scientific research value if their features are blurry or they lack typical behavior. The specific steps are as follows: First, the preset wildlife feature extraction model is also built on a deep convolutional neural network. It is optimized based on the feature extraction of the wildlife image recognition model. The preset wildlife feature extraction model is built on a deep convolutional neural network. The network structure includes an input layer, four 2D convolutional layers, two pooling layers, two fully connected layers, and a dual-branch output layer.

[0023] The convolutional layer has a kernel size of 3×3 and a stride of 1. Padding is used to maintain the size of the feature map, which is used to extract features such as the outline, feather texture and limb posture of wild animals. The pooling layer uses max pooling with a kernel size of 2×2 and a stride of 2 to reduce dimensionality and preserve key features; the fully connected layer maps the extracted features to feature clarity scores and behavioral saliency scores. The output layer uses the Sigmoid activation function to output feature clarity score and behavior saliency score in the range of 0-1. The feature clarity score is used to evaluate the clarity of core features such as outline, feathers, and facial features of wild animals in the image. The score range is 0-1. The higher the score, the clearer the features and the stronger the recognizability. The behavioral salience score is used to assess whether the behaviors of wild animals in an image (such as foraging, reproduction, fighting, and migration) have scientific research value. The model is trained on labeled samples of wild animal behavior and can identify typical behaviors with scientific research value. The score ranges from 0 to 1, with higher scores indicating stronger scientific research value. After inputting the content information of a wild animal image into the model, the model simultaneously outputs a feature clarity score and a behavioral salience score. Then, based on the feature clarity score and behavior salience score, the wildlife feature salience index is obtained. In this way, the wildlife feature salience index can distinguish between high-value images and low-value images, allowing clear images with behavior to obtain higher value weights, and filtering out blurry and featureless invalid data. Specific steps: Since the wildlife rarity score reflects the scientific research value of the wildlife itself and is the core foundation of image value, the weight is set to 0.55. The Wildlife Feature Salience Index reflects the quality and behavioral value of the image itself and is an important supplement to the image value. The weight is set to 0.45. This weight allocation ensures the core value of rare wild animals while taking into account the quality and behavioral value of the image itself, avoiding the unreasonable situation that "the value of blurry images of rare wild animals is higher than that of high-definition typical behavioral images of common wild animals". The calculation formula is: Initial value assessment factor = Wildlife rarity score × 0.55 + Wildlife characteristic significance index × 0.45. This step achieves a comprehensive quantification of the overall value of wildlife images through the weighted fusion of multi-dimensional indicators.

[0024] Then, an initial value assessment factor is obtained based on the wildlife rarity score and the wildlife feature salience index. The initial value assessment factor formed in this way can reflect the comprehensive value index of species value (rarity) and image quality (feature salience). In terms of specific implementation steps, since the wildlife rarity score reflects the scientific research value of the wildlife itself and is the core foundation of image value, the weight is set to 0.55. The Wildlife Feature Salience Index reflects the quality and behavioral value of the image itself and is an important supplement to the image value. The weight is set to 0.45. This weight allocation ensures the core value of rare wild animals while taking into account the quality and behavioral value of the image itself, avoiding the unreasonable situation that "the value of blurry images of rare wild animals is higher than that of high-definition typical behavioral images of common wild animals". The calculation formula is: Initial value assessment factor = Wild animal rarity score × 0.55 + Wild animal characteristic significance index × 0.45; Finally, a preset historical image value benchmark threshold is obtained, and the value assessment coefficient is obtained by the ratio of the initial value assessment factor and the historical image value benchmark threshold. In this way, the ratio of the comprehensively quantified initial value assessment factor to the historical benchmark threshold is calculated to obtain the normalized value assessment coefficient, so that the value assessment coefficients of wild animal images of different periods and types have a unified reference standard and comparability. At the same time, this coefficient can be directly used to formulate subsequent compression and transmission strategies, realizing the connection between quantitative indicators and practical applications. The specific steps involve setting a benchmark threshold for the value of historical images. This threshold is obtained by statistically analyzing the initial value assessment factors of a large number of images from past wildlife monitoring. Specifically, the average value of the initial value assessment factors of historical images is used as the benchmark threshold. This threshold is dynamically updated by the system based on historical monitoring data, which can fit the actual monitoring scenarios and scientific research needs.

[0025] Next, the value assessment coefficient is obtained by calculating the ratio between the initial value assessment factor and the historical image value benchmark threshold. The calculation formula is: Value assessment coefficient = Initial value assessment factor ÷ Historical image value benchmark threshold. This coefficient normalizes the value of an image. If the coefficient is greater than 1, it means that the overall value of the image is higher than the historical average and it is a high-value image. If the coefficient is equal to 1, it means that its value is comparable to the historical average level; If the coefficient is less than 1, it indicates that its value is below the historical average, belonging to a low-value image. This step, through normalization, transforms the value assessment coefficient into a quantitative indicator that can be directly used for strategy formulation.

[0026] In one embodiment, step S3, which involves obtaining network bandwidth data and image transmission priority based on the terminal network transmission status information, and obtaining the breakpoint resumption scheduling coefficient based on the network bandwidth data and image transmission priority, includes: S31. Based on the terminal network transmission status information, obtain multiple real-time uplink bandwidth values ​​of the wildlife image acquisition terminal within a preset time window, and obtain a network transmission stability index based on the multiple real-time uplink bandwidth values. S32. Obtain the device type and task priority of the wildlife image acquisition terminal according to the terminal identification information, and obtain the image transmission priority based on the device type and the task priority through a preset priority decision tree; S33. Obtain the amount of data already transmitted and the total amount of data for the wildlife image to be transmitted, and obtain the image transmission progress based on the amount of data already transmitted and the total amount of data. S34. Obtain the initial scheduling factor based on the network transmission stability index, the image transmission priority, and the image transmission progress; S35. Obtain the system's preset scheduling baseline coefficient, and obtain the breakpoint resume scheduling coefficient by ratio of the initial scheduling factor and the scheduling baseline coefficient.

