UAV-assisted intelligent breeding management method and system

Abstract

The embodiment of the present invention relates to the field of smart farming technology, and discloses a method for smart farming management based on drone assistance, including: receiving first image information collected when the first drone is at a first flight altitude; inputting the preprocessed first image information into a pre-built animal recognition model for identification to determine the image information of each animal in the first image information; determining the corresponding flight trajectory information according to the cluster center point; sending the flight trajectory information to the first drone to control the first drone to collect the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude. The method for smart farming management based on drone assistance in the embodiment of the present invention can intelligently plan the flight trajectory of the drone based on the density information of the animal distribution. This not only improves the working efficiency of the drone, but also ensures the comprehensiveness and accuracy of data collection.

Description

Technical Field

[0001] The present invention relates to the technical field of smart breeding management, and in particular to a method and system for unmanned aerial vehicle-assisted smart breeding management. Background Art

[0002] At present, as the breeding industry continues to develop towards scale and industrialization, many large-scale breeding companies have introduced information and intelligent breeding management systems. However, most of these systems focus on monitoring and regulating the breeding environment and improving the quality of breeding. As for the control of the breeding herd structure, it currently mainly relies on manual experience and professional judgment, and there is no more in-depth research on open breeding environments. This also means that there are still certain blind spots in smart breeding, and it cannot achieve better auxiliary breeding functions. Summary of the invention

[0003] In view of the above-mentioned defects, an embodiment of the present invention discloses a method for intelligent breeding management based on the assistance of drones, which can achieve efficient breeding management.

[0004] The first aspect of the embodiment of the present invention discloses a method for intelligent farming management based on drone assistance, including:

[0005] receiving first image information collected when the first unmanned aerial vehicle is at a first flight altitude, and performing an image preprocessing operation on the first image information;

[0006] Inputting the preprocessed first image information into a pre-built animal recognition model for recognition to determine image information of each animal in the first image information, and determining the center point position of each animal in the first image information according to the animal image information;

[0007] Determine density distribution image information of each animal in the first image information according to the center point position of each animal in the first image information and the first image information, perform clustering operation on the density distribution image information to determine multiple cluster center points in the density distribution image information, and determine corresponding flight trajectory information according to the cluster center points;

[0008] The flight trajectory information is sent to the first UAV to control the first UAV to collect the second image information group and the infrared image information group at a second flight altitude along the set flight trajectory, and the acquired second image information group and infrared image information group are sent to the background server; wherein, the second image information group includes multiple second image information, the infrared image information group includes multiple infrared image information, and the second image information corresponds to the infrared image information one by one.

[0009] As an optional implementation manner, in the first aspect of the embodiment of the present invention, after sending the acquired second image information group and infrared image information group to the background server, the method further includes:

[0010] Performing feature recognition on each second image information in the second image information group according to a pre-built target object recognition model to determine corresponding target object information and target position information corresponding to the target object information;

[0011] Matching the determined target object information with the constructed target object-animal number relationship table to determine the corresponding animal number information;

[0012] The infrared image information is subjected to temperature analysis to determine the temperature information corresponding to the corresponding target object, and the animal number information is data-associated with the temperature information according to the target position information; if the temperature information corresponding to the target object does not match the set threshold range, the animal number information with the temperature abnormality will be sent to the breeding management personnel.

[0013] As an optional implementation manner, in the first aspect of the embodiment of the present invention, the receiving first image information collected when the first drone is at the first flight altitude, and performing an image preprocessing operation on the first image information includes:

[0014] A first image information group collected by a first unmanned aerial vehicle at a first flight altitude is received, wherein the first image information group includes a plurality of first image information; the first image information group is obtained by the following steps:

[0015] Acquire first image information captured at a corresponding collection point through an onboard camera of the first unmanned aerial vehicle, and acquire attitude information corresponding to the first image information through an attitude sensor;

[0016] Determining, according to the posture data, whether a shooting state of the onboard camera when shooting the first image information satisfies a preset shaking condition;

[0017] When it is determined that the shooting state meets a preset shaking condition, marking the first image information, and performing a grayscale operation on the first image information meeting the shaking condition to obtain a first grayscale image;

[0018] The contour feature recognition is performed on the first grayscale image by using a background segmentation model to determine the contour image information of each animal in the first image information.

