Wild animal monitoring system based on wireless ad hoc network technology

Abstract

The invention belongs to the technical field of wild animal monitoring, and particularly relates to a wild animal monitoring system based on a wireless ad hoc network technology, which comprises an ad hoc network infrared camera used for automatically acquiring wild animal image data; the central station is used for receiving the wild animal image data and transmitting the wild animal image data to the backbone network; the backbone network is used for transmitting wild animal image data; through the arrangement of the image filtering module, after the Ad-Hoc network infrared camera monitors the signs of wild animals based on the pyroelectric sensor, multiple wild animal images can be shot in a short time, and then based on the image filtering module, false triggering images, namely images without the wild animals in the images, are screened out. In addition, based on the multiple remaining wild animal images, the similarity degree of the multiple remaining wild animal images can be analyzed, and if multiple wild animal images are similar to one another, only one wild animal image is reserved, so that low-value images can be effectively screened out, namely, the data volume is reduced, and the data uploading efficiency is prevented from being reduced.

Description

Technical Field

[0001] The invention belongs to the technical field of wild animal monitoring, in particular to a wild animal monitoring system based on wireless ad hoc network technology. Background Art

[0002] The wildlife monitoring system is a means and method to monitor the number, species, living environment and migration routes of regional wildlife. By monitoring regional wildlife, it can help relevant departments better protect and develop wildlife resources and maintain species diversity.

[0003] In the prior art, for example, the patent with publication number CN112614148A discloses a wildlife monitoring system based on edge-cloud intelligent architecture. The system collects wildlife images by setting edge devices, and then uses gateway devices and data centers to transmit and process wildlife images. In addition, the system uses a false trigger image pre-screening mechanism and optimizes the adaptive neural network algorithm to pre-screen the images stored in the edge device, that is, the infrared camera. Specifically, the images without objects taken by the infrared camera are deleted, thereby reducing the volume of data upload and improving data transmission efficiency.

[0004] Based on the above, in the above-mentioned prior art, it is mentioned that due to mechanism problems, infrared cameras may capture images without objects, which may lead to excessively large data upload volume and affect data transmission efficiency. However, in addition to capturing images without objects, infrared cameras, based on the captured objects, assume that the object does not take any action in a short period of time. In the multiple images captured in a short period of time, the key features that may be extracted by the infrared camera are consistent. Therefore, when processing multiple similar images, it is difficult to extract different key features. That is to say, if the multiple images captured by the infrared camera in a short period of time are similar images, repeated images will also lead to excessive data volume, thereby affecting data transmission efficiency.

[0005] To this end, the present invention provides a wild animal monitoring system based on wireless ad hoc network technology. Summary of the invention

[0006] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0007] The technical solution adopted by the present invention to solve its technical problem is: a wildlife monitoring system based on wireless ad hoc network technology described in the present invention includes an ad hoc network infrared camera for automatic collection of wildlife image data;

[0008] A central station for receiving wildlife image data and transmitting it to the backbone network;

[0009] Backbone network for transmitting wildlife image data;

[0010] Power supply system, used to supply power to the ad hoc network infrared camera and data processing center;

[0011] A central computer room, including a data processing server, for receiving wildlife image data transmitted by the backbone network and analyzing the images to automatically identify wildlife species;

[0012] The central computer room also includes a wildlife species matching module, which is used to match the identified wildlife species with the pre-stored wildlife species atlas:

[0013] When a wildlife species does not match any wildlife species in the Wildlife Species Atlas, the wildlife is marked as an invasive species;

[0014] When a wildlife species matches any wildlife species in the Wildlife Species Atlas, the wildlife is marked as a known species;

[0015] It also includes an early warning module for issuing early warning information when a wildlife species is identified as an invasive species.

[0016] Preferably, the ad hoc network infrared camera further comprises an image filtering module, and the image filtering module is used to remove low-value images, wherein the low-value images include false triggering images and similar images;

[0017] The image filtering module includes an image processing unit, which processes the wild animal image based on the grayscale technology to separate the foreground image from the background image in the image.

[0018] Preferably, the method for eliminating false trigger images is:

[0019] Based on the image processing unit, the foreground image and background image are separated for N wildlife images within a unit time:

[0020] If any wildlife image has a foreground image, it is defined as a valid image;

[0021] If there is no foreground image in any wildlife image, it is defined as a false trigger image;

[0022] All wildlife images defined as false trigger images were removed.

