Animal monitoring method and device

Abstract

The present application provides an animal monitoring method and device, which belongs to the field of monitoring technology. The method includes: obtaining a monitoring video collected by a target monitoring device; determining whether the monitoring video contains an animal image through an animal detection algorithm; if the monitoring video contains an animal image, extracting animal features from the animal image through an animal feature extraction algorithm; judging whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collected the monitoring video; if the animal contained in the animal image belongs to the animal category to be monitored, outputting the first warning information of the animal category to which the animal belongs. The use of the present application can improve search efficiency and reduce missed detection rate.

Description

Technical Field

[0001] The present application relates to the field of monitoring technology, and in particular to an animal monitoring method and device. Background Art

[0002] There are usually some animals in the area where people live. For example, there may be stray animals due to negligence or deliberate abandonment by their owners, or there may be some wild animals. These animals may affect public health and even endanger people's personal safety. In order to solve the above problems, urban management personnel will manually search for stray animals or wild animals and properly manage the animals found.

[0003] However, manual searches for stray animals or wild animals are relatively inefficient and may miss certain items. Summary of the invention

[0004] The purpose of the embodiment of the present application is to provide an animal monitoring method and device to solve the problems of low search efficiency and high missed detection rate. The specific technical solution is as follows:

[0005] In a first aspect, a method for monitoring an animal is provided, the method comprising:

[0006] Obtain surveillance video collected by the target surveillance device;

[0007] Determining whether the surveillance video contains an animal image by using an animal detection algorithm;

[0008] If the surveillance video contains an animal image, extracting animal features from the animal image using an animal feature extraction algorithm;

[0009] Determining whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video;

[0010] If the animal contained in the animal image belongs to the animal category to be monitored, first warning information of the animal category to which the animal belongs is output.

[0011] Optionally, judging whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information at which the monitoring device that collects the monitoring video is located includes:

[0012] For each animal category to be monitored, calculating the similarity between the animal feature corresponding to the animal category and the extracted animal feature;

[0013] If the similarity is greater than a preset similarity threshold, determining whether the terrain information corresponding to the animal category is the same as the terrain information of the monitoring device that collects the monitoring video;

[0014] If they are the same, it is determined that the animal contained in the animal image belongs to the animal category; if they are not the same, it is determined that the animal contained in the animal image does not belong to the animal category.

[0015] Optionally, if the surveillance video contains an animal image, after extracting the animal features from the animal image using an animal feature extraction algorithm, the method further includes:

[0016] Determining whether the surveillance video contains a pedestrian image by using a pedestrian detection algorithm;

[0017] If the surveillance video contains a pedestrian image, extract the pedestrian's features from the pedestrian image using a pedestrian feature extraction algorithm, and determine whether the pedestrian and the animal contained in the animal image appear together for a duration greater than a preset duration threshold based on an object tracking algorithm;

[0018] If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is not greater than the preset duration threshold, the step of determining whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring equipment that collects the monitoring video is executed.

[0019] Optionally, the method further includes:

[0020] If the surveillance video contains an animal image, determining the animal behavior of the animal contained in the animal image by using an animal behavior detection algorithm;

[0021] If the animal behavior is a preset dangerous behavior, a second warning message is output.

[0022] Optionally, the outputting the first warning information of the animal category to which the animal belongs includes:

[0023] Determine the target warning terminal corresponding to the animal category to which the animal belongs according to the preset correspondence between the animal category and the warning terminal;

[0024] The first warning information of the animal category to which the animal belongs is sent to the target warning terminal.

[0025] In a second aspect, an animal monitoring device is provided, the device comprising:

[0026] An acquisition module is used to acquire the surveillance video collected by the target surveillance device;

[0027] A first determination module, used to determine whether the surveillance video contains an animal image through an animal detection algorithm;

[0028] An extraction module, configured to extract animal features from the animal images by using an animal feature extraction algorithm if the surveillance video contains animal images;

[0029] A judgment module, used to judge whether the animal contained in the animal image belongs to the animal category to be monitored according to the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information where the monitoring device that collects the monitoring video is located;

[0030] The first output module is used for outputting first warning information of the animal category to which the animal belongs if the animal contained in the animal image belongs to the animal category to be monitored.