[0027] As described in steps S31-S35 above, since wildlife monitoring terminals are mostly located in remote, weak network environments, it is necessary to first quantify the network quality and fluctuation level in order to determine whether interrupted transmission is required. Therefore, this invention obtains multiple real-time uplink bandwidth values ​​of the wildlife image acquisition terminal within a preset time window based on the terminal network transmission status information, and obtains a network transmission stability index based on the multiple real-time uplink bandwidth values. The specific steps are as follows: Obtain the start and end times according to the preset time window; The average real-time uplink bandwidth value is calculated based on multiple real-time uplink bandwidth values, the start time, and the end time, wherein the calculation formula is: ; in, This represents the average real-time uplink bandwidth value. Indicates the start time. Indicates the end time. This represents the real-time uplink bandwidth value, where n represents the number of real-time uplink bandwidth values, and n = 1, 2, 3...n; The standard deviation of the signal-to-interference-plus-noise ratio (SIR) is calculated based on multiple real-time uplink bandwidth values ​​and the mean SIR, wherein the calculation formula is: ; in, Indicates the standard deviation of bandwidth. This represents the i-th real-time uplink bandwidth value. This indicates the number of real-time uplink bandwidth values, where i represents the index of the real-time uplink bandwidth value, i = 1, 2, 3...n. This represents the average real-time uplink bandwidth value; The bandwidth fluctuation coefficient is obtained based on the average real-time uplink bandwidth value and the bandwidth standard deviation, wherein the calculation formula is: ; in, Indicates the bandwidth fluctuation coefficient. This represents the average real-time uplink bandwidth value. Indicates the standard deviation of bandwidth; Using the bandwidth fluctuation coefficient as a network transmission stability index, the core performance parameters of network transmission are extracted, and the dynamically changing network state is transformed into a standardized quantitative index, thereby achieving an objective assessment of network transmission capacity and stability, and providing core network-level basis for subsequent transmission scheduling quantification. Next, the device type and task priority of the wildlife image acquisition terminal are obtained according to the terminal identification information. Based on the device type and task priority, the image transmission priority is obtained through a preset priority decision tree. In this way, by associating the hardware attributes of the device with the monitoring task requirements through the terminal identification, the device type and task priority are transformed into standardized image transmission priorities. This enables an objective determination of the importance of the transmission tasks of images acquired by different terminals, providing a core basis at the task level for subsequent transmission scheduling quantification. Specific implementation process: The terminal identification information is a unique identification code for each data acquisition terminal. This code is pre-bound to information such as the terminal's device type and task priority and stored in the terminal device management database. Next, based on the terminal identification information, the corresponding device type and task priority can be directly retrieved from the database. The device types are mainly divided into solar-powered 4G infrared cameras, high-definition network PTZ cameras, and portable acquisition terminals. Different device types have different acquisition capabilities, power consumption, and network compatibility. High-definition network PTZ cameras acquire images with higher resolution and relatively greater scientific research value, while solar-powered 4G infrared cameras are conventional monitoring devices. Secondly, the task priorities are set according to the importance of the monitoring area and the target type of the wild animals to be monitored, and are divided into three levels: Level 1, Level 2, and Level 3. Level 1 priority is key monitoring areas such as exclusive monitoring points for rare wild animals and core migratory bird passages; Level 2 priority is routine wild animal monitoring areas; and Level 3 priority is auxiliary monitoring areas. The preset priority decision tree is a hierarchical decision model built based on the equipment performance and task requirements of wildlife monitoring. The decision tree takes equipment type and task priority as input nodes. The first layer uses task priority as the basis for judgment. The first-level task priority directly corresponds to high transmission priority, the third-level task priority directly corresponds to low transmission priority, and the second-level task priority enters the second layer based on equipment type. Images acquired by high-definition network PTZ cameras correspond to medium-high transmission priority, while images acquired by solar-powered 4G infrared cameras correspond to medium-low transmission priority. The output nodes of the decision tree are the image transmission priority quantization values ​​in the range of 0-1, with high transmission priority corresponding to 0.8-1.0, medium-high transmission priority corresponding to 0.6-0.8, medium-low transmission priority corresponding to 0.4-0.6, and low transmission priority corresponding to 0-0.4. In this way, this step achieves standardized quantification of transmission priority through matching device and task attributes and decision tree determination.

[0028] Next, the amount of data already transmitted and the total amount of data for the wildlife images to be transmitted are obtained, and the image transmission progress is obtained based on the amount of data already transmitted and the total amount of data. In this way, by calculating the amount of data transmitted, the image transmission completion status is transformed into a standardized quantitative indicator, enabling accurate determination of the image transmission stage. This makes up for the one-sidedness of scheduling quantification based solely on network status and task priority, because images with the same network status and task priority have different scheduling requirements depending on their transmission progress. For example, images that are close to completion of transmission do not require complex breakpoint resume scheduling, while images that have just started transmission require more comprehensive scheduling guarantees. Specific steps: The amount of data already transmitted and the total amount of data for the wildlife images to be transmitted are counted in real time by the transmission module of the acquisition terminal; The total data volume is the total number of bytes in the wildlife image data packet after processing with the image compression strategy; The amount of data transmitted is the number of bytes that the terminal has successfully sent and that the server has acknowledged as received. Image transmission progress is calculated by the ratio of the amount of data transmitted to the total amount of data. The formula is: Image transmission progress = Amount of data transmitted ÷ Total amount of data; The value of this indicator ranges from 0 to 1. A value of 0 indicates that the image has not yet started to be transmitted. The closer the value is to 1, the higher the image transmission completion rate. A value of 1 indicates that the image transmission is complete. This step achieves accurate determination of image transmission progress through a simple and direct quantification method, providing real-time transmission status basis for subsequent multi-index fusion scheduling.