[0023] Preferably, it further comprises a calculation unit, which calculates the similarity of the outer enclosing contour by traversing the outer enclosing contour corresponding to the foreground image in the valid image;

[0024] According to the formula:

[0025]

[0026] Among them, x and y represent two images respectively;

[0027] μ x , μ y are the average grayscale values ​​of images x and y respectively;

[0028] σ x , σ y are the standard deviations of images x and y, respectively, used to characterize the contrast of images x and y;

[0029] σ xy is the covariance of images x and y, which is used to characterize the structure of images x and y;

[0030] C1 and C2 are small constants used to maintain a stable divisor. Usually C1 = (0.01 × L) 2 and C2 = (0.03 × L) 2 , L is the dynamic range of image pixel values, for an 8-bit image L=255.

[0031] Preferably, it also includes an analysis unit, if the similarity between the two images is greater than a threshold, only one of the images is retained, if the similarity between the two images is less than the threshold, both images are retained;

[0032] Traverse all remaining valid images within a unit time and eliminate similar images.

[0033] Preferably, it also includes a drone module, and the drone module is used to collect wild animal image data, and the method is:

[0034] Obtain the layout of the ad hoc network infrared cameras and central stations;

[0035] Develop the data collection route of the drone according to the layout of the self-organizing network infrared camera and the central station;

[0036] The drone approaches any self-organizing network infrared camera or central station according to the data collection route, and wakes up any self-organizing network infrared camera or central station;

[0037] Adopt long-distance wireless transmission to receive wildlife image data from current ad hoc network infrared cameras or central stations;

[0038] After the data collection is completed, the drone returns or continues to perform the data collection task according to the data collection route.

[0039] Preferably, the drone module further includes a response judgment unit. When the drone approaches any ad hoc network infrared camera or central station and wakes up any ad hoc network infrared camera or central station:

[0040] If any of the ad hoc network infrared cameras or the central station does not respond within the set time, the drone will return or continue to perform the collection task according to the data collection route;

[0041] The failure of any ad hoc network infrared camera or central station to respond within the set time includes a wake-up response and a link establishment response, and the set time includes a wake-up response countdown and a link establishment response countdown; the wake-up response corresponds to the wake-up response countdown; the link establishment response corresponds to the link establishment response countdown.

[0042] Preferably, the central computer room also includes a storage module for storing ad hoc network infrared camera information, wherein the ad hoc network infrared camera information includes the ad hoc network infrared camera number and ad hoc network infrared camera coordinate information; the storage module is also used to store wild animal species illustrations.

[0043] Preferably, it also includes a tracing and tracking module, which is based on multiple wild animal images taken by multiple self-organizing network infrared cameras, and takes the shooting time of the wild animal images as a condition, sorts the images taken by multiple self-organizing network infrared cameras in time, and then draws the movement trajectory corresponding to the wild animal in sequence according to the order of the self-organizing network infrared cameras that have completed the sorting.

[0044] Preferably, when any of the wild animals is marked as an invasive species, the method for issuing warning information is:

[0045] Obtaining the drawn historical movement trajectory corresponding to the wild animal;

[0046] Update the movement trajectory based on the wild animal images taken by any infrared camera in the ad hoc network and received in real time by the central computer room;

[0047] According to the updated moving trajectory, a fixed distance around the updated moving trajectory is set as a dangerous activity area;

[0048] The dangerous activity areas and the real-time received wildlife images are combined into early warning information and released.

[0049] The beneficial effects of the present invention are as follows:

[0050] 1. The wildlife monitoring system based on wireless ad hoc network technology described in the present invention can capture multiple wildlife images in a short time after the ad hoc network infrared camera detects signs of wildlife based on the pyroelectric sensor by setting an image filtering module, and then filter out the false triggering images, that is, the images that do not contain wildlife, based on the image filtering module. In addition, the similarity between the remaining multiple wildlife images can be analyzed. If there are multiple wildlife images that are similar to each other, only one wildlife image is retained, so that low-value images can be effectively filtered out, that is, the volume of data is reduced, and the efficiency of data upload is avoided;

[0051] 2. The wildlife monitoring system based on wireless ad hoc network technology described in the present invention sets an image filtering module and uses an algorithm to identify falsely triggered images and similar images, thereby eliminating low-value images such as falsely triggered images and similar images, and retaining images that are high-value images. The images can be used to identify and analyze wildlife species, observe wildlife activities, etc. In addition, the similarity SSIM is used to calculate the wildlife images stored in the ad hoc network infrared camera, and the similarity between the two images is judged by the image brightness, i.e., grayscale value, contrast, and structural information. By calculating the similarity of the two images, it is possible to determine whether the two images are similar images. If they are similar images, one of them is retained; if not, both are retained. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The present invention will be further described below in conjunction with the accompanying drawings.