[0031] Optionally, the judgment module is specifically used to:

[0032] For each animal category to be monitored, calculating the similarity between the animal feature corresponding to the animal category and the extracted animal feature;

[0033] If the similarity is greater than a preset similarity threshold, determining whether the terrain information corresponding to the animal category is the same as the terrain information of the monitoring device that collects the monitoring video;

[0034] If they are the same, it is determined that the animal contained in the animal image belongs to the animal category; if they are not the same, it is determined that the animal contained in the animal image does not belong to the animal category.

[0035] Optionally, the device further comprises:

[0036] A second determination module is used to determine whether the monitoring video contains a pedestrian image through a pedestrian detection algorithm;

[0037] A third determination module is used for extracting features of the pedestrian from the pedestrian image by using a pedestrian feature extraction algorithm if the surveillance video contains a pedestrian image, and determining whether a co-appearance duration of the pedestrian and the animal contained in the animal image is greater than a preset duration threshold according to an object tracking algorithm;

[0038] If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is not greater than the preset duration threshold, the judgment module is triggered to execute the step of judging whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring equipment that collects the monitoring video.

[0039] Optionally, the device further comprises:

[0040] A fourth determination module is used to determine the animal behavior of the animal contained in the animal image by using an animal behavior detection algorithm if the surveillance video contains an animal image;

[0041] The second output module is used to output second warning information if the animal behavior is a preset dangerous behavior.

[0042] Optionally, the first output module is specifically used to:

[0043] Determine the target warning terminal corresponding to the animal category to which the animal belongs according to the preset correspondence between the animal category and the warning terminal;

[0044] The first warning information of the animal category to which the animal belongs is sent to the target warning terminal.

[0045] In a third aspect, a server is provided, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0046] Memory, used to store computer programs;

[0047] The processor is used to implement any method step described in the first aspect when executing a program stored in the memory.

[0048] In a fourth aspect, a computer-readable storage medium is provided, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, any method step described in the first aspect is implemented.

[0049] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute any of the animal monitoring methods described above.

[0050] Beneficial effects of the embodiments of the present application:

[0051] The embodiment of the present application provides an animal monitoring method, which can obtain the monitoring video collected by the target monitoring device, determine whether the monitoring video contains animal images through an animal detection algorithm, and if the monitoring video contains animal images, extract animal features from the animal images through an animal feature extraction algorithm, and determine whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video. If the animal contained in the animal image belongs to the animal category to be monitored, output the first warning information of the animal category to which the animal belongs. In this way, animals of the animal type to be monitored can be automatically identified from the monitoring video and an alarm can be issued without manual search, thereby improving the search efficiency and reducing the missed detection rate.

[0052] Of course, implementing any product or method of the present application does not necessarily require achieving all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0054] Figure 1 A flowchart of an animal monitoring method provided in an embodiment of the present application;

[0055] Figure 2 A flowchart of another animal monitoring method provided in an embodiment of the present application;

[0056] Figure 3 A flowchart of another animal monitoring method provided in an embodiment of the present application;

[0057] Figure 4 A flowchart of another animal monitoring method provided in an embodiment of the present application;

[0058] Figure 5 A flowchart of another animal monitoring method provided in an embodiment of the present application;

[0059] Figure 6 A schematic diagram of the structure of an animal monitoring device provided in an embodiment of the present application;

[0060] Figure 7 A schematic diagram of the structure of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] The embodiment of the present application provides an animal monitoring method, which can be applied to a server, wherein the server can be a background server in a monitoring system. The server can be connected to each monitoring device and receive monitoring videos sent by each monitoring device. The server can detect animal images in the monitoring video through an animal detection algorithm, and then extract animal features from the animal image through an animal feature extraction algorithm. Then, it can be determined whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device where the monitoring video is collected. If the animal contained in the animal image belongs to the animal category to be monitored, the first warning information of the animal category to which the animal belongs is output. Thereby, the animal of the animal type to be monitored can be automatically identified and an alarm can be issued without manual search, thereby improving the search efficiency and reducing the missed detection rate.

