Hydro-fluctuation belt ecological environment monitoring method, system and equipment and storage medium

Abstract

The invention relates to the technical field of hydro-fluctuation belt ecological environment monitoring, and provides a hydro-fluctuation belt ecological environment monitoring method, system and device and a storage medium, and the method comprises the steps: obtaining monitoring information of a falling belt coverage canceling region, ecological vegetation and a character target; decoding the monitoring image of the hydro-fluctuation belt based on the monitoring information, and performing frame extraction according to a preset time interval to obtain a to-be-analyzed image; and calculating and analyzing the proportion of the bare land area of the hydro-fluctuation belt area in the whole hydro-fluctuation belt area, the proportion of the ecological vegetation area in the whole hydro-fluctuation belt area and the appearance duration of the figure target based on the to-be-analyzed image, and outputting early warning information. Therefore, by comparing the bare land area of the hydro-fluctuation belt, the ecological vegetation area and the appearance duration of the person target with the preset threshold value and recording the change of the area of the concerned area according to the time change, the purposes of ecological monitoring and early warning management are achieved, and the behavior of artificially entering the ecological area of the hydro-fluctuation belt is sensed and early warned in real time. Man-made interference on the ecological environment of the hydro-fluctuation belt is avoided.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of ecological environment monitoring in a drawdown zone, and in particular relates to a method, system, device and storage medium for monitoring the ecological environment in a drawdown zone. Background Art

[0002] The drawdown zone of the Three Gorges Reservoir is a large area of "water in winter and land in summer" formed by the periodic water level fluctuations in the main and tributary rivers of the Yangtze River. The drawdown zone plays a vital role in maintaining the balance of the reservoir bank ecosystem and protecting reservoirs, lakes and river water bodies. However, the periodic flooding stress has caused drastic changes in the vegetation community in the drawdown zone, which has had a serious impact on the normal operation of the Three Gorges Dam and the ecological security of the reservoir area.

[0003] Due to periodic inundation and drawdown, research on the ecological environment of the drawdown zone in the Three Gorges Reservoir area mainly focuses on the ecological reconstruction and restoration of vegetation communities and species. Monitoring changes in the area and ecological vegetation of the drawdown zone can provide more timely information and early warnings for ecological management and restoration of the drawdown zone. At the same time, monitoring of people in the drawdown zone can also prevent ecological damage caused by interference.

[0004] At present, there are relatively few means and methods to monitor the land distribution and vegetation distribution in the drawdown zone, and the measures to prevent interference with the ecology of the drawdown zone are also relatively limited. Summary of the Invention

[0005] To solve the above problems, the present disclosure provides a method, system, equipment and storage medium for monitoring the ecological environment of the drawdown zone. By monitoring the coverage area, ecological vegetation and human targets of the drawdown zone, the real-time perception capability of the changes in the ecological environment of the drawdown zone is achieved, and early warning and ecological management are carried out according to abnormal ecological changes. At the same time, real-time perception and early warning are carried out for human behavior of entering the ecological area of the drawdown zone to avoid human interference with the ecological environment of the drawdown zone.

[0006] A first aspect of the present disclosure provides a method for monitoring the ecological environment of a water-logging zone, comprising: obtaining monitoring information of the coverage area, ecological vegetation, and human targets of the water-logging zone; decoding the monitoring image of the water-logging zone based on the monitoring information, extracting frames at preset time intervals, and obtaining images to be analyzed; calculating and analyzing the proportion of the bare land area of the water-logging zone to the entire water-logging zone area, the proportion of the ecological vegetation area to the entire water-logging zone area, and the duration of the appearance of human targets based on the images to be analyzed, and outputting early warning information.

[0007] This setting makes it easier to monitor changes in the area of bare land in the drawdown zone, the survival status of vegetation, and changes in the area of vegetation. At the same time, it can also monitor the appearance of people in the drawdown zone. By analyzing the monitoring image of the bare land area in the drawdown zone, the area of ecological vegetation, the duration of the appearance of human targets and other information, and comparing them with the preset ratio and time, the changes in the area of the focus area are recorded according to the time change to achieve the purpose of ecological monitoring and early warning management, and real-time perception and early warning of human behavior entering the ecological area of the drawdown zone are carried out to avoid human interference with the ecological environment of the drawdown zone.