[0053] Figure 1 is a flow chart of the present invention;

[0054] Figure 2 It is a flow chart of the backbone network using a bridge mode in the present invention;

[0055] Figure 3 It is a flowchart of the backbone network adopting the MESH chain networking mode in the present invention;

[0056] Figure 4 It is a flow chart of the backbone network using a bridge in conjunction with a drone in the present invention;

[0057] Figure 5 It is a flow chart of the data collection by the UAV in the present invention. DETAILED DESCRIPTION

[0058] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0059] like Figures 1 to 4 As shown, a wildlife monitoring system based on wireless ad hoc network technology according to an embodiment of the present invention includes an ad hoc network infrared camera for automatic collection of wildlife image data;

[0060] A central station for receiving wildlife image data and transmitting it to the backbone network;

[0061] Backbone network for transmitting wildlife image data;

[0062] Power supply system, used to supply power to the ad hoc network infrared camera and data processing center;

[0063] A central computer room, including a data processing server, for receiving wildlife image data transmitted by the backbone network and analyzing the images to automatically identify wildlife species;

[0064] The central computer room also includes a wildlife species matching module, which is used to match the identified wildlife species with the pre-stored wildlife species atlas:

[0065] When a wildlife species does not match any wildlife species in the Wildlife Species Atlas, the wildlife is marked as an invasive species;

[0066] When a wildlife species matches any wildlife species in the Wildlife Species Atlas, the wildlife is marked as a known species;

[0067] It also includes an early warning module for issuing early warning information when a wildlife species is identified as an invasive species.

[0068] In the prior art, it is mentioned that due to the mechanism problem, the infrared camera may take images without objects, which leads to excessive data upload volume and affects the data transmission efficiency. However, in addition to taking images without objects, the infrared camera, based on the photographed object, assumes that the object does not take any action in a short period of time, then the infrared camera may be able to extract the same key features in multiple images taken in a short period of time. Therefore, when processing multiple similar images, it is difficult to extract different key features. That is to say, if the multiple images taken by the infrared camera in a short period of time are similar images, the repeated images will also lead to excessive data volume, which affects the data transmission efficiency. Based on the above, as an embodiment of the present invention, by setting an image filtering module, after the self-organizing network infrared camera detects signs of wild animals based on the pyroelectric sensor, it can take multiple images of wild animals in a short period of time, and then based on the image filtering module, the false triggering image, that is, the image that does not contain wild animals in the image is filtered out. In addition, it can also analyze the similarity between the remaining multiple wild animal images. If there are multiple wild animal images that are similar to each other, only one wild animal image is retained, so that low-value images can be effectively filtered out, that is, the volume of data is reduced, and the efficiency of data upload is avoided.

[0069] In addition, the local area network is formed by self-organizing infrared cameras and central stations, using OFDM and MIMO technology, flat system architecture design, supporting multi-band and multi-bandwidth automatic allocation, with strong penetration and diffraction capabilities, and covering forest areas blocked by trees and mountains, with a coverage radius of 1-3km (line of sight can reach 5km). It also has the characteristics of 30Mbps high speed, strong anti-interference and low power consumption. The wildlife image data captured by the self-organizing network infrared camera is transmitted to the central station in real time. After receiving the data, the central station stores it in the micro server or transmits it to the gateway through the backbone network and then to the central computer room. In areas where the network communication is relatively poor, if there is a terminal or the data transmission rate is slow, the wildlife image to be uploaded can be saved in the central station by identifying the current network environment, and then continued to be transmitted when the network environment is good. The use of wireless self-organizing network technology solves the problem of monitoring data backhaul in areas without network coverage, and improves the convenience and efficiency of monitoring work. When the central computer room receives the wildlife image, it is displayed through the command system or display screen in the central computer room. The central computer room is also equipped with an AI server, which uses the AI ​​server to call the image data in the database for artificial intelligence analysis and processing, and automatically identifies species information, classification, statistics, event warning, etc.

[0070] Specifically, in one embodiment of the present invention, by pre-storing the wildlife species atlas of the area, after the AI ​​server identifies the wildlife, it is matched with the wildlife atlas. If there is a match, the wildlife is marked as a known species; the known species include common species and endangered species; if there is no match, the wildlife is marked as an invasive species. It is worth noting that in this area, known species can form a complete food chain and maintain biodiversity. When invasive species appear, they may destroy species diversity or destroy the food chain, making it difficult for the remaining known species to survive. Therefore, when the wild animals discovered in real time are marked as invasive species, an early warning module is needed to issue an early warning and at the same time notify the relevant departments to deal with the invasive species.

[0071] Furthermore, the central station includes a communication micro base station and a micro server, the backbone network includes any one of a public network base station, an optical cable, a bridge, a drone, and a MESH chain network; the power supply system includes any one of solar panels and wind-solar complementary power generation equipment.