[0019] As an optional implementation, in the first aspect of the embodiment of the present invention, the animal recognition model is constructed by the following steps:

[0020] Acquire sample image information, where the sample image information is a sample image taken by the drone at a first set height, and manually mark the animals that appear;

[0021] Dividing the sample image information to obtain a model training set and a model test set;

[0022] Inputting the model training set into a pre-built animal recognition model for recognition, and continuously adjusting network parameters until the animal recognition model meets the set conditions;

[0023] The model test set is input into an animal recognition model that meets the set conditions for testing and verification to determine whether the set requirements are met. If so, the corresponding animal recognition model is output; if not, it returns to continue training.

[0024] As an optional implementation, in the first aspect of the embodiment of the present invention, the inputting the pre-processed first image information into a pre-built animal recognition model for recognition to determine the image information of each animal in the first image information includes:

[0025] Divide the preprocessed first image information into a first recognition area, a second recognition area and a third recognition area; wherein the first recognition area, the second recognition area and the third recognition area have the same center point; and the distances from the edge of the first recognition area, the edge of the second recognition area and the edge of the third recognition area to the center point increase in sequence;

[0026] Inputting the first recognition area, the second recognition area and the third recognition area into the pre-built first animal recognition model, the second animal recognition model and the third animal recognition model respectively for recognition to determine the image information of each animal in the first image information;

[0027] When the first UAV collects the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude, the sensor parameters of each area are acquired by the sensor component arranged at the first UAV. The sensor component includes a gas sensor, a temperature sensor and a humidity sensor, and the environmental parameter information of the corresponding area is determined according to the sensor parameters of each area.

[0028] As an optional implementation manner, in the first aspect of the embodiment of the present invention, the breeding management method further includes:

[0029] Receiving motion sensing signals collected by motion sensors installed at each animal;

[0030] The animal's motion state is identified by the motion sensing signal, and the animal's health state is determined based on the animal's motion state.

[0031] As an optional implementation manner, in the first aspect of the embodiment of the present invention, after determining the corresponding flight trajectory information according to the cluster center point, the method further includes:

[0032] Performing signal disassembly on the flight trajectory information to determine the position information of each collection point in the flight trajectory information, and generating corresponding delivery information according to the position information of each collection point;

[0033] The flight trajectory information and the delivery information are sent to the second drone so that the second drone delivers the medicine or feed at the corresponding collection point;

[0034] The smart breeding management method further includes:

[0035] In the client management module, a corresponding key pair is generated according to the identity of the client, the key pair includes an encryption key and a decryption key, the encryption key is extracted from the key pair and sent to the client corresponding to the identity, and the decryption key is extracted from the key pair, the decryption key is bound to the identity to generate decryption information, and the decryption information is sent to the client management node at the blockchain system;

[0036] In the data management module, the collected breeding management information is split into multiple public data packets and private data packets, and the private data packets are encrypted using the encryption key, wherein the public data packets include an identity identifier and timestamp information of the breeding management information, and the private data packets include breeding quantity, animal health status and animal number information; the public data packets and the private data packets are distributed and stored in each breeding data block of the blockchain system;

[0037] When acquiring the breeding management information of the data management module, the breeding management node extracts the public data packet and the private data packet from each breeding data block according to the corresponding timestamp information and the identity of the client, queries the decryption information according to the identity, extracts the corresponding decryption key to decrypt the private data packet, and restores the breeding management information based on the public data packet and the private data packet using a preset data template.

[0038] The second aspect of the embodiment of the present invention discloses a drone-assisted smart farming management system, comprising:

[0039] A receiving module: used for receiving first image information collected when the first UAV is at a first flight altitude, and performing image preprocessing operations on the first image information;

[0040] A center point determination module: used for inputting the pre-processed first image information into a pre-built animal recognition model for recognition to determine the image information of each animal in the first image information, and determining the center point position of each animal in the first image information according to the animal image information;

[0041] A flight trajectory determination module: determines the density distribution image information of each animal in the first image information according to the center point position of each animal in the first image information and the first image information, and performs a clustering operation on the density distribution image information to determine a plurality of cluster center points in the density distribution image information, and determines the corresponding flight trajectory information according to the cluster center points;

[0042] Image acquisition module: used for sending the flight trajectory information to the first UAV to control the first UAV to acquire the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude, and sending the acquired second image information group and infrared image information group to the background server; wherein, the second image information group includes a plurality of second image information, the infrared image information group includes a plurality of infrared image information, and the second image information corresponds to the infrared image information one by one.