[0008] In some embodiments, the acquisition of monitoring information of the area covered by the water-drawing zone, ecological vegetation, and human targets includes: the monitoring information includes monitoring images of the area covered by the water-drawing zone, the area of ecological vegetation, and the duration of appearance of human targets.

[0009] This setting makes it easier to monitor the ecological environment and human interference in the drawdown zone based on the changes in the area covered by the drawdown zone, the area of ecological vegetation, and the duration of the appearance of human targets, so as to maintain the stability of the ecological environment in the drawdown zone.

[0010] In some embodiments, the ratio of the bare land area in the drawdown zone to the entire drawdown zone area, the ratio of the ecological vegetation area to the entire drawdown zone area, and the duration of appearance of the human target are calculated and analyzed based on the image to be analyzed, and the output warning information includes: the drawdown zone area includes water areas and bare land areas; the ecological vegetation in the drawdown zone includes normal herbaceous plants, dead herbaceous plants, normal shrubs and dead shrubs and other areas.

[0011] This setting makes it easier to subdivide the water areas and bare land areas, herbaceous plants and shrub plants in the drawdown zone, and helps to conduct detailed monitoring of ecological vegetation on herbaceous and shrub plants.

[0012] In some embodiments, the ratio of the bare land area in the drawdown zone to the entire drawdown zone area, the ratio of the ecological vegetation area to the entire drawdown zone area, and the duration of appearance of human targets are calculated and analyzed based on the image to be analyzed, and the output warning information includes: a semantic segmentation model, which is used to segment the drawdown zone water area, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs and dead shrubs and other areas and record the corresponding pixel areas in the monitoring image; a human target detection model, which is used to detect human targets in the monitoring image.

[0013] This setting helps to calculate the changes over a period of time based on the area of bare land in the water-logging zone and the proportion of the ecological vegetation area to the entire water-logging zone area. The pixel area is calculated from this, which helps to judge the changes in the bare land area and ecological vegetation over a period of time. At the same time, the entry of people into the water-logging zone area can be known through the human target model.

[0014] In some embodiments, the semantic segmentation model is used to segment areas such as water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs and dead shrubs and record the pixel areas in the monitoring image, including: recording the pixel areas of areas such as water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs and dead shrubs in the monitoring image once every 3-6 hours.

[0015] With this setting, the ecological environment pixel area of the ebb and flow zone is recorded 4-8 times a day, which helps to record the changes in the ebb and flow zone in detail within a day.

[0016] In some embodiments, the semantic segmentation model is used to segment areas such as water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs and dead shrubs and record the pixel area in the monitoring image, including: when the bare land area in the monitoring image occupies more than 20% of the total area of the drawdown zone, or the normal herbaceous plant area plus the normal shrub plant area occupies less than 60% of the total area of the drawdown zone, an early warning message is output.

[0017] This setting makes it easier to judge whether an early warning is needed based on the proportion of bare land area and vegetation area, thereby achieving the purpose of ecological monitoring and early warning management.

[0018] In some embodiments, the human target detection model is used to detect human targets in surveillance images, including: detecting the surveillance image every 1 second, and outputting a warning message if the human target appears in the surveillance image and stays for more than 30 seconds.

[0019] This setting makes it possible to determine whether an early warning is needed based on the time when the human target appears in the monitoring, thereby achieving the early warning purpose of preventing human interference with the ecology of the ebb and flow zone.

[0020] The second aspect of the present disclosure provides an ecological environment monitoring system for a water-logging zone, the system comprising: a monitoring module for obtaining monitoring information of the coverage area, ecological vegetation, and human targets of the water-logging zone; a network transmission module for decoding the monitoring image of the water-logging zone based on the monitoring information, extracting frames at preset time intervals, and obtaining images to be analyzed; a calculation module for calculating and analyzing the proportion of the bare land area of the water-logging zone to the entire water-logging zone area, the proportion of the ecological vegetation area to the entire water-logging zone area, and the duration of appearance of human targets based on the images to be analyzed, and outputting early warning information.