[0063] The following will be combined with specific implementation methods to describe in detail an animal monitoring method provided in an embodiment of the present application. Figure 1 As shown, the specific steps are as follows:

[0064] Step 101, obtaining the surveillance video collected by the target surveillance device.

[0065] In the embodiment of the present application, monitoring equipment can be set up in the area to be monitored. For example, cameras can be deployed in the suburbs, or monitoring equipment can be set up near areas where animals are prohibited from entering, such as subway stations, railway stations, shopping malls, etc. Among them, the monitoring equipment can be a monitoring equipment with the characteristics of small size, strong concealment, and self-sustaining. After the monitoring equipment is running, it can collect monitoring videos of the monitoring area to which it belongs in real time, and then the collected monitoring videos can be sent to the background server, and the server can receive the monitoring videos sent by each monitoring device.

[0066] Step 102: Determine whether the surveillance video contains animal images through an animal detection algorithm.

[0067] In the embodiment of the present application, an animal detection algorithm may be pre-stored in the server. After receiving the surveillance video, the server may perform animal detection in the surveillance video according to the surveillance image contained in the surveillance video and the animal detection algorithm, thereby determining whether the surveillance video contains an animal image.

[0068] Specifically, the server can perform detection periodically. Accordingly, for any surveillance video received by the server, after receiving the surveillance image of the surveillance video, the server can extract the surveillance image contained in the surveillance video within a preset time period to perform this calculation. When the next cycle is reached, the calculation is performed again to detect in real time whether there is an animal in the surveillance video. Among them, the server can extract the surveillance image from the surveillance video through OPENCV.

[0069] When the server obtains the surveillance image of the target surveillance video within a preset time length, it can identify whether the surveillance image contains an animal through a preset animal detection algorithm, and can extract the outline of the animal contained in the surveillance image through the animal detection algorithm, and then determine the candidate area frame corresponding to the animal according to the outline of the animal, and then extract the image contained in the candidate area frame to obtain the animal image of the animal. Among them, the animal detection algorithm can be any existing algorithm with animal detection function, such as the YOLO framework algorithm, deep learning neural network, etc. The deep learning neural network can specifically be Faster-RCNN (Faster Region-based Convolutional Neural Network, convolutional neural network based on high-speed region proposal), RCNN, etc.

[0070] In addition, after the server extracts the animal images, it can further screen these animal images. Specifically, for each animal image, the server can determine whether the animal image meets the preset clarity condition and whether the confidence of the animal image is greater than the preset threshold. If the animal image meets the preset clarity condition and the confidence of the animal image is greater than or equal to the preset threshold, it means that the animal image is a usable image and the animal image is saved. If the animal image does not meet the preset clarity condition or the confidence of the animal image is less than the preset threshold, it means that the animal image is an unusable image and the animal image is discarded.

[0071] Optionally, in order to share the pressure on the server, the monitoring system can also be provided with an intelligent terminal (such as a smart box) for edge computing. The intelligent terminal can cover multiple monitoring devices through wifi and connect wires and network cables to improve battery life and network stability. In addition, the intelligent terminal can also communicate with the server. In this scenario, the processing of step 102 can be completed by the intelligent terminal, and the intelligent terminal can report the screened animal image to the server for subsequent processing by the server. The server can determine which monitoring video the received animal image belongs to through the interface with the intelligent terminal or the identification of each monitoring video sent by the intelligent terminal.

[0072] Step 103: If the surveillance video contains an animal image, extract animal features from the animal image using an animal feature extraction algorithm.

[0073] In the embodiment of the present application, after the server obtains the animal image, the animal features can be extracted from the animal image by an animal feature extraction algorithm. The animal feature extraction algorithm can be a ResNet50 network model, or a machine learning algorithm, a deep learning neural network, etc. The deep learning neural network can be Faster-RCNN, RCNN, etc., which is not limited in the embodiment of the present application.

[0074] Step 104 , judging whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information where the monitoring device that collects the monitoring video is located.

[0075] In an embodiment of the present application, technicians can set the animal category that needs to be monitored in the background system. For example, they can upload animal images of the animal category, and the server can store these animal images, and can extract animal features from the animal images and store them.