[0029] Because scheduling only considers network conditions and not task importance, rare data cannot be guaranteed. Secondly, scheduling only considers priority and ignores network status, forcing transmission under weak network conditions leads to frequent failures. At the same time, it does not consider transmission progress, and a large portion of the transmitted images are abandoned, resulting in huge waste. Under these combined factors, it is necessary to obtain an initial scheduling factor based on the network transmission stability index, the image transmission priority, and the image transmission progress. This way, three independent standardized quantitative indicators are weighted and fused to obtain an initial scheduling factor that can comprehensively reflect the overall transmission scheduling needs of wildlife images. This realizes the transformation from single-dimensional evaluation to multi-dimensional comprehensive evaluation, making the quantification of transmission scheduling more comprehensive, objective, and in line with the actual transmission scenario requirements. Among the specific steps, the network transmission stability index is the basis of transmission scheduling and determines whether breakpoint resumption needs to be enabled, so its weight is set to 0.4. Image transmission priority is the core of transmission scheduling, determining the priority of transmission resource allocation; therefore, its weight is set to 0.4. Image transmission progress is a supplement to transmission scheduling and determines the fineness of breakpoint resume transmission. Therefore, the weight is set to 0.2. This weight allocation ensures that the core influence of network status and task priority is taken into account, while also taking into account the supplementary role of real-time transmission progress, and avoids unreasonable scheduling judgments that ignore the transmission stage. The calculation formula is: Initial scheduling factor = (1 - network transmission stability index) × 0.4 + image transmission priority × 0.4 + image transmission progress × 0.2; The network transmission stability index has a negative correlation with the scheduling factor; that is, the lower the network transmission stability index (the worse the network condition), the higher the scheduling demand. Therefore, in actual calculations, this index is processed inversely, using "1 − network transmission stability index". The initial scheduling factor ranges from 0 to 1. The closer the value is to 1, the higher the image transmission scheduling requirement, and the more necessary it is to enable breakpoint resume and prioritize the allocation of transmission resources. The closer the value is to 0, the lower the image transmission scheduling requirement, and the more likely the conventional single transmission mode can be used.

[0030] Finally, the system's preset scheduling benchmark coefficient is obtained, and the breakpoint resume scheduling coefficient is obtained by the ratio of the initial scheduling factor and the scheduling benchmark coefficient. In this way, the ratio of the comprehensively quantified initial scheduling factor and the system's preset scheduling benchmark coefficient is calculated to obtain the normalized breakpoint resume scheduling coefficient. This allows the transmission scheduling requirements of different periods, different terminals, and different images to have a unified reference standard and comparability. At the same time, this coefficient can be directly compared with the preset resume scheduling threshold, realizing the automatic determination of the breakpoint resume mode and completing the connection between the quantitative indicators and the actual transmission strategy. The specific steps are as follows: Resume interruption scheduling coefficient = Initial scheduling factor ÷ Scheduling baseline coefficient. This coefficient normalizes the image transmission scheduling requirements. If the coefficient is greater than or equal to the system's preset resume scheduling threshold, it indicates that the image transmission scheduling requirements meet the criteria for enabling resume interruption, and resume interruption mode needs to be activated. If the coefficient is less than the resume scheduling threshold, it means that the transmission scheduling requirement of the image is low, and there is no need to enable the breakpoint resume mode; the single transmission mode is sufficient. The resume scheduling threshold is set according to the system transmission performance and actual application requirements, and is generally 1.0.

[0031] This step uses normalization to make the breakpoint resume scheduling coefficient a quantitative indicator that can be directly used to determine the transmission strategy. For example, the system's preset scheduling baseline coefficient is 0.5, and the initial scheduling factor of a certain image is 0.74. The calculated breakpoint resume scheduling coefficient is 0.74 ÷ 0.5 = 1.48. This coefficient is greater than the resume scheduling threshold of 1.0, so it is determined that the breakpoint resume mode needs to be enabled. The initial scheduling factor for a certain image is 0.3. After calculation, the breakpoint resume scheduling coefficient is 0.3 ÷ 0.5 = 0.6. This coefficient is less than the resume scheduling threshold of 1.0, so it is determined that the single transmission mode is adopted.

[0032] In one embodiment, step S4, which involves obtaining the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient, includes: S41. Obtain the current remaining battery power and storage space of the wildlife image acquisition terminal, and obtain the terminal resource stress based on the current remaining battery power and storage space. S42. Based on the value assessment coefficient, the breakpoint resume scheduling coefficient, and the terminal resource tension, determine the initial compression algorithm and the initial compression rate through a preset compression strategy decision matrix. S43. Obtain a preset image quality loss threshold, and perform simulated compression on the wildlife image according to the initial compression algorithm and the initial compression rate to obtain the simulated compressed image quality loss value; S44. Determine whether the image quality loss value is greater than the image quality loss threshold; S45. If the image quality loss value is not greater than the image quality loss threshold, then the initial compression algorithm and the initial compression ratio are used as the image compression strategy. S46. If the image quality loss value is greater than the image quality loss threshold, the initial compression ratio is adjusted by gradient descent until the image quality loss value corresponding to the adjusted compression ratio is not greater than the image quality loss threshold, and an image compression strategy is obtained based on the adjusted compression ratio and the initial compression algorithm.