[0021] The technical effects brought about by any possible implementation of the second aspect can refer to the technical effects brought about by the above-mentioned first aspect, and will not be repeated here.

[0022] A third aspect of the present disclosure provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the method for monitoring the ecological environment of a drawdown zone as described in the first aspect when executing the computer program.

[0023] The technical effects brought about by any possible implementation of the third aspect can refer to the technical effects brought about by the above-mentioned first aspect, and will not be repeated here.

[0024] A fourth aspect of the present disclosure provides a computer storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for monitoring the ecological environment of a drawdown zone as described in the first aspect is implemented.

[0025] The technical effects brought about by any possible implementation of the fourth aspect can refer to the technical effects brought about by the above-mentioned first aspect, and will not be repeated here.

[0026] Compared with the prior art, the present disclosure has the following advantages:

[0027] (1) It can realize the real-time perception of the ecological environment changes in the drawdown zone, and carry out early warning and ecological management according to abnormal ecological changes. At the same time, it can also realize real-time perception and early warning of human behavior entering the ecological area of the drawdown zone, so as to avoid human interference in the ecological environment of the drawdown zone.

[0028] (2) A large amount of ecological data of the water-fluctuation zone can be accumulated to provide more data support for the ecological environment management of the water-fluctuation zone.

[0029] Other features and advantages of the present disclosure will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present disclosure. The purposes and other advantages of the present disclosure can be realized and obtained by the structures indicated in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0031] Figure 1 A structural diagram of a video surveillance system provided in an embodiment of the present disclosure;

[0032] Figure 2 A flowchart of video surveillance data processing provided by an embodiment of the present disclosure;

[0033] Figure 3 A flow chart of the method for monitoring the ecological environment of the drawdown zone provided in an embodiment of the present disclosure;

[0034] Figure 4 Schematic diagram of the segmentation of the ecological area of the water-fluctuation zone using the DeepLabV3Plus model provided in an embodiment of the present disclosure;

[0035] Figure 5 Schematic diagram of the CBAM structure and attention module provided in an embodiment of the present disclosure.

[0036] Figure 6 A schematic diagram of human target detection using the YOLO model provided in an embodiment of the present disclosure;

[0037] Figure 7 A block diagram of a water-fluctuation zone ecological environment monitoring system provided in an embodiment of the present disclosure;

[0038] Figure 8 A structural block diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure more clear, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present disclosure without making any creative efforts shall fall within the scope of protection of the present disclosure.

[0040] During the entire monitoring process of the ecological environment of the water-fluctuating zone, surveillance cameras, network transmission modules, edge computing gateways, and water-fluctuating zone video monitoring software are used to realize intelligent monitoring of changes in the water-fluctuating zone area and ecological vegetation changes, and to identify human targets to achieve early warning of human interference with the water-fluctuating zone ecology. Figure 1 and Figure 2 As shown, Figure 1 A structural diagram of a video surveillance system provided in an embodiment of the present disclosure; Figure 2 A flowchart of video surveillance data processing provided in an embodiment of the present disclosure.

[0041] exist Figure 1In this scenario, network cameras are deployed in the watershed zone to collect real-time video data. They are connected to edge computing devices via network cables. The edge computing devices receive the video data transmitted by the network cameras, perform preliminary processing and analysis on it, and then transmit the processed data to the central server via the 5G network. As a data transmission medium, the 5G network ensures fast and stable transmission of video data to the central server. The central server receives the video data from the edge computing devices, further processes and analyzes the data, and generates monitoring results and warning information.

[0042] exist Figure 2 In the monitoring process, network cameras capture images and video streams in real time. Users can set up electronic fences within the monitoring area to detect activity within a specific area. The edge computing gateway then decodes the received video stream, extracts key frames from the video stream for analysis, and segments the extracted frames into ecological regions to identify different ecological zones. It then detects human targets in the video and, based on the electronic fence settings, determines whether the detected person has invaded a specific area. If an intrusion is detected, the time of the intrusion is recorded. It determines whether the intruder's stay in the area exceeds a set threshold and caches the relevant video data for subsequent processing. The central server receives warning signals and related video data from the edge computing device, manually confirms the warning information, and conducts ecological governance based on the warning information and data analysis results to monitor abnormal changes in ecological zones and issue warnings. It also records relevant images and area data for each ecological zone to facilitate law enforcement management based on the warning and analysis results.