[0043] The third aspect of an embodiment of the present invention discloses an electronic device, comprising: a memory storing executable program code; a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the drone-assisted smart farming management method disclosed in the first aspect of the embodiment of the present invention.

[0044] A fourth aspect of an embodiment of the present invention discloses a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the drone-assisted smart farming management method disclosed in the first aspect of an embodiment of the present invention.

[0045] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0046] The drone-assisted intelligent farming management method in the embodiment of the present invention can intelligently plan the flight trajectory of the drone based on the density information of animal distribution. This not only improves the working efficiency of the drone, but also ensures the comprehensiveness and accuracy of data collection. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 It is a flow chart of a method for intelligent farming management based on drone assistance disclosed in an embodiment of the present invention;

[0049] Figure 2 is a schematic diagram of a process for acquiring first image information disclosed in an embodiment of the present invention;

[0050] Figure 3 It is a schematic diagram of the process of constructing an animal identification model disclosed in an embodiment of the present invention;

[0051] Figure 4 It is a structural schematic diagram of a drone-assisted smart farming management system provided by an embodiment of the present invention;

[0052] Figure 5 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] It should be noted that the terms "first", "second", "third", "fourth", etc. in the specification and claims of the present invention are used to distinguish different objects rather than to describe a specific order. The terms "including" and "having" in the embodiments of the present invention and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0055] At present, with the continuous development of the breeding industry towards scale and industrialization, many large-scale breeding companies have introduced information-based and intelligent breeding management systems. However, most of these systems focus on the monitoring and regulation of the breeding environment and the improvement of breeding quality. As for the management and control of the structure of breeding herds, it currently mainly relies on manual experience and professional judgment, and there is no more in-depth research on open breeding environments. This also means that there are still certain blind spots in smart breeding, and it is not possible to achieve better auxiliary breeding functions. Based on this, the embodiment of the present invention discloses a method, system, electronic device and storage medium for intelligent breeding management based on drones. According to the density information of animal distribution, the system can intelligently plan the flight trajectory of the drone. This not only improves the working efficiency of the drone, but also ensures the comprehensiveness and accuracy of data collection.

[0056] Example 1: Please refer to Figure 1 , Figure 1 It is a flow chart of the method for intelligent farming management based on drone assistance disclosed in the embodiment of the present invention. Among them, the execution subject of the method described in the embodiment of the present invention is an execution subject composed of software and / or hardware, and the execution subject can receive relevant information by wired or / and wireless means, and can send certain instructions. Of course, it can also have certain processing functions and storage functions. The execution subject can control multiple devices, such as a remote physical server or cloud server and related software, or it can be a local host or server and related software that performs related operations on a device placed somewhere. In some scenarios, multiple storage devices can also be controlled, and the storage devices can be placed in the same place or different places as the devices. For example Figure 1 As shown, the drone-assisted intelligent farming management method includes the following steps:

[0057] S101: receiving first image information collected by a first drone at a first flight altitude, and performing an image preprocessing operation on the first image information;

[0058] S102: inputting the preprocessed first image information into a pre-built animal recognition model for recognition to determine image information of each animal in the first image information, and determining the center point position of each animal in the first image information according to the animal image information;

[0059] S103: determining density distribution image information of each animal in the first image information according to the center point position of each animal in the first image information and the first image information, performing a clustering operation on the density distribution image information to determine a plurality of cluster center points in the density distribution image information, and determining corresponding flight trajectory information according to the cluster center points;

[0060] S104: Send the flight trajectory information to the first UAV to control the first UAV to collect the second image information group and the infrared image information group at a second flight altitude along the set flight trajectory, and send the acquired second image information group and infrared image information group to the background server; wherein, the second image information group includes a plurality of second image information, the infrared image information group includes a plurality of infrared image information, and the second image information corresponds to the infrared image information one by one.