[0072] The central station integrates low-power long-distance communication micro base stations and micro servers, and an external high-gain fiberglass omnidirectional antenna. It has a small size, low power consumption, and strong diffraction capability. It can be installed by pole, which is convenient for construction and installation, with small engineering workload, low cost, and little impact on the environment. Since the monitoring system is deployed in the wild, the images of wild animals captured need to be uploaded to the server so that the images of wild animals can be identified and analyzed. For the local area network composed of the self-organizing network infrared camera and the central station, wild animals can be photographed in areas without network coverage, and the images of wild animals can be uploaded and saved in the central station, and then uploaded to the central computer room via the backbone network. Among them, the equipment used in the backbone network is selected according to the actual situation of the protected area. For the protected area based on the wild environment, power supply is the basis for maintaining the central station and the self-organizing network infrared camera. Therefore, solar panels and wind-solar complementary power generation equipment are used as power supply; it is worth noting that the self-organizing network infrared camera can use 9 large-capacity rechargeable 18650 batteries with 5W low-power solar panels; the central station uses a large-capacity rechargeable large-capacity battery and a high-power wind-solar complementary power generation system to ensure uninterrupted power supply on rainy days for 15 days; among them, in order to save power consumption, the self-organizing network infrared camera, the central station and the backbone network can adopt the appointment working mode, and the data taken the previous day will be uploaded uniformly when the appointment time arrives, and the rest of the time is in low-power sleep or power-off state.

[0073] In one embodiment, the ad hoc network infrared camera further includes an image filtering module, and the image filtering module is used to remove low-value images, and the low-value images include false trigger images and similar images;

[0074] The image filtering module includes an image processing unit, which processes the wild animal image based on the grayscale technology to separate the foreground image from the background image in the image.

[0075] Based on the above, in the process of automatically shooting wildlife images, traditional infrared cameras may simultaneously contain a large number of waste films, that is, store a large number of images that do not contain wildlife images. Since infrared cameras use pyroelectric sensors to monitor signs of wildlife, they are prone to false triggering images, that is, images that do not contain wildlife. If the infrared camera cannot filter out such images, it is easy to cause the infrared camera to upload too much data. In areas without network coverage or areas with poor network quality, the system burden is increased, resulting in a high failure rate. In addition, in addition to shooting false triggering images, since the existing infrared camera shooting generally selects a short-term multiple shooting mode in theory, the movement trajectory of the wildlife can be fed back through multiple time-series images, so as to facilitate the server to analyze and identify the species of wildlife. However, in actual application, wild animals may not maintain a fixed movement trajectory in a short period of time. For example, if a wild animal enters the recognition range of the infrared camera and remains relatively still, there may be multiple similar images in the large number of images shot by the infrared camera in a short period of time, which may not be helpful for identifying and analyzing the species of wild animals. On the contrary, the same as false triggering images, it will reduce the efficiency of data upload and increase the system burden. Based on this, in one embodiment of the present invention, an image filtering module is set in the ad hoc network infrared camera, and an algorithm is used to identify false trigger images and similar images, so that low-value images such as false trigger images and similar images are eliminated, and the retained images are high-value images, which can be used to identify and analyze wild animal species, observe wild animal activities, etc.;

[0076] Based on the above, in order to eliminate the low-value images taken by the ad hoc network infrared camera, in one embodiment of the present invention, the collected wildlife images are processed by an image processing unit, and the wildlife images are processed by grayscale technology to generate a mask image. According to the input image, a grayscale image with the same size as the wildlife image is first established, and then the grayscale image is compared with the historical background image pixel by pixel. If the difference is less than the threshold, it is considered that the comparison is successful, and the pixels with successful comparison in the grayscale image are eliminated. The outer surrounding contour formed by the remaining pixels that have not been successfully compared is the foreground part. Finally, all pixels included in the outer surrounding contour are subjected to pixel bitwise operation with the wildlife image to obtain the foreground image; based on the processing of all wildlife images in a unit time, it is possible to analyze and determine whether there is a false triggering image, and then eliminate the false triggering image, which can reduce the system burden and reduce the amount of uploaded data.

[0077] In one embodiment, the method for eliminating false trigger images is:

[0078] Based on the image processing unit, the foreground image and background image are separated for N wildlife images within a unit time:

[0079] If any wildlife image has a foreground image, it is defined as a valid image;

[0080] If there is no foreground image in any wildlife image, it is defined as a false trigger image;

[0081] All wildlife images defined as false trigger images were removed.

[0082] Based on the above, after all wildlife images within a unit time are processed by the image processing unit, the wildlife images with foreground images are marked as valid images, and the wildlife images without foreground images are marked as falsely triggered images. Through analysis and judgment, all wildlife images marked as falsely triggered images are eliminated, and the remaining wildlife images are valid images. In addition, the valid images may also contain similar images, so the similar images in the valid images are eliminated again, and the remaining wildlife images are high-value images.