[0043] This embodiment provides a method for monitoring the ecological environment of the water-fluctuating zone. By combining edge computing with the processing power of a central server, it can realize intelligent monitoring and management of the ecological environment of the water-fluctuating zone, and can timely detect and handle ecological anomalies and human interference, such as Figure 3 As shown, Figure 3 A flow chart of a method for monitoring the ecological environment of a water-fluctuating zone provided in an embodiment of the present disclosure, the method comprising:

[0044] S100: Obtain monitoring information of the area covered by the ebb and flow zone, ecological vegetation, and human targets.

[0045] In this embodiment, the monitoring information includes monitoring images of the area covered by the ebb and flow zone, the area of ecological vegetation, and the duration of appearance of human targets. The monitoring cameras collect the monitoring images and transmit them to the edge computing gateway device.

[0046] S200 , decoding the monitoring image of the ebb and flow zone based on the monitoring information, extracting frames at preset time intervals, and obtaining an image to be analyzed.

[0047] In this embodiment, the edge computing gateway decodes the surveillance image and extracts frames at set time intervals to obtain the image to be analyzed. The user needs to mark the area to be analyzed (electronic fence) on the initial template image. Surveillance images are usually composed of a large number of continuous frames. Directly processing all frames will consume a lot of computing resources and time. By decoding and extracting frames, the surveillance image can be decomposed into discrete image frames, reducing the amount of data to be processed; the extracted image can be further processed by image enhancement, denoising, etc. to improve the clarity and quality of the image, providing a better data foundation for subsequent analysis.

[0048] S300: Based on the image to be analyzed, the ratio of the bare land area in the water-drawing zone to the entire water-drawing zone area, the ratio of the ecological vegetation area to the entire water-drawing zone area, and the duration of the human target appearance are calculated and analyzed, and warning information is output.

[0049] In the embodiment of the present disclosure, the drawdown zone area includes water areas and bare land areas; the ecological vegetation in the drawdown zone includes normal herbaceous plants, dead herbaceous plants, normal shrub plants and dead shrub plants and other areas.

[0050] For example, analyzing collected surveillance images on an edge computing gateway requires two models. The semantic segmentation model is used to segment areas such as water in the drawdown zone, bare land, healthy herbaceous plants, dead herbaceous plants, healthy shrubs, and dead shrubs, and records the corresponding pixel areas in the surveillance image. The human target detection model is used to detect human targets in the surveillance image.

[0051] For example, the semantic segmentation model is mainly an improved DeepLabV3Plus model, such as Figure 4 As shown, Figure 4 This is a schematic diagram of the improved DeepLabV3Plus model structure provided by the embodiment of the present disclosure. In the figure, an input image, such as a cat picture, is passed through the encoder and decoder, and the semantic segmentation result of the image is output, such as Figure 4 Image segmentation mask of a cat shown in . Figure 4 The Deep Convolutional Neural Network (DCNN) in the encoder is used to extract image features. Atrous convolution (ATV) uses atrous convolutions at different dilation rates (6, 12, and 18) to capture multi-scale features. Image pooling performs global pooling on the image to extract global context. Finally, the results of atrous convolutions at different dilation rates and image pooling are fused to generate multi-scale features.

[0052] Figure 4 In the decoder, 1x1 Conv (1x1 convolution) is used to adjust the number of feature channels; Convolutional Block Attention Module (CBAM) enhances feature expression through channel attention and spatial attention; Upsample by 4 (upsampling 4 times) enlarges the feature map by 4 times, restoring it to a resolution close to that of the input image; Concat (feature splicing): splices the features of the encoder and decoder to fuse multi-scale information; 3x3 Conv (3x3 convolution) further processes the spliced features; Upsample by 4 (upsampling 4 times) again enlarges the feature map to generate the final segmentation result.