[0061] The technical effects of the drone-assisted intelligent farming management method of the embodiment of the present invention are mainly reflected in the following aspects: by receiving and processing the image information collected by the first drone, the animal in the image can be quickly identified and its position can be determined. This process significantly improves the monitoring efficiency of the farming environment and reduces the time and cost of manual inspections. By determining the position of the animal center point in the first image information and further calculating the density distribution image information, the distribution of animals in the farming environment can be accurately grasped. This helps farmers or managers to reasonably feed and manage animals and ensure the healthy growth of animals. According to the density information of animal distribution, the system can intelligently plan the flight trajectory of the drone. This not only improves the working efficiency of the drone, but also ensures the comprehensiveness and accuracy of data collection. By collecting the second image information group and the infrared image information group and ensuring that the two correspond one to one, the system can obtain more comprehensive and rich farming environment data. Infrared image information can reveal key information such as animal body temperature, providing an important reference for farming management. The entire management process is automated and intelligent, reduces human intervention and errors, and improves the efficiency and accuracy of farming management. At the same time, by acquiring and analyzing farming environment data in real time, problems can be discovered in time and measures can be taken to reduce farming risks.

[0062] During the implementation process, there may be a situation where animals gather in a certain area. At this time, if image collection is performed in the usual order, some empty images will appear, that is, images without animals. This will reduce the overall image collection efficiency. Therefore, when carrying out the specific implementation, clustering is used to determine the density of the corresponding animals, and then the corresponding flight route is determined according to the center point of the density. This can maximize the overall monitoring efficiency and reduce the overall system usage cost.

[0063] When implementing it, it is also necessary to determine the location information of the animals in the sub-area to the edge of the sub-area, because when the image is collected for the second time, it is necessary to determine that the corresponding animals are still in the corresponding area and have not run to other areas, otherwise it will easily cause confusion in the number of statistics. The movement speed of the drone should be greater than the movement speed of the animals so that the drone can reach the specific image collection point more quickly, so that the most accurate graphic data can be obtained.

[0064] The scheme of the embodiment of the present invention determines the travel routes of different drones from different heights. When at a higher position, only the information of the corresponding points can be obtained through images, so that quantitative statistics can be performed. In addition to quantitative statistics, the current group movement status of animals can also be counted; when monitoring the overall movement status of animals, the images obtained by the drone at the first height can also be recognized and processed, but what needs to be obtained here is video information, because only video information can be used for the next step of density change statistics, extracting key frames in the video information, and then performing regional segmentation on the key frames, dividing them into 100*100 square images, and counting the number of each animal in the corresponding area as the corresponding feature vector, judging the feature changes corresponding to adjacent frames, and then determining the movement rules of animal groups.

[0065] During specific implementation, the boundary area is determined based on the different movement state information and movement point information of the animals. The distance between the boundary area and the nearest animal is used to determine the selection range.

[0066] More preferably, after sending the acquired second image information group and infrared image information group to the backend server, the method further includes:

[0067] S105: performing feature recognition on each second image information in the second image information group according to a pre-built target object recognition model to determine corresponding target object information and target position information corresponding to the target object information;

[0068] S106: Matching the determined target object information with the constructed target object-animal number relationship table to determine the corresponding animal number information;

[0069] S107: Perform temperature analysis on the infrared image information to determine the temperature information corresponding to the corresponding target object, and perform data association between the animal number information and the temperature information according to the target position information; if the temperature information corresponding to the target object does not match the set threshold range, the animal number information with the temperature abnormality will be sent to the breeding management personnel.

[0070] In the above implementation, the system can accurately identify the target in the second image information group and determine its corresponding target location information through the pre-constructed target recognition model. This greatly enhances the monitoring capability of specific targets in breeding management. By matching the identified target information with the constructed target-animal number relationship table, the system can quickly determine the animal number information related to the target. This step effectively associates the image recognition results with the breeding management database, providing a basis for subsequent data analysis and processing.