[0033] As described in steps S41-S46 above, the current remaining power and storage space of the wildlife image acquisition terminal are obtained, and the terminal resource stress is obtained based on the current remaining power and storage space. In this way, the two core hardware resource states of the terminal, power and storage space, are transformed into standardized quantitative indices, realizing an objective assessment of the terminal's resource carrying capacity. This provides a hardware-level constraint basis for the formulation of subsequent compression strategies, and avoids the formulation of compression strategies exceeding the actual resource execution capacity of the terminal. Specifically, the remaining battery power of the wildlife image acquisition terminal is collected in real time by the terminal's built-in battery monitoring module and presented as a percentage of the remaining battery power. The remaining storage space is collected by the terminal's storage management module and presented as a percentage of the remaining storage space relative to the total storage space. Both types of data are reported to the system in real time by the terminal and are the raw data for assessing the terminal's resource status. After obtaining the two raw data points, they are standardized, and the remaining power percentage and storage space percentage are directly mapped to the 0-1 range. The closer the value is to 1, the more abundant the resources are, and the closer it is to 0, the more scarce the resources are. Since most wildlife monitoring terminals are powered by solar energy and outdoor charging conditions are limited, the power supply has a more critical impact on the continuous operation of the terminal. Therefore, the weight of the current remaining power supply is set to 0.6, and the weight of the storage space remaining capacity is set to 0.4. Simultaneously, the standardized resource data is processed in reverse, using 1 minus the standardized value in the calculation, so that the calculation result is positively correlated with the resource scarcity level. The calculation formula is: Terminal resource scarcity level = (1 - current remaining power standardized value) × 0.6 + (1 - storage space remaining standardized value) × 0.4. The value of this index ranges from 0 to 1. The closer the value is to 1, the more scarce the terminal resources are, and the lower the resource requirements for the compression strategy. The closer the value is to 0, the more abundant the terminal resources are, and the more suitable it is for higher quality compression requirements. Next, based on the value assessment coefficient, the breakpoint resume scheduling coefficient, and the terminal resource stress, the initial compression algorithm and initial compression rate are determined through a preset compression strategy decision matrix. This accurately matches the three core influencing factors—image value, transmission scheduling requirements, and terminal resource constraints—with the compression algorithm and compression rate, quickly obtaining an initial compression strategy that meets multi-dimensional requirements. This achieves preliminary adaptation of the compression strategy to the actual situation of the image, transmission, and terminal, providing a basic solution for subsequent quality verification. The specific confirmation process is achieved through a pre-set compression strategy decision matrix, which is a multi-dimensional matching model constructed by combining the image research needs of wildlife monitoring, transmission network characteristics, and terminal hardware capabilities. The row dimension of the matrix is ​​composed of a combination of value assessment coefficient, breakpoint resume scheduling coefficient, and terminal resource tension intervals. Each indicator is divided into three intervals: low (0-0.3), medium (0.3-0.7), and high (0.7-1.0), forming 27 indicator combination scenarios. Secondly, the column dimensions of the matrix are compression algorithms and compression ratios. The compression algorithms cover four categories: lossless compression (such as PNG), lightly lossy compression (such as JPEG with low compression ratio), moderately lossy compression (such as JPEG with medium compression ratio), and highly lossy compression (such as WebP with high compression ratio), to adapt to different image quality requirements. Meanwhile, the compression ratio is divided into six levels: 5%, 10%, 20%, 30%, 50%, and 70%, with 5% being low compression ratio (high quality) and 70% being high compression ratio (high compression ratio).

[0034] Each combination of indicators in the decision matrix corresponds to a unique initial compression algorithm and initial compression ratio. The matching rules follow the core principle of "high-value images are adapted to low compression ratios and lossless / lightly lossy compression, high breakpoint resume scheduling coefficients are adapted to high compression ratios and efficient compression algorithms, and high terminal resource stress is adapted to high compression ratios and low computing power compression algorithms, while taking into account the comprehensive impact of the three factors. Next, a preset image quality loss threshold is obtained, and the wildlife image is simulated and compressed according to the initial compression algorithm and the initial compression rate to obtain the image quality loss value after simulated compression. In this way, the degree of loss of image quality caused by the initial compression strategy is calculated in advance through simulated compression, providing a quantitative basis for subsequent quality judgment and avoiding the problem that the image quality loss exceeds the acceptable range for scientific research due to the direct adoption of the initial compression strategy. The compression process involves setting a standardized threshold for image quality loss based on the image identification requirements of wildlife ecology research, with the structural similarity index (SSIM) of the image used as the evaluation metric. The value of SSIM ranges from 0 to 1. The closer the value is to 1, the higher the structural similarity between the compressed image and the original image, and the smaller the quality loss. Based on the minimum requirements for image features and behavior identification in wildlife research, the image quality loss threshold is set to SSIM≥0.8, meaning that the SSIM corresponding to the quality loss value of the compressed image must not be lower than 0.8. This threshold is uniformly set and stored by the system according to research needs.

[0035] After obtaining the initial compression algorithm and initial compression ratio, the system calls the corresponding compression algorithm module to simulate compression of the original wildlife image according to the initial compression ratio. The simulated compression is executed on the system side and does not occupy the hardware resources of the acquisition terminal. After compression, the SSIM value of the simulated compressed image and the original image is calculated by the image quality assessment algorithm. The 1-SSIM value is used as the image quality loss value. The value ranges from 0 to 0.2. The closer the value is to 0, the smaller the quality loss. The closer it is to 0.2, the larger the quality loss. This realizes the quantitative measurement of the quality loss of the initial compression strategy. For example, when a lossless compression algorithm with an initial compression rate of 5% was used to simulate compression of a high-resolution image of the Cabot's Tragopan, the calculated SSIM value was 0.99 and the image quality loss value was 0.01, which is far lower than the quality loss threshold of 0.2, indicating that the quality loss of the initial compression strategy meets the requirements. If an image of a rare wild animal is initially matched with moderate lossy compression and an initial compression rate of 30% due to limited terminal resources, the simulated compression will result in an SSIM value of 0.75 and an image quality loss value of 0.25, which exceeds the threshold range of 0.2. This indicates that the quality loss of the initial compression strategy does not meet the needs of scientific research and parameter adjustments are required.

[0036] At the same time, it is determined whether the image quality loss value is greater than the image quality loss threshold. By comparing the actual quality loss value obtained from the simulated compression with the system's preset quality loss threshold, a clear quality judgment result is formed, providing a decision-making basis for the determination or adjustment of subsequent compression strategies. This is a key node for achieving closed-loop control of compression quality. During the comparison process, the preset image quality loss threshold (i.e., 1-SSIM=0.2) is used as the judgment benchmark. If the image quality loss value obtained by simulated compression is greater than 0.2, it indicates that the quality loss of the initial compression strategy exceeds the acceptable range for wildlife research, and the initial compression rate needs to be adjusted to reduce the quality loss. If the image quality loss value is no greater than 0.2, it indicates that the quality loss of the initial compression strategy meets the research requirements, and no parameter adjustment is needed. The initial compression algorithm and initial compression rate can be directly used as the final image compression strategy. This simple and direct judgment process serves as a bridge between initial compression strategy matching and adaptive compression ratio adjustment, preventing unqualified compression strategies from being used directly and providing a basis for qualified compression strategies to take effect directly. Secondly, if the image quality loss value is not greater than the image quality loss threshold, the initial compression algorithm and the initial compression ratio are used as the image compression strategy. This confirms the initial compression strategy that meets the quality requirements after simulated compression verification and uses it as the final image compression strategy. This provides a clear execution plan for subsequent actual image compression operations and realizes the implementation of the strategy after multi-dimensional index matching and quality verification. Meanwhile, if the image quality loss value is greater than the image quality loss threshold, the initial compression ratio is adjusted by gradient descent until the image quality loss value corresponding to the adjusted compression ratio is not greater than the image quality loss threshold. Then, based on the adjusted compression ratio and the initial compression algorithm, an image compression strategy is obtained. In this way, the initial compression ratio is adaptively and iteratively adjusted using the gradient descent method. While retaining the initially adapted compression algorithm, the compression ratio is reduced to reduce image quality loss until the compression strategy meets the quality requirements. This allows for the optimal adjustment of the compression ratio under quality constraints, ensuring that the image quality meets research needs while also taking into account terminal resources and transmission requirements as much as possible. In this step, the gradient descent method adjusts the initial compression ratio while keeping the compression algorithm unchanged. This is because the compression algorithm is better suited to the core needs of image value and terminal computing power, while adjusting the compression ratio allows for flexible control of the degree of image quality loss without changing the algorithm.