[0083] In one embodiment, it further includes a calculation unit, which calculates the similarity of the outer enclosing contour by traversing the outer enclosing contour corresponding to the foreground image in the valid image;

[0084] According to the formula:

[0085]

[0086] Among them, x and y represent two images respectively;

[0087] μ x , μ y are the average grayscale values ​​of images x and y respectively;

[0088] σ x , σ y are the standard deviations of images x and y, respectively, used to characterize the contrast of images x and y;

[0089] σ xy is the covariance of images x and y, which is used to characterize the structure of images x and y;

[0090] C1 and C2 are small constants used to maintain a stable divisor. Usually C1 = (0.01 × L) 2 and C2 = (0.03 × L) 2 , L is the dynamic range of image pixel values, for an 8-bit image L=255.

[0091] Based on the above, after eliminating the falsely triggered images in all wildlife images within a unit time, similar images in the remaining valid images need to be eliminated. Unlike falsely triggered images, the similarity between two images or multiple images cannot be intuitively judged based on grayscale processing, and calculation is required to reflect the similarity between the two images. In one embodiment of the present invention, similarity SSIM calculation is adopted, and the similarity between two images is judged by image brightness, i.e. grayscale value, contrast and structural information, wherein the value range of SSIM is [-1, 1], and the larger the value, the more similar the images are, and the smaller the value, the greater the difference between the images. By calculating the similarity of two images, it is possible to obtain whether the two images are similar images. If they are similar images, one of them is retained, and if not, both are retained.

[0092] In one embodiment, the system further includes an analyzing unit, which retains only one of the images if the similarity between the two images is greater than a threshold, and retains both images if the similarity between the two images is less than the threshold;

[0093] Traverse all remaining valid images within a unit time and eliminate similar images.

[0094] Among them, similarity is used to calculate the similarity of two images. Assuming that all valid images remaining in a unit time are recorded as Pi, similarity calculations are performed on P1 and P2-Pi respectively to filter out images similar to P1. The remaining images are sequentially and then similarity calculations are performed with the remaining images in sequence. For example, if P1 and P2 are both retained after the first round of calculations, similarity calculations are performed on P2 and P3-Pi respectively to filter out images similar to P2, until similarity calculations are performed on Pi-1 and Pi. After all valid images are calculated, all remaining images are retained as high-value images.

[0095] In one embodiment, if Figure 5 As shown, it also includes a drone module, and the drone module is used to collect wild animal image data, and the method is as follows:

[0096] Obtain the layout of the ad hoc network infrared cameras and central stations;

[0097] Develop the data collection route of the drone according to the layout of the self-organizing network infrared camera and the central station;

[0098] The drone approaches any self-organizing network infrared camera or central station according to the data collection route, and wakes up any self-organizing network infrared camera or central station;

[0099] Adopt long-distance wireless transmission to receive wildlife image data from current ad hoc network infrared cameras or central stations;

[0100] After the data collection is completed, the drone returns or continues to perform the data collection task according to the data collection route.

[0101] Since the ad hoc network infrared cameras and the central station need to upload the stored wildlife images to the central computer room through the backbone network, if a certain link is broken, the data may not be uploaded in time at the scheduled time, resulting in data loss. Therefore, alternative data collection is required. In one embodiment of the present invention, a data collection route is formulated based on the local area network formed by the ad hoc network infrared cameras and the central station, and drones are used to perform collection tasks corresponding to the data collection route. The central computer room can issue data collection tasks and corresponding data collection routes to the drones based on the points corresponding to the ad hoc network infrared cameras or central stations that have not uploaded data in time. The drones perform collection tasks according to the data collection routes, which can overcome the problems of manual data collection difficulty and untimely data collection in the prior art. Specifically When the UAV performs the collection task according to the data collection route, it approaches the ad hoc network infrared camera or central station that has not uploaded data in time, wakes up the infrared camera through low-power networking technology, and then collects the data of the ad hoc network infrared camera or central station through the long-distance wireless transmission module. The collected data is stored in the UAV storage module on the UAV. When the data collection of any ad hoc network infrared camera or central station is completed, the UAV chooses to return or continue to perform the collection task according to the data collection route. It is worth noting that if the current ad hoc network infrared camera or central station is the last collection point, the UAV returns, if not, it continues to perform the collection task; in addition, if the storage space of the current UAV storage module is limited, it returns and arranges another UAV to perform the collection task.