[0071] Temperature analysis of infrared image information can obtain the temperature information of the target object in real time and associate the data with the animal number information. When the temperature information of the target object (usually an animal) does not match the set threshold range, the system can quickly identify the animal with abnormal temperature and send relevant information to the breeding manager. This function can detect animal health problems in time and provide timely and effective early warning for breeding management. The introduction of the above steps makes breeding management more refined. By acquiring and analyzing key data in the breeding environment (such as animal location, temperature, etc.) in real time, the system can provide breeding managers with more comprehensive, accurate and timely breeding management suggestions to help them better understand the breeding situation and improve breeding efficiency and quality.

[0072] The entire process is highly automated and intelligent, reducing manual intervention and errors, and improving the efficiency and accuracy of breeding management. At the same time, through real-time data analysis and early warning functions, the system can autonomously discover and handle problems in the breeding process, reducing breeding risks.

[0073] And when carrying out the specific implementation, each animal needs to be marked on its back. This mark can be used as the marking feature of the corresponding animal in color, shape, or a combination of the two, so as to complete the information correspondence and facilitate subsequent management.

[0074] More preferably, Figure 2 As shown, the receiving of first image information collected when the first drone is at a first flight altitude, and performing an image preprocessing operation on the first image information, includes:

[0075] S1011: Receive a first image information group collected by a first drone at a first flight altitude, wherein the first image information group includes a plurality of first image information; the first image information group is obtained by the following steps:

[0076] S1012: Acquire first image information captured at a corresponding collection point through an onboard camera of the first unmanned aerial vehicle, and acquire attitude information corresponding to the first image information through an attitude sensor;

[0077] S1013: Determine, according to the posture data, whether a shooting state of the onboard camera when shooting the first image information satisfies a preset shaking condition;

[0078] S1014: When it is determined that the shooting state meets a preset shaking condition, marking the first image information, and performing a grayscale operation on the first image information meeting the shaking condition to obtain a first grayscale image;

[0079] S1015: Performing contour feature recognition on the first grayscale image by using a background segmentation model to determine contour image information of each animal in the first image information.

[0080] The above-mentioned implementation method simultaneously obtains image information and attitude information through the airborne camera and attitude sensor of the first drone, ensuring that the environmental state during image acquisition, such as the jitter of the drone, can be taken into account in the subsequent processing, thereby improving the accuracy and integrity of data processing. In step S1013, it is determined whether the image is blurred due to the jitter of the drone according to the attitude data. In step S1014, the image that meets the jitter condition is grayed, which can not only reduce the amount of data processing, but also alleviate the image blur problem caused by jitter to a certain extent and improve the image quality. In step S1015, the contour feature recognition of the grayscale image is performed by adopting the background segmentation model, and the contour information of the animal in the image can be accurately extracted. This provides important basic data for subsequent target object recognition, positioning and temperature analysis, and enhances the accuracy of breeding management. The preprocessing process is highly automated, which reduces manual intervention and improves processing efficiency. At the same time, through intelligent image processing and feature recognition technology, a more intelligent and efficient means is provided for breeding management. The pre-processed image information is clearer and more accurate, which can provide more reliable data support for breeding managers, helping them to better understand the breeding situation and make management decisions in a timely manner, thereby improving breeding management efficiency. The above-mentioned more preferred image pre-processing process effectively improves the accuracy and availability of image information by introducing technical means such as posture information, jitter correction, grayscale processing and contour feature recognition, providing a more solid data foundation and technical support for subsequent breeding management.

[0081] The background segmentation model of the present invention mainly sets different grayscale thresholds for image processing, sets tolerance values ​​to identify images after grayscale processing, and obtains animal contour images. The tolerance value here can be set according to actual needs. When the color difference between the two is not much, the tolerance value can be set to be relatively small, which can also be more convenient for extraction. Since the contour is collected, and the color difference between grassland and grazing animals (such as cattle and sheep) is large, when such a tolerance value is used to obtain the contour image, the calculation speed will be faster, and there is no need to build an animal model, which is a more efficient way to obtain the contour.