[0076] The server can determine the terrain information of the monitoring device that collects the monitoring video, and then determine the animal category corresponding to the terrain information based on the correspondence between the preset terrain information and the animal category. The determined animal category is the category of animals that may appear around the monitoring device in the animal category to be monitored. The server can obtain the animal features corresponding to these animal categories, and for each determined animal category, the server can calculate the similarity between the extracted animal features and the animal features corresponding to the animal category. If the similarity is greater than or equal to the preset similarity threshold, it can be determined that the animal contained in the animal image belongs to the animal category to be monitored. If the similarity is less than the preset similarity threshold, it can be determined that the animal contained in the animal image does not belong to the animal category to be monitored.

[0077] Step 105: If the animal contained in the animal image belongs to the animal category to be monitored, output first warning information of the animal category to which the animal belongs.

[0078] In an embodiment of the present application, if the server determines that the animal contained in the animal image belongs to a certain animal category to be monitored (which may be referred to as the target animal category), the first warning information corresponding to the target animal category may be output. Among them, the first warning information may include the timestamp of the monitoring image to which the animal image belongs (i.e., the shooting time of the monitoring image), the location information of the monitoring device, and the identification information of the target animal category. In addition, the first warning information may also include other information, such as the extracted animal image, the similarity corresponding to the animal image, etc., which is not limited in the embodiment of the present application. In addition, in an embodiment of the present application, the server may also store monitoring videos for users to review history and determine the specific information of the animal.

[0079] Optionally, if the server determines that the animal contained in the animal image does not belong to any category of animals to be monitored, it means that the animal is not an animal that needs to be monitored and no processing is required.

[0080] Optional, such as Figure 2 As shown, the above method may further include the following steps.

[0081] Step 106: Determine whether the surveillance video contains a pedestrian image by using a pedestrian detection algorithm.

[0082] In the embodiment of the present application, a pedestrian detection algorithm may be pre-stored in the server. After receiving the surveillance video, the server may also perform pedestrian detection in the surveillance video according to the surveillance image and the pedestrian detection algorithm contained in the surveillance video, thereby determining whether the surveillance video contains a pedestrian image. Pedestrian detection and animal detection may be performed simultaneously, and the process of pedestrian detection is similar to that of animal detection.

[0083] Specifically, for any surveillance video received by the server, after receiving the surveillance image of the surveillance video, the server can extract the surveillance image contained in the surveillance video, and then can identify whether the surveillance image contains pedestrians through a preset pedestrian detection algorithm, and can extract the outline of the pedestrian contained in the surveillance image through the pedestrian detection algorithm, determine the candidate area frame corresponding to the pedestrian according to the outline of the pedestrian, and then extract the image contained in the candidate area frame to obtain the pedestrian image of the pedestrian. Among them, the pedestrian detection algorithm can be any existing algorithm with pedestrian detection function, such as the YOLO framework algorithm, deep learning neural network, etc. The deep learning neural network can specifically be Faster-RCNN (Faster Region-based Convolutional Neural Network, convolutional neural network based on high-speed region proposal), RCNN, etc.

[0084] In addition, after the server extracts the pedestrian images, it can further screen these pedestrian images. Specifically, for each pedestrian image, the server can determine whether the pedestrian image meets the preset clarity condition and whether the confidence of the pedestrian image is greater than the preset threshold. If the pedestrian image meets the preset clarity condition and the confidence of the pedestrian image is greater than or equal to the preset threshold, it means that the pedestrian image is a usable image and the pedestrian image is saved. If the pedestrian image does not meet the preset clarity condition, or the confidence of the pedestrian image is less than the preset threshold, it means that the pedestrian image is an unusable image and the pedestrian image is discarded.

[0085] If the surveillance video contains pedestrian images, step 107 is executed; if not, the process ends.

[0086] Step 107 , extracting pedestrian features from the pedestrian image using a pedestrian feature extraction algorithm, and determining whether the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is greater than a preset duration threshold based on an object tracking algorithm.