[0053] It should be noted that the DeepLabV3Plus model uses MobileNet as the backbone, and the minimum scale of the output feature map is selected as 1 / 16 to further reduce the amount of computation. The network decoder part adds a convolutional block attention module (CBAM module) after each feature fusion, so that the fused features can infer the attention map from two different dimensions, channel and space, to perform adaptive feature refinement.

[0054] like Figure 5 As shown, Figure 5 Schematic diagram of the CBAM structure and attention module provided in an embodiment of the present disclosure. Figure 5 The Input Feature of the CBAM structure is the feature map input to the CBAM module. The Channel Attention Module is used to calculate the channel-level attention weights, extract channel-level statistical information through maximum pooling (MaxPool) and average pooling (AvgPool), use a shared multi-layer perceptron (Shared MLP) to calculate the channel attention weights, and weight the weights to the input features to enhance the information of important channels. The Spatial Attention Module is used to calculate the spatial-level attention weights, perform maximum pooling and average pooling on the features processed by channel attention, extract spatial-level statistical information, use a convolutional layer (conv layer) to calculate the spatial attention weights, and weight the weights to the features to enhance the information of important spatial regions. The Refined Feature is the output feature map after channel attention and spatial attention processing.

[0055] Figure 5In the Channel Attention Module, MaxPool (maximum pooling) is used to perform maximum pooling on the input features and extract the maximum value information at the channel level. AvgPool (average pooling) performs average pooling on the input features and extracts the average value information at the channel level. The Shared MLP (shared multi-layer perceptron) uses a shared multi-layer perceptron to perform a nonlinear transformation on the pooled results to generate channel attention weights. Channel Attention (channel attention) adds the results of maximum pooling and average pooling through the shared MLP to obtain channel attention weights. The Channel-refined feature F' (channel refined feature) in the Spatial Attention Module is the feature processed by channel attention. [MaxPool, AvgPool] performs maximum pooling and average pooling on the channel refined features to extract spatial statistical information. The conv layer (convolutional layer) uses a convolutional layer to process the pooled results and generate spatial attention weights. Spatial Attention adds the results of maximum pooling and average pooling through the convolutional layer to obtain the spatial attention weight.

[0056] It can be understood that by combining channel attention and spatial attention, CBAM can more effectively capture important information in feature maps and improve the performance of the model.

[0057] For example, the human target detection model uses the improved YOLO algorithm to realize the recognition of human targets in the watershed area, so that the detection accuracy is still high when the human target pixel size is smaller than 32x32. Figure 6 As shown, Figure 6 A schematic diagram of human target detection using the YOLO model provided in an embodiment of the present disclosure uses YOLOV8 as the base model, replaces the model's backbone with a simplified self-attention mechanism network Conv2Former-S, and adds a branch from the P2 feature layer in the feature fusion layer to better detect smaller targets.

[0058] exist Figure 6 In the image processing framework, the input network receives an image, typically a live feed from a surveillance camera. The backbone network, using Conv2Former-S, extracts image features. The neck network fuses and enhances these features to generate multi-scale feature maps. The head network generates the final detection results, including bounding boxes and categories.

[0059] Specifically, Patch Embed in Backbone (backbone network) divides the input image into multiple patches and embeds each patch into the feature space; Conv Block is used to extract features; in different stages (P2, P3, P4, P5), the number of P2 channels is 72, including 3 convolution blocks, the number of P3 channels is 144, including 3 convolution blocks, the number of P4 channels is 288, including 12 convolution blocks, and the number of P5 channels is 576, including 3 convolution blocks; feature extraction is to gradually extract higher-level features in each stage to generate feature maps of different resolutions.

[0060] Specifically, the 3x3 Conv layer in the Neck network is used for feature extraction; the C2f module is a feature fusion module for fusing features of different scales; U (upsampling) performs upsampling to amplify the feature map; and C (channel adjustment) adjusts the number of channels in the feature map. Feature fusion generates multi-scale feature maps for subsequent object detection.