[0102] In one embodiment, if Figure 5 As shown, the drone module also includes a response judgment unit. When the drone approaches any ad hoc network infrared camera or central station and wakes up any ad hoc network infrared camera or central station:

[0103] If any of the ad hoc network infrared cameras or the central station does not respond within the set time, the drone will return or continue to perform the collection task according to the data collection route;

[0104] The failure of any ad hoc network infrared camera or central station to respond within the set time includes a wake-up response and a link establishment response, and the set time includes a wake-up response countdown and a link establishment response countdown; the wake-up response corresponds to the wake-up response countdown; the link establishment response corresponds to the link establishment response countdown.

[0105] In addition to the fact that storage space will affect whether the drone returns or continues to perform the collection task, it is also necessary to consider whether the ad hoc network infrared camera or central station can respond to the drone within the set time. If it cannot respond in time, it is assumed that the ad hoc network infrared camera may be faulty. In one embodiment of the present invention, in order to optimize the process of automatic data collection by the drone, when any ad hoc network infrared camera or central station does not respond to the drone within the set time, the drone chooses to return or continue to perform the collection task. It is worth noting that if the current ad hoc network infrared camera or central station is the last collection point, the drone returns. If not, it continues to perform the collection task. If necessary, if there is an ad hoc network infrared camera or central station that has not responded within the set time, the current ad hoc network infrared camera or central station needs to be marked to facilitate later maintenance.

[0106] It is worth noting that the above-mentioned set time includes the wake-up response countdown and the link establishment response countdown, and the corresponding responses include the wake-up response and the link establishment response. When the drone wakes up any self-organizing network infrared camera or central station, it is possible that any of the self-organizing network infrared camera or central station fails and cannot respond, or the response is normal but cannot establish a link with the drone normally. Among them, the establishment of a link between any self-organizing network infrared camera or central station and the drone is the basis for long-distance wireless transmission of data. If the link is not established, data transmission cannot be performed. Therefore, it is necessary to set the response countdown and the link establishment response countdown to feedback the problem that any self-organizing network infrared camera and the central station fail to upload data within the scheduled time.

[0107] In one embodiment, the central computer room further includes a storage module for storing ad hoc network infrared camera information, wherein the ad hoc network infrared camera information includes the ad hoc network infrared camera number and ad hoc network infrared camera coordinate information; the storage module is also used to store wild animal species illustrations.

[0108] A storage module is set in the central computer room, which is used to store the point layout of the self-organizing network infrared camera on the one hand, and to store the wild animal species atlas on the other hand. In one embodiment of the present invention, when the data processing server receives the wild animal image, it will analyze and identify the wild animal image based on AI. Specifically, after AI identifies the wild animal species, it will also extract the relevant features of the wild animal image, and then match it with the wild animal species in the wild animal species atlas in the storage module. If the match is successful, the wild animal is marked as a known species, otherwise it is an invasive species; it is worth noting that the wild animal species atlas claimed here is the known species in the area. After the AI ​​identification is completed, the wild animal species atlas is used for comparative analysis to identify whether the wild animal belongs to the known species in the area. If it is an invasive species, it may cause the collapse of species diversity or food chain, and then cause serious problems. In addition, if the invasive species is aggressive, it may affect human activities in the area. Therefore, after the wild animal is identified, it is particularly important to judge whether the wild animal belongs to the area.

[0109] In one embodiment, it also includes a traceability and tracking module, which is based on multiple wild animal images taken by multiple self-organizing network infrared cameras. The multiple self-organizing network infrared cameras are time-sorted based on the shooting time of the wild animal images, and then the movement trajectory corresponding to the wild animal is drawn in sequence according to the order of the self-organizing network infrared cameras that have completed the sorting.

[0110] Based on the above, if after AI has identified a wild animal, it is analyzed that the wild animal is an invasive species, it may cause irreversible damage to other animals and plants in the area. In order to prevent the food chain in the area from collapsing, the wild animal needs to be tracked and traced. In one embodiment of the present invention, a pre-arranged ad hoc network infrared camera can automatically capture images of the wild animal when the wild animal passes by. If the wild animal is an invasive species, the movement trajectory of the wild animal can be drawn by the ad hoc network infrared camera with known coordinates and numbers. The movement trajectory corresponding to the invasive species is drawn, which is helpful for relevant departments to deal with the invasive species and prevent the invasive species from causing the collapse of the food chain in the area and even affecting human activities.

[0111] In addition, if the wild animal is an endangered animal, it is also necessary to use the traceability and tracking module to track the wild animal. If the wild animal has survival problems and needs rescue by relevant departments, the traceability and tracking module will send the movement trajectory to the early warning module and issue an alarm simultaneously.