[0087] In the embodiment of the present application, the server can extract the features of pedestrians from the pedestrian image through the pedestrian feature extraction algorithm, thereby determining the pedestrians appearing around the animal. The server can track and detect each pedestrian and each animal in the surveillance video according to the object tracking algorithm and the extracted pedestrian features and animal features, thereby determining the duration of simultaneous appearance of the pedestrian and the animal. If the duration of simultaneous appearance of the animal contained in the pedestrian and animal images is not greater than the preset duration threshold, then return to step 104. Otherwise, it is indicated that the animal is the pet of the pedestrian, and the processing ends.

[0088] Optional, such as Figure 3 As shown, the above method may further include the following steps.

[0089] Step 108: If the surveillance video contains an animal image, the animal behavior of the animal contained in the animal image is determined by using an animal behavior detection algorithm.

[0090] In the embodiment of the present application, the server may also be provided with an animal behavior detection algorithm. If the server determines that the surveillance video contains an animal image, the animal behavior of the animal contained in the animal image may be determined by the animal behavior detection algorithm. The animal behavior detection algorithm may be implemented by any model with recognition function in the prior art, such as a deep learning CNN (Convolutional Neural Networks) model, RNN (Recurrent Neural Network), etc., which is not limited in the embodiment of the present application.

[0091] Step 109: If the animal behavior is a preset dangerous behavior, a second warning message is output.

[0092] In the embodiment of the present application, dangerous behaviors that animals may perform can be preset, such as biting or jumping on people. If the server determines that the animal behavior is a preset dangerous behavior, it outputs a second warning message, which may include the location information of the monitoring device, the shooting time, and the identification of the preset dangerous behavior, so that the user can handle it in time.

[0093] Optional, such as Figure 4 As shown, the specific process of outputting the first warning information may include the following steps.

[0094] Step 401, according to the preset correspondence between animal categories and warning terminals, determine the target warning terminal corresponding to the animal category to which the animal belongs.

[0095] In the embodiment of the present application, different animals can be managed by different management structures. For example, stray animals can be managed by a stray animal management association, and wild animals can be managed by a wildlife protection association. Therefore, a correspondence between animal categories and warning terminals can be set. For example, the warning terminal corresponding to stray cats or stray dogs is a terminal of the stray animal management association, and the warning terminal corresponding to wild boars and wild monkeys is a terminal of the wildlife protection association. After the server determines the target animal type to which the animal in the surveillance video belongs, it can determine the target warning terminal corresponding to the target animal category according to the preset correspondence between the animal category and the warning terminal.

[0096] Step 402: Send the first warning information of the animal category to which the animal belongs to the target warning terminal.

[0097] Optional, such as Figure 5 As shown, the specific process of determining whether the animal contained in the animal image belongs to the category of animals to be monitored may include the following steps.

[0098] Step 501 : for each animal category to be monitored, calculate the similarity between the animal feature corresponding to the animal category and the extracted animal feature.

[0099] In the embodiment of the present application, for each animal category to be monitored, the server may also first calculate the similarity between the extracted animal features and the animal features corresponding to the animal category. The animal features may be represented in the form of vectors, and the similarity of the vectors may be calculated.

[0100] Step 502: determine whether the similarity is greater than a preset similarity threshold.

[0101] If the similarity is greater than or equal to the preset similarity threshold, step 503 is executed. If the similarity is less than the preset similarity threshold, the process ends.

[0102] Step 503, determining whether the terrain information corresponding to the animal category is the same as the terrain information where the monitoring device that collects the monitoring video is located.

[0103] If they are the same, then go to step 504. If they are not the same, then go to step 506.

[0104] Step 504: determine whether the animal contained in the animal image belongs to the animal category.

[0105] Step 505: determine whether the animal included in the animal image does not belong to the animal category.

[0106] In the embodiment of the present application, the monitoring video collected by the target monitoring device can be obtained, and the animal detection algorithm can be used to determine whether the monitoring video contains animal images. If the monitoring video contains animal images, the animal features are extracted from the animal images by the animal feature extraction algorithm. According to the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video, it is determined whether the animal contained in the animal image belongs to the animal category to be monitored. If the animal contained in the animal image belongs to the animal category to be monitored, the first warning information of the animal category to which the animal belongs is output. In this way, the animal of the animal type to be monitored can be automatically identified from the monitoring video, and an alarm can be issued without manual search, thereby improving the search efficiency and reducing the missed detection rate.