[0037] The parameter settings of the gradient descent method are formulated in accordance with the actual needs of wildlife image compression. The image quality loss value is less than or equal to 0.2 as the objective function, the compression rate is the independent variable, and the learning rate is set to 5%, that is, the compression rate gradient is 5% each time, and the adjustment direction is to reduce the compression rate from high to low. Secondly, because the higher the compression ratio, the greater the loss of image quality, reducing the compression ratio can effectively reduce the loss of quality.

[0038] The specific adjustment process is as follows: starting from the initial compression rate, the compression rate is reduced by 5% each time, and the wildlife images are simulated and compressed according to the adjusted compression rate and the initial compression algorithm, and the corresponding image quality loss value is calculated. If the adjusted quality loss value is still greater than 0.2, continue to reduce the compression ratio by 5% and repeat the simulation compression and quality calculation steps until the image quality loss value corresponding to the adjusted compression ratio is no greater than 0.2, at which point the adjustment is stopped. The adjusted compression ratio and the initial compression algorithm are then used as the final image compression strategy.

[0039] In one embodiment, step S5, which involves adaptively compressing and scheduling breakpoint resume transmission of wildlife images acquired by the wildlife image acquisition terminal according to the image compression strategy and the breakpoint resume transmission scheduling coefficient, to obtain wildlife images after adaptive compression and breakpoint resume transmission scheduling, includes: S51. According to the image compression strategy, the original wildlife images acquired by the wildlife image acquisition terminal are compressed to obtain a compressed wildlife image data packet, and the compression parameters are recorded. S52. Based on the breakpoint resume scheduling coefficient, determine whether the compressed wildlife image data packet needs to enable the breakpoint resume mode. S53. If the breakpoint resume scheduling coefficient is less than the preset resume scheduling threshold, then the single transmission mode is adopted to send the compressed wildlife image data packet to the identification server all at once. S54. If the breakpoint resume scheduling coefficient is greater than or equal to the preset resume scheduling threshold, the breakpoint resume mode is enabled, the compressed wildlife image data packet is divided into multiple data blocks, and sent sequentially according to the preset transmission order, while recording the transmission status of each data block and the information of the data blocks that have been confirmed to be received. S55. When the connection is restored after the transmission is interrupted, the transmission continues from the first unacknowledged data block according to the confirmed data block information until all data blocks are sent, and the wildlife image after adaptive compression and breakpoint resume scheduling is obtained.

[0040] As described in steps S51-S55 above, the present invention first compresses the original wildlife images collected by the wildlife image acquisition terminal according to the image compression strategy to obtain compressed wildlife image data packets and records the compression parameters. In this way, the personalized image compression strategy formulated in the early stage is applied to the original wildlife images to complete the actual compression process and form a standardized data packet. At the same time, the key parameters in the compression process are fully recorded to provide core basis for decompression, integrity verification and quality assessment on the server side, and realize the effective connection between the compression link and the transmission and verification link. Secondly, based on the breakpoint resume scheduling coefficient, it is determined whether the compressed wildlife image data packet needs to enable the breakpoint resume mode. In this way, the quantitative result of the breakpoint resume scheduling coefficient is compared with the preset threshold to form a clear basis for determining the transmission mode. This allows the transmission mode of the image to be accurately adapted to its own scheduling needs, which is the core node for achieving differentiated transmission and avoids the problem that a single transmission mode cannot balance efficiency and reliability. If the breakpoint resume scheduling coefficient is less than the preset resume scheduling threshold, a single transmission mode is adopted, and the compressed wildlife image data packet is sent to the identification server all at once. This provides a reliable breakpoint resume mode for images with high transmission scheduling requirements. By standardizing the segmentation of data packets, the transmission pressure of a single data block is reduced. Combined with orderly transmission and refined status recording, precise control of the transmission process is achieved, providing core data basis for subsequent interruption recovery and ensuring the reliability of image transmission in weak network environments. This allows for efficient single-transmission mode matching for images with low transmission scheduling requirements. The complete compressed data packet is sent directly to the corresponding server, maximizing transmission efficiency and avoiding efficiency loss for low-demand images due to complex transmission processes. At the same time, the binding relationship between the terminal identifier and the server enables targeted transmission of data packets. If the breakpoint resume scheduling coefficient is greater than or equal to the preset resume scheduling threshold, then the breakpoint resume mode is enabled, and the compressed wildlife image data packet is divided into multiple data blocks. Then, the data is sent sequentially according to the preset transmission order, while recording the transmission status of each data block and the information of the data blocks that have been confirmed to be received. This enables a reliable breakpoint resume mode to be matched for images with high transmission scheduling requirements. Secondly, by standardizing and segmenting data packets, the transmission pressure of single data blocks is reduced. Combined with orderly sending and refined status recording, precise control of the transmission process is achieved, providing core data basis for subsequent interruption recovery and resumption of transmission, and ensuring the reliability of image transmission in weak network environments. Finally, when the connection is restored after a transmission interruption, the transmission continues from the first unacknowledged data block according to the confirmed received data block information until all data blocks are sent, resulting in an image of wildlife after adaptive compression and breakpoint resume scheduling. This can solve the problem of transmission stagnation caused by network interruption in weak network environments, and rely on the previously recorded confirmed received data block information to achieve accurate resumption of transmission from the breakpoint, avoiding the retransmission of the entire image and saving network bandwidth and transmission time. Ultimately, this ensures the complete transmission of compressed image data packets to the server, completing the entire process of adaptive compression and breakpoint resume scheduling. This solves the problem of transmission stagnation caused by network interruptions in weak network environments. Relying on the previously recorded confirmed received data block information, it enables precise recovery and resumption of transmission from the breakpoint, avoiding the retransmission of the entire image, saving network bandwidth and transmission time, and ultimately ensuring the complete transmission of compressed image data packets to the server, completing the entire process of adaptive compression and breakpoint resume scheduling.