[0061] The Detection Head in the Head network is the detection head, used to generate the final detection results; Bbox is bounding box regression, using the CIoU+DFL (distributed focus loss) loss function; Cls is classification, using the BCE (binary cross entropy) loss function. Multi-scale detection performs object detection at four scales: P2, P3, P4, and P5, ensuring detection of objects of different sizes.

[0062] It can be understood that by combining advanced feature extraction and fusion technologies, efficient and accurate human target detection is achieved, providing technical support for protecting the ecological environment of the drawdown zone.

[0063] In some embodiments, the semantic segmentation model can record the pixel area of water areas, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs, and dead shrubs in the monitoring image every 3-6 hours. When the bare land area in the monitoring image accounts for more than 20% of the total area of the entire drawdown zone, or the area of normal herbaceous plants plus the area of normal shrubs accounts for less than 60% of the total area of the drawdown zone, an early warning message is output, triggering an early warning for ecological management of the drawdown zone. It should be noted that the specific ratio threshold can be set according to user needs.

[0064] In some embodiments, the human target detection model detects the surveillance image every 1 second. If the human target appears in the surveillance image and stays for more than 30 seconds, an early warning message is output, which triggers the edge device to report the human interference decay band ecological alarm and 30s intrusion video. At the same time, the edge device saves the subsequent video data until the human target leaves, and uploads the complete intrusion video data to the server for storage.

[0065] In some embodiments, the edge computing device will actively push monitoring images and calculation results to the central server. The server will store historical images and record and analyze the changes in the area of these areas, give early warning information for abnormal changes, and remind relevant personnel to carry out ecological governance; generally, when the area of the drawdown zone shows a trend of continuous decrease, a drawdown zone flooding warning will be prompted; when the area of the drawdown zone shows a trend of continuous increase, a drawdown zone exposure prompt will be prompted and ecological governance judgments and early warnings will be made based on the proportion of each ecological area in the drawdown zone.

[0066] Based on the above method, the embodiment of the present disclosure also provides a water-fluctuation zone ecological environment monitoring system 2000 corresponding to the above method, such as Figure 7 As shown, Figure 7 This is a block diagram of a water-fluctuation zone ecological environment monitoring system provided by an embodiment of the present disclosure. The system 2000 includes a monitoring module 210, a network transmission module 220, and a computing module 230.

[0067] The monitoring module 210 is used to obtain monitoring information of the area covered by the drawdown zone, ecological vegetation, and human targets; the network transmission module 220 is used to decode the monitoring image of the drawdown zone based on the monitoring information, extract frames according to a preset time interval, and obtain the image to be analyzed; the calculation module 230 is used to calculate and analyze the proportion of the bare land area in the drawdown zone to the entire area of the drawdown zone, the proportion of the ecological vegetation area to the entire area of the drawdown zone, and the duration of the appearance of human targets based on the image to be analyzed, and output warning information.

[0068] In this embodiment, the monitoring module can be a surveillance camera, the network transmission module can be a 4G / 5G communication module, a fiber optic transmission module, or an Ethernet transmission module, and the computing module can be an edge computing gateway. Therefore, by combining high-performance video surveillance equipment and technology with network communication technology, data processing technology, and an intelligent management platform, comprehensive, real-time, and remote monitoring of targets is achieved. Furthermore, by incorporating intelligent image processing algorithms and intelligent processing and analysis of real-time monitoring images and videos, automatic monitoring and early warning can be achieved.

[0069] Based on the same inventive concept as the above disclosure, the present disclosure also provides an electronic device 3000. Figure 8 As shown, Figure 8 This is a block diagram of an electronic device structure provided in an embodiment of the present disclosure. The electronic device 3000 in an embodiment of the present disclosure includes at least one processor 310 and at least one memory 320 electrically connected to each other. The memory 320 is electrically connected to the processor 310, wherein the memory 320 stores instructions that can be executed by the at least one processor 310. The instructions are executed by the at least one processor 310 to enable the at least one processor 310 to perform the method described above.

[0070] It should be noted that the electrical connection between the above-mentioned units does not necessarily mean the connection between lines. An indirect connection method can be applied to the embodiments of the present disclosure as long as the purpose of the present disclosure is achieved.