[0107] The embodiment of the present application also provides a specific example of an animal monitoring method, which is as follows.

[0108] First, the types of animals that may appear in a certain area, such as wild boars and pythons, are entered into the monitoring background system, and pictures of wild boars and pythons can be uploaded.

[0109] The staff sets up cameras C1 to C8 at locations where animals may appear, and places smart box A1 at a location that can be covered by wifi.

[0110] Late at night, a stray dog ​​appears at C4 and a wild boar appears at C7. The smart box identifies the animal images I4 and I7 and uploads them to the server.

[0111] The server calculates the uploaded images I4 and I7 to extract animal features, and compares them with the preset animal features of wild boar and python. It is found that the similarity between image I4 and wild boar is 40%, and between image I7 and python is 2%, and the similarity between image I4 and wild boar is 40%, and between image I7 and python is 2%.

[0112] If the similarity between picture I7 and wild boar is 80%, which is higher than the warning threshold of 70%, the server will send information such as the shooting time of picture I7, the location information of camera C7, and the similarity between picture I7 and wild boar to the staff's mobile phone to remind the staff to handle it.

[0113] The staff judges the behavior and emotions of the animals based on the warning results and captured images I7, and arranges corresponding capture or expulsion operations.

[0114] The embodiment of the present application provides an animal monitoring method, which can obtain the monitoring video collected by the target monitoring device, determine whether the monitoring video contains animal images through an animal detection algorithm, and if the monitoring video contains animal images, extract animal features from the animal images through an animal feature extraction algorithm, and determine whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video. If the animal contained in the animal image belongs to the animal category to be monitored, output the first warning information of the animal category to which the animal belongs. In this way, animals of the animal type to be monitored can be automatically identified from the monitoring video and an alarm can be issued without manual search, thereby improving the search efficiency and reducing the missed detection rate.

[0115] Based on the same technical concept, the present application also provides an animal monitoring device, such as Figure 6 As shown, the device comprises:

[0116] The acquisition module 610 is used to acquire the surveillance video collected by the target surveillance device;

[0117] A first determination module 620, configured to determine whether the surveillance video contains an animal image by using an animal detection algorithm;

[0118] An extraction module 630 is used to extract animal features from the animal images by using an animal feature extraction algorithm if the surveillance video contains animal images;

[0119] A judgment module 640 is used to judge whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information where the monitoring device that collects the monitoring video is located;

[0120] The first output module 650 is configured to output first warning information of the animal category to which the animal belongs if the animal contained in the animal image belongs to the animal category to be monitored.

[0121] Optionally, the determination module 640 is specifically configured to:

[0122] For each animal category to be monitored, calculating the similarity between the animal features corresponding to the animal category and the extracted animal features;

[0123] If the similarity is greater than a preset similarity threshold, determining whether the terrain information corresponding to the animal category is the same as the terrain information where the monitoring device that collects the monitoring video is located;

[0124] If they are the same, it is determined that the animal contained in the animal image belongs to the animal category; if they are not the same, it is determined that the animal contained in the animal image does not belong to the animal category.

[0125] Optionally, the device further comprises:

[0126] A second determination module is used to determine whether the surveillance video contains a pedestrian image through a pedestrian detection algorithm;

[0127] A third determination module is used to extract the features of the pedestrian from the pedestrian image by using a pedestrian feature extraction algorithm if the surveillance video contains a pedestrian image, and determine whether the simultaneous appearance time of the pedestrian and the animal contained in the animal image is greater than a preset time threshold according to an object tracking algorithm;

[0128] If the duration of simultaneous appearance of pedestrians and animals contained in the animal image is not greater than the preset duration threshold, the judgment module 640 is triggered to execute the step of judging whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring equipment that collects the monitoring video.

[0129] Optionally, the device further comprises:

[0130] A fourth determination module is used to determine the animal behavior of the animal contained in the animal image by using an animal behavior detection algorithm if the surveillance video contains an animal image;

[0131] The second output module is used to output a second warning message if the animal behavior is a preset dangerous behavior.