[0041] In one embodiment, step S6, which involves receiving and verifying the wildlife image through the identification server to obtain a verification result, acquiring a coordination optimization strategy based on the verification result and the image compression strategy, and performing coordination optimization on the wildlife image according to the coordination optimization strategy to obtain an optimized wildlife image, includes: S61. The identification server receives the wildlife image after adaptive compression and breakpoint resume scheduling, and decompresses and verifies the integrity of the received wildlife image data according to the compression parameters in the image compression strategy, to obtain the decompressed wildlife image and the integrity verification result. S62. Based on the decompressed wildlife image and the integrity verification result, obtain the image quality verification result, wherein the image quality verification result includes whether the image is complete, whether there is a decompression error, and whether the image resolution meets the preset standard. S63. If any of the image quality verification results fails, the verification result is determined to be that optimization is needed, and a retransmission request is sent to the wildlife image acquisition terminal to request the retransmission of the corresponding original wildlife image or compressed wildlife image data packet. S64. If all the image quality verification results pass, then based on the decompressed wildlife image and the basic information of the wildlife, an image optimization parameter set is obtained through a preset image enhancement model, wherein the image optimization parameter set includes contrast adjustment parameters, sharpness adjustment parameters and color correction parameters. S65. Enhance and optimize the decompressed wildlife image according to the image optimization parameter set to obtain the wildlife image optimization result. Store the image compression strategy, the integrity verification result, and the image optimization parameter set as a coordinated optimization strategy for use as a reference for optimizing subsequently received wildlife images.

[0042] As described in steps S61-S65 above, the identification server of the present invention receives the wildlife images after adaptive compression and breakpoint resume scheduling, and decompresses and verifies the integrity of the received wildlife image data according to the compression parameters in the image compression strategy, to obtain the decompressed wildlife images and integrity verification results. In this way, the image is received in a targeted manner through a dedicated identification server, and accurate decompression is achieved by relying on the compression parameters transmitted on the terminal side. At the same time, the integrity verification confirms whether the image data has been lost or tampered with during transmission, laying an effective data foundation for subsequent quality verification and optimization, and realizing seamless connection between the transmission link and the server processing link. Secondly, based on the decompressed wildlife images and the integrity verification results, image quality verification results are obtained. These results include whether the image is complete, whether there are decompression errors, and whether the image resolution meets preset standards. In this way, based on the integrity verification, a multi-dimensional quality judgment is further carried out from the image presentation level, setting a more stringent image quality threshold. This comprehensively confirms whether the images meet the basic requirements for wildlife research, providing a clear quality basis for subsequent retransmission or optimization decisions and preventing unqualified images from entering the subsequent processing stage. If any of the image quality verification results fails, the verification result is determined to require optimization, and a retransmission request is sent to the wildlife image acquisition terminal to request the retransmission of the corresponding original wildlife image or compressed wildlife image data packet. In this way, the image that has passed the quality verification initiates a precise retransmission request, and the retransmission on the terminal side completes the valid image data, ensuring that the server can obtain wildlife images that meet the quality requirements. This avoids data failure due to a single unqualified image transmission and improves the effectiveness of image data in the entire monitoring system. If all the image quality verification results pass, then based on the decompressed wildlife image and the basic information of the wildlife, an image optimization parameter set is obtained through a preset image enhancement model. The image optimization parameter set includes contrast adjustment parameters, sharpness adjustment parameters, and color correction parameters. In this way, by combining the actual image quality of the qualified image with the corresponding basic information of the wildlife, a targeted optimization parameter set is generated through a customized image enhancement model. This allows the image enhancement optimization to be tailored to the characteristics of different wildlife and the actual image quality issues of the image, avoiding the situation where a single enhancement optimization leads to the indistinct features of some wildlife or unresolved image quality issues. This provides accurate parameter basis for subsequent image enhancement optimization. The specific steps involve a pre-defined image enhancement model that is a customized model built on deep learning, specifically optimized for the features and shooting environment of wildlife images. The model's network structure includes an input layer, convolutional layer, pooling layer, fully connected layer, and output layer. The convolutional layer consists of 5 two-dimensional convolutional layers with a kernel size of 3×3 and a stride of 1. Padding is used to maintain the feature map size. The first 3 convolutional layers are used to extract basic image quality features, such as contrast, sharpness, and color deviation. The last 2 convolutional layers combine basic information about wild animals to extract specific features of different wild animals. For features such as feather texture, color characteristics, and body contour, the pooling layer uses a combination of max pooling and average pooling, with a pooling kernel size of 2×2 and a stride of 2. This preserves key image quality and wildlife features while reducing dimensionality. Three fully connected layers are set to map the extracted features to corresponding optimization parameter values. The output layer outputs contrast adjustment parameters, sharpness adjustment parameters, and color correction parameters, forming an image optimization parameter set. The model was trained on a large number of wildlife image samples labeled with wildlife information, image quality issues, and optimization parameters. It can accurately identify the image quality shortcomings of the input decompressed wildlife image and basic wildlife information, and generate suitable optimization parameters in combination with wildlife features.