[0071] Based on the same inventive concept, the present disclosure further provides a computer storage medium having a computer program stored thereon, which, when executed by a processor, performs the method described above. The storage medium may include any medium capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0072] Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for monitoring the ecological environment of a fluctuating zone, characterized in that: The method comprises, Obtain monitoring information on the coverage area of the drawdown zone, ecological vegetation, and human targets; Decoding the monitoring image of the fluctuating zone based on the monitoring information, extracting frames at preset time intervals, and obtaining an image to be analyzed; Based on the image to be analyzed, the proportion of the bare land area in the water-logging zone to the entire water-logging zone area, the proportion of the ecological vegetation area to the entire water-logging zone area, and the duration of the human target appearance are calculated and analyzed, and warning information is output.

2. The method according to claim 1, characterized in that The acquisition of monitoring information of the area covered by the water-fluctuation zone, ecological vegetation, and human targets includes: The monitoring information includes monitoring images of the area covered by the ebb and flow zone, the area of ecological vegetation, and the duration of appearance of human targets.

3. The method according to claim 1, characterized in that The calculation and analysis of the ratio of the bare land area in the water-fluctuation zone to the entire water-fluctuation zone area, the ratio of the ecological vegetation area to the entire water-fluctuation zone area, and the duration of the human target appearance based on the image to be analyzed, and the output of the warning information includes: The drawdown zone area includes water areas and bare land areas; The ecological vegetation in the water-fluctuation zone includes areas such as normal herbaceous plants, dead herbaceous plants, normal shrub plants and dead shrub plants.

4. The method according to claim 3, characterized in that The calculation and analysis of the ratio of the bare land area in the water-fluctuation zone to the entire water-fluctuation zone area, the ratio of the ecological vegetation area to the entire water-fluctuation zone area, and the duration of the human target appearance based on the image to be analyzed, and the output of the warning information includes: Semantic segmentation model, used to segment areas such as water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs, and dead shrubs, and record the corresponding pixel areas in the monitoring image; The human target detection model is used to detect human targets in surveillance images.

5. The method according to claim 4, characterized in that The semantic segmentation model is used to segment areas such as water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs, and dead shrubs, and the pixel areas recorded in the monitoring image include: The pixel areas of water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs and dead shrubs in the monitoring images are recorded every 3-6 hours.

6. The method according to claim 5, characterized in that The semantic segmentation model is used to segment areas such as water areas in the drawdown zone, bare land, normal herbaceous plants, dead herbaceous plants, normal shrubs, and dead shrubs, and the pixel areas recorded in the monitoring image include: When the proportion of bare land area in the monitoring image to the total area of the water-flow zone exceeds 20%, or the proportion of normal herbaceous plant area plus normal shrub plant area to the total area of the water-flow zone is less than 60%, an early warning message is output.

7. The method according to claim 4, characterized in that The human target detection model is used to detect human targets in monitoring images and includes: The monitoring image is detected every 1 second. If a human target appears in the monitoring image and stays for more than 30 seconds, an early warning message is output.

8. A fluctuating zone ecological environment monitoring system, characterized in that: The system comprises: Monitoring module: used to obtain monitoring information of the area covered by the water-fluctuation zone, ecological vegetation, and human targets; A network transmission module is used to decode the monitoring image of the fluctuating zone based on the monitoring information, extract frames at preset time intervals, and obtain images to be analyzed; The calculation module is used to calculate and analyze the proportion of the bare land area in the water-logging zone to the entire water-logging zone area, the proportion of the ecological vegetation area to the entire water-logging zone area, and the duration of the appearance of human targets based on the image to be analyzed, and output warning information.

9. An electronic device, characterized in that: The electronic device comprises: memory for storing computer programs; A processor is used to implement the method for monitoring the ecological environment of the drawdown zone as described in any one of claims 1 to 7 when executing the computer program.

10. A computer storage medium, characterized in that The computer storage medium stores a computer program, and when the computer program is executed by the processor, the method for monitoring the ecological environment of the drawdown zone according to any one of claims 1 to 7 is implemented.

Images