[0112] In one embodiment, when any of the wild animals is marked as an invasive species, the method for issuing warning information is:

[0113] Obtaining the drawn historical movement trajectory corresponding to the wild animal;

[0114] Update the movement trajectory based on the wild animal images taken by any infrared camera in the ad hoc network and received in real time by the central computer room;

[0115] According to the updated moving trajectory, a fixed distance around the updated moving trajectory is set as a dangerous activity area;

[0116] The dangerous activity areas and the real-time received wildlife images are combined into early warning information and released.

[0117] As an embodiment of the present invention, when there are invasive species in the area, it is necessary to update the movement trajectory of the invasive species in real time to facilitate the relevant departments to deal with the invasive species. The movement trajectory of the wild animal is updated through a self-organizing network infrared camera with known points and numbers. As an updated movement trajectory, a dangerous activity area needs to be set for aggressive invasive species. As mentioned above, according to the updated movement trajectory, 500m around the updated movement trajectory is used as a dangerous activity area, and the dangerous activity area is combined with the time of receiving the wild animal image, that is, the time when the latest sign of the wild animal is discovered, as a warning information release, to facilitate the relevant departments to deal with the invasive species, and at the same time, it can also warn humans in the dangerous activity area.

[0118] Working principle: In the prior art, it is mentioned that due to the mechanism problem, the infrared camera may take images without objects, which will lead to excessive data upload volume and affect the data transmission efficiency. However, in addition to taking images without objects, the infrared camera is based on the photographed object. Assuming that the object does not take any action in a short period of time, the key features that may be extracted by the infrared camera in multiple images taken in a short time are consistent. Therefore, when processing multiple similar images, it is difficult to extract different key features. That is to say, if the multiple images taken by the infrared camera in a short time are similar images, the repeated images will also lead to excessive data volume, which will affect the data transmission efficiency. Based on the above, as an embodiment of the present invention, by setting an image filtering module, after the self-organizing network infrared camera detects signs of wild animals based on the pyroelectric sensor, it can take multiple images of wild animals in a short time, and then based on the image filtering module, the false triggering image, that is, the image that does not contain wild animals in the image is filtered out. In addition, it can also analyze the similarity between the remaining multiple wild animal images. If there are multiple wild animal images that are similar to each other, only one wild animal image is retained, so that low-value images can be effectively filtered out, that is, the volume of data is reduced, and the efficiency of data upload is avoided.

[0119] In addition, the local area network is formed by self-organizing infrared cameras and central stations, using OFDM and MIMO technology, flat system architecture design, supporting multi-band and multi-bandwidth automatic allocation, with strong penetration and diffraction capabilities, and covering forest areas blocked by trees and mountains, with a coverage radius of 1-3km (line of sight can reach 5km). It also has the characteristics of 30Mbps high speed, strong anti-interference and low power consumption. The wildlife image data captured by the self-organizing network infrared camera is transmitted to the central station in real time. After receiving the data, the central station stores it in the micro server or transmits it to the gateway through the backbone network and then to the central computer room. In areas where the network communication is relatively poor, if there is a terminal or the data transmission rate is slow, the wildlife image to be uploaded can be saved in the central station by identifying the current network environment, and then continued to be transmitted when the network environment is good. The use of wireless self-organizing network technology solves the problem of monitoring data backhaul in areas without network coverage, and improves the convenience and efficiency of monitoring work. When the central computer room receives the wildlife image, it is displayed through the command system or display screen in the central computer room. The central computer room is also equipped with an AI server, which uses the AI ​​server to call the image data in the database for artificial intelligence analysis and processing, and automatically identifies species information, classification, statistics, event warning, etc.

[0120] Specifically, in one embodiment of the present invention, by pre-storing the wildlife species atlas of the area, after the AI ​​server identifies the wildlife, it is matched with the wildlife atlas. If there is a match, the wildlife is marked as a known species; the known species include common species and endangered species; if there is no match, the wildlife is marked as an invasive species. It is worth noting that in this area, known species can form a complete food chain and maintain biodiversity. When invasive species appear, they may destroy species diversity or destroy the food chain, making it difficult for the remaining known species to survive. Therefore, when the wild animals discovered in real time are marked as invasive species, an early warning module is needed to issue an early warning and at the same time notify the relevant departments to deal with the invasive species.

[0121] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A wildlife monitoring system based on wireless ad hoc network technology, characterized in that: include: Self-organizing infrared cameras for automated collection of wildlife image data; A central station for receiving wildlife image data and transmitting it to the backbone network; Backbone network for transmitting wildlife image data; Power supply system, used to supply power to the ad hoc network infrared camera and data processing center; A central computer room, including a data processing server, for receiving wildlife image data transmitted by the backbone network and analyzing the images to automatically identify wildlife species; The central computer room also includes a wildlife species matching module, which is used to match the identified wildlife species with the pre-stored wildlife species atlas: When a wildlife species does not match any wildlife species in the Wildlife Species Atlas, the wildlife is marked as an invasive species; When a wildlife species matches any wildlife species in the Wildlife Species Atlas, the wildlife is marked as a known species; It also includes an early warning module for issuing early warning information when a wildlife species is identified as an invasive species.