[0043] Finally, the decompressed wildlife images are enhanced and optimized according to the image optimization parameter set to obtain the wildlife image optimization result. The image compression strategy, the integrity verification result, and the image optimization parameter set are stored as a coordinated optimization strategy for subsequent received wildlife images as a reference for optimization. In this way, the image quality of qualified wildlife images is enhanced according to the personalized image optimization parameter set, resulting in wildlife image optimization results that meet the high requirements of scientific research. At the same time, the core strategies and parameters of this processing are stored as a coordinated optimization strategy to realize the iterative reuse of processing experience. This allows subsequent images of similar wildlife, with the same image quality issues, and in the same transmission scenarios to directly refer to this strategy, improving processing efficiency and optimization effect, and forming a closed-loop iterative system for image processing.

[0044] like Figure 2 As shown, this application also provides an adaptive compression and breakpoint resume scheduling system for wildlife images, including: The first acquisition module is used to acquire image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal, wherein the image acquisition information includes terminal identification information and terminal network transmission status information; The second acquisition module is used to acquire basic information about wild animals based on the content information of the wild animal images, and to acquire a value assessment coefficient based on the basic information about wild animals. The third acquisition module is used to acquire network bandwidth data and image transmission priority based on the terminal network transmission status information, and to acquire breakpoint resume scheduling coefficient based on the network bandwidth data and the image transmission priority. The fourth acquisition module is used to acquire the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient. The fifth acquisition module is used to adaptively compress and schedule breakpoint resume transmission of the wildlife images acquired by the wildlife image acquisition terminal according to the image compression strategy and the breakpoint resume transmission scheduling coefficient, so as to obtain the wildlife images after adaptive compression and breakpoint resume transmission scheduling. The sixth acquisition module is used to acquire an identification server based on the terminal identification information, receive the wildlife image through the identification server and verify it to obtain a verification result, acquire a coordination optimization strategy based on the verification result and the image compression strategy, and perform coordination optimization on the wildlife image based on the coordination optimization strategy to obtain an optimized wildlife image result.

[0045] In one embodiment, the second acquisition module includes: The first acquisition unit is used to acquire the wild animal recognition result in the image based on the wild animal image content information through a preset wild animal image recognition model, and to acquire the corresponding wild animal basic information from a preset wild animal information database based on the wild animal recognition result, wherein the wild animal basic information includes the wild animal name, rarity level and protection level; The second acquisition unit is used to acquire the rarity score of wild animals according to the rarity level and the protection level through a preset rarity score mapping table; The third acquisition unit is used to acquire the feature clarity score and behavior salience score of the wild animals in the image based on the content information of the wild animal image, through a preset wild animal feature extraction model, and to acquire the wild animal feature salience index based on the feature clarity score and behavior salience score. The fourth acquisition unit is used to acquire initial value assessment factors based on the wildlife rarity score and the wildlife characteristic salience index; The fifth acquisition unit is used to acquire a preset historical image value benchmark threshold, and to acquire a value assessment coefficient by ratio based on the initial value assessment factor and the historical image value benchmark threshold.

[0046] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0047] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0048] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0049] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0050] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for adaptive compression and breakpoint resume scheduling of wildlife images, characterized in that, include: The system acquires image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal, wherein the image acquisition information includes terminal identification information and terminal network transmission status information. Basic information about wild animals is obtained based on the content information of the wild animal images, and a value assessment coefficient is obtained based on the basic information about wild animals. The network bandwidth data and image transmission priority are obtained based on the terminal network transmission status information, and the breakpoint resume scheduling coefficient is obtained based on the network bandwidth data and the image transmission priority. The image compression strategy is obtained based on the value assessment coefficient and the breakpoint resume scheduling coefficient. Based on the image compression strategy and the breakpoint resume scheduling coefficient, the wildlife images acquired by the wildlife image acquisition terminal are adaptively compressed and resumable from breakpoints to obtain wildlife images after adaptive compression and resumable from breakpoints. The identification server is obtained based on the terminal identification information. The wildlife image is received and verified through the identification server to obtain the verification result. A coordination optimization strategy is obtained based on the verification result and the image compression strategy. The wildlife image is then coordinated and optimized based on the coordination optimization strategy to obtain the wildlife image optimization result.

2. The method of claim 1, wherein, The steps of obtaining basic information about wild animals based on the content information of the wild animal images, and obtaining a value assessment coefficient based on the basic information about wild animals, include: Based on the content information of the wildlife images, the wildlife recognition results in the images are obtained through a preset wildlife image recognition model, and the corresponding basic wildlife information is obtained from a preset wildlife information database based on the wildlife recognition results. The basic wildlife information includes the wildlife name, rarity level, and protection level. Based on the rarity level and the protection level, the rarity score of wild animals is obtained through a preset rarity score mapping table; Based on the content information of the wildlife images, a preset wildlife feature extraction model is used to obtain the feature clarity score and behavioral salience score of the wildlife in the images, and the wildlife feature salience index is obtained based on the feature clarity score and behavioral salience score. Initial value assessment factors are obtained based on the wildlife rarity score and the wildlife characteristic significance index; Obtain a preset historical image value benchmark threshold, and obtain a value assessment coefficient by ratio between the initial value assessment factor and the historical image value benchmark threshold.

3. The method of claim 1, wherein, The step of obtaining network bandwidth data and image transmission priority based on the terminal network transmission status information, and obtaining the breakpoint resumption scheduling coefficient based on the network bandwidth data and image transmission priority, includes: Based on the terminal network transmission status information, multiple real-time uplink bandwidth values ​​of the wildlife image acquisition terminal within a preset time window are obtained, and a network transmission stability index is obtained based on the multiple real-time uplink bandwidth values. The device type and task priority of the wildlife image acquisition terminal are obtained based on the terminal identification information, and the image transmission priority is obtained through a preset priority decision tree based on the device type and the task priority. Obtain the amount of data already transmitted and the total amount of data for the wildlife images currently to be transmitted, and obtain the image transmission progress based on the amount of data already transmitted and the total amount of data. The initial scheduling factor is obtained based on the network transmission stability index, the image transmission priority, and the image transmission progress. Obtain the system's preset scheduling baseline coefficient, and obtain the breakpoint resume scheduling coefficient by ratio of the initial scheduling factor and the scheduling baseline coefficient.