[0132] Optionally, the first output module 650 is specifically configured to:

[0133] According to the preset correspondence between the animal category and the warning terminal, determine the target warning terminal corresponding to the animal category to which the animal belongs;

[0134] The first warning information of the animal category to which the animal belongs is sent to the target warning terminal.

[0135] The embodiment of the present application provides an animal monitoring device, which can obtain the monitoring video collected by the target monitoring device, determine whether the monitoring video contains animal images through an animal detection algorithm, and if the monitoring video contains animal images, extract animal features from the animal images through an animal feature extraction algorithm, and determine whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video. If the animal contained in the animal image belongs to the animal category to be monitored, output the first warning information of the animal category to which the animal belongs. In this way, animals of the animal type to be monitored can be automatically identified from the monitoring video and an alarm can be issued without manual search, thereby improving the search efficiency and reducing the missed detection rate.

[0136] Based on the same technical concept, the embodiment of the present invention also provides a server, such as Figure 7 As shown, it includes a processor 701, a communication interface 702, a memory 703 and a communication bus 704, wherein the processor 701, the communication interface 702, and the memory 703 communicate with each other through the communication bus 704.

[0137] Memory 703, used for storing computer programs;

[0138] The processor 701 is used to execute the program stored in the memory 703 to implement the following steps:

[0139] Obtain surveillance video collected by the target surveillance device;

[0140] Determining whether the surveillance video contains an animal image by an animal detection algorithm;

[0141] If the surveillance video contains an animal image, extracting animal features from the animal image using an animal feature extraction algorithm;

[0142] Determining whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video;

[0143] If the animal contained in the animal image belongs to the animal category to be monitored, first warning information of the animal category to which the animal belongs is output.

[0144] Optionally, judging whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information at which the monitoring device that collects the monitoring video is located includes:

[0145] For each animal category to be monitored, calculating the similarity between the animal feature corresponding to the animal category and the extracted animal feature;

[0146] If the similarity is greater than a preset similarity threshold, determining whether the terrain information corresponding to the animal category is the same as the terrain information of the monitoring device that collects the monitoring video;

[0147] If they are the same, it is determined that the animal contained in the animal image belongs to the animal category; if they are not the same, it is determined that the animal contained in the animal image does not belong to the animal category.

[0148] Optionally, if the surveillance video contains an animal image, after extracting the animal features from the animal image using an animal feature extraction algorithm, the method further includes:

[0149] Determining whether the surveillance video contains a pedestrian image by using a pedestrian detection algorithm;

[0150] If the surveillance video contains a pedestrian image, extract the pedestrian's features from the pedestrian image using a pedestrian feature extraction algorithm, and determine whether the pedestrian and the animal contained in the animal image appear together for a duration greater than a preset duration threshold based on an object tracking algorithm;

[0151] If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is not greater than the preset duration threshold, the step of determining whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring equipment that collects the monitoring video is executed.

[0152] Optionally, the method further includes:

[0153] If the surveillance video contains an animal image, determining the animal behavior of the animal contained in the animal image by using an animal behavior detection algorithm;

[0154] If the animal behavior is a preset dangerous behavior, a second warning message is output.

[0155] Optionally, the outputting the first warning information of the animal category to which the animal belongs includes:

[0156] Determine the target warning terminal corresponding to the animal category to which the animal belongs according to the preset correspondence between the animal category and the warning terminal;

[0157] The first warning information of the animal category to which the animal belongs is sent to the target warning terminal.

[0158] The communication bus mentioned in the above server can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0159] The communication interface is used for communication between the above server and other devices.

[0160] The memory may include a random access memory (RAM) or a non-volatile memory (NVM), such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0161] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0162] In another embodiment of the present invention, a computer-readable storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned animal monitoring methods are implemented.

[0163] In another embodiment of the present invention, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute any of the animal monitoring methods in the above embodiments.

[0164] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk Solid State Disk (SSD)), etc.