2. The wildlife monitoring system based on wireless ad hoc network technology according to claim 1 is characterized in that: The ad hoc network infrared camera also includes an image filtering module, which is used to remove low-value images, including false trigger images and similar images; The image filtering module includes an image processing unit, which processes the wild animal image based on the grayscale technology to separate the foreground image from the background image in the image.

3. The wildlife monitoring system based on wireless ad hoc network technology according to claim 2 is characterized in that: The method for eliminating false triggering images is: Based on the image processing unit, the foreground image and background image are separated for N wildlife images within a unit time: If any wildlife image has a foreground image, it is defined as a valid image; If there is no foreground image in any wildlife image, it is defined as a false trigger image; All wildlife images defined as false trigger images were removed.

4. The wildlife monitoring system based on wireless ad hoc network technology according to claim 3 is characterized in that: It also includes a calculation unit, which calculates the similarity of the outer enclosing contour by traversing the outer enclosing contour corresponding to the foreground image in the effective image; According to the formula: Among them, x and y represent two images respectively; μ x , μ y are the average grayscale values ​​of images x and y respectively; σ x , σ y are the standard deviations of images x and y, respectively, used to characterize the contrast of images x and y; σ xy is the covariance of images x and y, which is used to characterize the structure of images x and y; C1 and C2 are small constants used to maintain a stable divisor. Usually C1 = (0.01 × L) 2 and C2 = (0.03 × L) 2 , L is the dynamic range of image pixel values, for an 8-bit image L=255.

5. The wildlife monitoring system based on wireless ad hoc network technology according to claim 4 is characterized in that: It also includes an analysis unit, if the similarity between the two images is greater than a threshold, only one of the images is retained, if the similarity between the two images is less than the threshold, both images are retained; Traverse all remaining valid images within a unit time and eliminate similar images.

6. The wildlife monitoring system based on wireless ad hoc network technology according to claim 1 is characterized in that: It also includes a drone module, which is used to collect wildlife image data by: Obtain the layout of the ad hoc network infrared cameras and central stations; Develop the data collection route of the drone according to the layout of the self-organizing network infrared camera and the central station; The drone approaches any self-organizing network infrared camera or central station according to the data collection route, and wakes up any self-organizing network infrared camera or central station; Adopt long-distance wireless transmission to receive wildlife image data from current ad hoc network infrared cameras or central stations; After the data collection is completed, the drone returns or continues to perform the data collection task according to the data collection route.

7. The wildlife monitoring system based on wireless ad hoc network technology according to claim 6 is characterized by: The drone module also includes a response judgment unit. When the drone approaches any ad hoc network infrared camera or central station and wakes up any ad hoc network infrared camera or central station: If any of the ad hoc network infrared cameras or the central station does not respond within the set time, the drone will return or continue to perform the collection task according to the data collection route; The failure of any ad hoc network infrared camera or central station to respond within the set time includes a wake-up response and a link establishment response, and the set time includes a wake-up response countdown and a link establishment response countdown; the wake-up response corresponds to the wake-up response countdown; the link establishment response corresponds to the link establishment response countdown.

8. The wildlife monitoring system based on wireless ad hoc network technology according to claim 1 is characterized in that: The central computer room also includes a storage module for storing ad hoc network infrared camera information, wherein the ad hoc network infrared camera information includes the ad hoc network infrared camera number and ad hoc network infrared camera coordinate information; the storage module is also used to store wild animal species illustrations.

9. The wildlife monitoring system based on wireless ad hoc network technology according to claim 8 is characterized in that: It also includes a traceability and tracking module, which is based on multiple wild animal images taken by multiple self-organizing network infrared cameras. The images taken by multiple self-organizing network infrared cameras are sorted in time based on the shooting time of the wild animal images, and then the movement trajectory corresponding to the wild animal is drawn in sequence according to the order of the self-organizing network infrared cameras that have completed the sorting.

10. The wildlife monitoring system based on wireless ad hoc network technology according to claim 1, characterized in that: When any of the above wildlife is marked as an invasive species, the method of issuing warning information is: Obtaining the drawn historical movement trajectory corresponding to the wild animal; Update the movement trajectory based on the wild animal images taken by any infrared camera in the ad hoc network and received in real time by the central computer room; According to the updated moving trajectory, a fixed distance around the updated moving trajectory is set as a dangerous activity area; The dangerous activity areas and the real-time received wildlife images are combined into early warning information and released.

Images