4. The method of claim 1, wherein, The step of obtaining the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient includes: The system obtains the current remaining battery power and storage space of the wildlife image acquisition terminal, and determines the terminal resource stress based on the current remaining battery power and storage space. Based on the value assessment coefficient, the breakpoint resume scheduling coefficient, and the terminal resource stress, the initial compression algorithm and the initial compression rate are determined through a preset compression strategy decision matrix. A preset image quality loss threshold is obtained, and the wildlife image is simulated and compressed according to the initial compression algorithm and the initial compression ratio to obtain the simulated compressed image quality loss value. Determine whether the image quality loss value is greater than the image quality loss threshold; If the image quality loss value is not greater than the image quality loss threshold, then the initial compression algorithm and the initial compression ratio are used as the image compression strategy. If the image quality loss value is greater than the image quality loss threshold, the initial compression ratio is adjusted by gradient descent until the image quality loss value corresponding to the adjusted compression ratio is not greater than the image quality loss threshold, and an image compression strategy is obtained based on the adjusted compression ratio and the initial compression algorithm.

5. The method of claim 1, wherein, The step of adaptively compressing and resuming interrupted transmission of wildlife images acquired by the wildlife image acquisition terminal according to the image compression strategy and the interrupted transmission resumption scheduling coefficient, to obtain wildlife images after adaptive compression and interrupted transmission resumption scheduling, includes: According to the image compression strategy, the original wildlife images acquired by the wildlife image acquisition terminal are compressed to obtain compressed wildlife image data packets, and the compression parameters are recorded. Based on the breakpoint resume scheduling coefficient, determine whether the compressed wildlife image data packet needs to enable the breakpoint resume mode. If the breakpoint resume scheduling coefficient is less than the preset resume scheduling threshold, then the single transmission mode is adopted to send the compressed wildlife image data packet to the identification server all at once. If the breakpoint resume scheduling coefficient is greater than or equal to the preset resume scheduling threshold, the breakpoint resume mode is enabled, the compressed wildlife image data packet is divided into multiple data blocks, and sent sequentially according to the preset transmission order, while recording the transmission status of each data block and the information of the data blocks that have been confirmed to be received. When the connection is restored after a transmission interruption, the transmission continues from the first unacknowledged data block according to the confirmed data block information, until all data blocks are sent, resulting in a wildlife image after adaptive compression and breakpoint resume scheduling.

6. The method of claim 1, wherein, The steps of receiving the wildlife image through the identification server and verifying it to obtain a verification result, obtaining a coordination optimization strategy based on the verification result and the image compression strategy, and performing coordination optimization on the wildlife image according to the coordination optimization strategy to obtain an optimized wildlife image include: The identification server receives the wildlife images after adaptive compression and breakpoint resume scheduling, and decompresses and verifies the integrity of the received wildlife image data according to the compression parameters in the image compression strategy, to obtain the decompressed wildlife images and integrity verification results. Based on the decompressed wildlife image and the integrity verification result, an image quality verification result is obtained, wherein the image quality verification result includes whether the image is complete, whether there is a decompression error, and whether the image resolution meets a preset standard; If any of the image quality verification results fails, the verification result is determined to require optimization, and a retransmission request is sent to the wildlife image acquisition terminal to request the retransmission of the corresponding original wildlife image or compressed wildlife image data packet. If all the image quality verification results pass, then based on the decompressed wildlife image and the basic information of the wildlife, an image optimization parameter set is obtained through a preset image enhancement model, wherein the image optimization parameter set includes contrast adjustment parameters, sharpness adjustment parameters and color correction parameters; The decompressed wildlife image is enhanced and optimized according to the image optimization parameter set to obtain the wildlife image optimization result. The image compression strategy, the integrity verification result, and the image optimization parameter set are stored as a coordinated optimization strategy for use as a reference for optimizing subsequently received wildlife images.

7. A wild animal image adaptive compression and breakpoint continuation scheduling system, characterized in that, include: The first acquisition module is used to acquire image acquisition information and wildlife image content information reported by the wildlife image acquisition terminal, wherein the image acquisition information includes terminal identification information and terminal network transmission status information; The second acquisition module is used to acquire basic information about wild animals based on the content information of the wild animal images, and to acquire a value assessment coefficient based on the basic information about wild animals. The third acquisition module is used to acquire network bandwidth data and image transmission priority based on the terminal network transmission status information, and to acquire breakpoint resume scheduling coefficient based on the network bandwidth data and the image transmission priority. The fourth acquisition module is used to acquire the image compression strategy based on the value assessment coefficient and the breakpoint resume scheduling coefficient. The fifth acquisition module is used to adaptively compress and schedule breakpoint resume transmission of the wildlife images acquired by the wildlife image acquisition terminal according to the image compression strategy and the breakpoint resume transmission scheduling coefficient, so as to obtain the wildlife images after adaptive compression and breakpoint resume transmission scheduling. The sixth acquisition module is used to acquire an identification server based on the terminal identification information, receive the wildlife image through the identification server and verify it to obtain a verification result, acquire a coordination optimization strategy based on the verification result and the image compression strategy, and perform coordination optimization on the wildlife image based on the coordination optimization strategy to obtain an optimized wildlife image result.

8. The wild animal image adaptive compression and breakpoint transfer scheduling system according to claim 7, wherein, The second acquisition module includes: The first acquisition unit is used to acquire the wild animal recognition result in the image based on the wild animal image content information through a preset wild animal image recognition model, and to acquire the corresponding wild animal basic information from a preset wild animal information database based on the wild animal recognition result, wherein the wild animal basic information includes the wild animal name, rarity level and protection level; The second acquisition unit is used to acquire the rarity score of wild animals according to the rarity level and the protection level through a preset rarity score mapping table; The third acquisition unit is used to acquire the feature clarity score and behavior salience score of the wild animals in the image based on the content information of the wild animal image, through a preset wild animal feature extraction model, and to acquire the wild animal feature salience index based on the feature clarity score and behavior salience score. The fourth acquisition unit is used to acquire initial value assessment factors based on the wildlife rarity score and the wildlife characteristic salience index; The fifth acquisition unit is used to acquire a preset historical image value benchmark threshold, and to acquire a value assessment coefficient by ratio based on the initial value assessment factor and the historical image value benchmark threshold. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images