[0165] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0166] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. An animal monitoring method, characterized in that: The method comprises: Obtain surveillance video collected by the target surveillance device; Determining whether the surveillance video contains an animal image by an animal detection algorithm; If the surveillance video contains an animal image, extracting animal features from the animal image using an animal feature extraction algorithm; Determining whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video; If the animal contained in the animal image belongs to the animal category to be monitored, outputting first warning information of the animal category to which the animal belongs; Wherein, if the surveillance video contains an animal image, after extracting the animal features from the animal image by using an animal feature extraction algorithm, the method further includes: Determining whether the surveillance video contains a pedestrian image by using a pedestrian detection algorithm; If the surveillance video contains a pedestrian image, extract the pedestrian's features from the pedestrian image using a pedestrian feature extraction algorithm, and determine whether the pedestrian and the animal contained in the animal image appear together for a duration greater than a preset duration threshold based on an object tracking algorithm; If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is not greater than the preset duration threshold, the step of determining whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video; If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is greater than the preset duration threshold, the step of determining whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring equipment that collects the monitoring video will not be performed.

2. The method according to claim 1, characterized in that: The determining whether the animal included in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information where the monitoring device that collects the monitoring video is located includes: For each animal category to be monitored, calculating the similarity between the animal feature corresponding to the animal category and the extracted animal feature; If the similarity is greater than a preset similarity threshold, determining whether the terrain information corresponding to the animal category is the same as the terrain information of the monitoring device that collects the monitoring video; If they are the same, it is determined that the animal contained in the animal image belongs to the animal category; if they are not the same, it is determined that the animal contained in the animal image does not belong to the animal category.

3. The method according to claim 1, characterized in that The method further comprises: If the surveillance video contains an animal image, determining the animal behavior of the animal contained in the animal image by using an animal behavior detection algorithm; If the animal behavior is a preset dangerous behavior, a second warning message is output.

4. The method according to claim 1, characterized in that: The outputting of the first warning information of the animal category to which the animal belongs includes: Determine the target warning terminal corresponding to the animal category to which the animal belongs according to the preset correspondence between the animal category and the warning terminal; The first warning information of the animal category to which the animal belongs is sent to the target warning terminal.

5. An animal monitoring device, characterized in that: The device comprises: An acquisition module is used to acquire the surveillance video collected by the target surveillance device; A first determination module, used to determine whether the surveillance video contains an animal image through an animal detection algorithm; An extraction module, configured to extract animal features from the animal images by using an animal feature extraction algorithm if the surveillance video contains animal images; A judgment module, used to judge whether the animal contained in the animal image belongs to the animal category to be monitored according to the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information where the monitoring device that collects the monitoring video is located; A first output module, configured to output first warning information of the animal category to which the animal belongs if the animal contained in the animal image belongs to the animal category to be monitored; The device also includes: A second determination module is used to determine whether the monitoring video contains a pedestrian image through a pedestrian detection algorithm; A third determination module is used for extracting features of the pedestrian from the pedestrian image by using a pedestrian feature extraction algorithm if the surveillance video contains a pedestrian image, and determining whether a co-appearance duration of the pedestrian and the animal contained in the animal image is greater than a preset duration threshold according to an object tracking algorithm; If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is not greater than the preset duration threshold, the judgment module is triggered to execute the step of judging whether the animal contained in the animal image belongs to the animal category to be monitored according to the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring device that collects the monitoring video; If the duration of simultaneous appearance of the pedestrian and the animal contained in the animal image is greater than the preset duration threshold, the step of determining whether the animal contained in the animal image belongs to the animal category to be monitored based on the animal features corresponding to the animal category to be monitored, the extracted animal features, and the terrain information of the monitoring equipment that collects the monitoring video will not be performed.

6. The device according to claim 5, characterized in that The judgment module is specifically used for: For each animal category to be monitored, calculating the similarity between the animal feature corresponding to the animal category and the extracted animal feature; If the similarity is greater than a preset similarity threshold, determining whether the terrain information corresponding to the animal category is the same as the terrain information of the monitoring device that collects the monitoring video; If they are the same, it is determined that the animal contained in the animal image belongs to the animal category; if they are not the same, it is determined that the animal contained in the animal image does not belong to the animal category.

7. A server, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, for implementing the method steps described in any one of claims 1 to 5 when executing a program stored in a memory.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps described in any one of claims 1 to 5 are implemented.

Images