[0082] More preferably, Figure 3 As shown, the animal recognition model is constructed by the following steps:

[0083] S1021: Acquire sample image information, where the sample image information is a sample image taken by the drone at a first set height, and manually mark the animals that appear;

[0084] S1022: Divide the sample image information to obtain a model training set and a model test set;

[0085] S1023: inputting the model training set into a pre-built animal recognition model for recognition, and continuously adjusting network parameters until the animal recognition model meets the set conditions;

[0086] S1024: Input the model test set into the animal recognition model that meets the set conditions for testing and verification to determine whether the set requirements are met. If so, output the corresponding animal recognition model; if not, return to continue training.

[0087] The above implementation method uses the sample images taken by the drone at the first set height as the basic data for model training, which can ensure that the model learns the image features of animals in the real environment. At the same time, the animals that appear are marked by manual annotation to ensure the accuracy and reliability of the sample data. The sample image information is divided into a model training set and a model test set, following the common training-testing process in machine learning. The animal recognition model is trained with the training set, and the network parameters are continuously adjusted to optimize the model performance; then the trained model is tested and verified with the test set to ensure the accuracy and generalization ability of the model in practical applications. During the training process, the network parameters are continuously adjusted until the animal recognition model meets the set conditions, and this process realizes the automatic optimization of the model. This can not only improve the recognition accuracy of the model, but also reduce manual intervention and improve the efficiency and stability of model training.

[0088] In the model testing phase, the model is verified through the test set to ensure that the model can meet the set requirements in actual applications. If the model fails to meet the requirements, it will return to continue training until the model performance reaches the preset standard. This iterative optimization method can ensure that the final output animal recognition model has a high degree of reliability and stability.

[0089] By building a high-quality animal recognition model, intelligent monitoring and management of the breeding environment can be achieved. The model can automatically identify animals in the breeding environment, classify and count them, and provide accurate data support for breeding managers, helping them to better understand the breeding situation and improve breeding efficiency and management level. The construction of the above animal model can make its adaptability wider and not limited to the site.

[0090] More preferably, the step of inputting the pre-processed first image information into a pre-built animal recognition model for recognition to determine the image information of each animal in the first image information comprises:

[0091] Divide the preprocessed first image information into a first recognition area, a second recognition area and a third recognition area; wherein the first recognition area, the second recognition area and the third recognition area have the same center point; and the distances from the edge of the first recognition area, the edge of the second recognition area and the edge of the third recognition area to the center point increase in sequence;

[0092] Inputting the first recognition area, the second recognition area and the third recognition area into the pre-built first animal recognition model, the second animal recognition model and the third animal recognition model respectively for recognition to determine the image information of each animal in the first image information;

[0093] When the first UAV collects the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude, the sensor parameters of each area are acquired by the sensor component arranged at the first UAV. The sensor component includes a gas sensor, a temperature sensor and a humidity sensor, and the environmental parameter information of the corresponding area is determined according to the sensor parameters of each area.

[0094] Because not all images taken by drones are of the front and back of animals, it is necessary to build contour models for different areas. For example, the farther away from the border, the more likely it is that the image of a cow will be taken from the side, so there is an angle deviation problem. Therefore, when implementing it, different models can be set for recognition according to the distance. This not only facilitates the construction of the contour model, but also facilitates subsequent rapid recognition, greatly improving the recognition accuracy and efficiency.

[0095] In addition to the above-mentioned angle deviation, there may be two animals close together during the specific implementation. Similarly, the farther away from the drone, the greater the probability of this happening. This is because there is an angle deviation. Therefore, during the specific implementation, it is necessary to build a model for the situation where multiple cows are stacked together. During the specific implementation, when the outline of the cow is detected to be larger than the set value, the animal stacking recognition model is activated; this can improve the recognition accuracy.

[0096] The solution of the embodiment of the present invention can also obtain environmental parameter information of each area in real time through the sensor components (including gas sensors, temperature sensors and humidity sensors) set at the first drone. These environmental parameters are of great significance for understanding the health of the breeding environment, predicting animal behavior and formulating scientific management strategies. Through real-time monitoring, breeding managers can promptly detect environmental anomalies and take corresponding measures to intervene to ensure the stability of the breeding environment and the health of the animals.

[0097] Combining animal identification and environmental parameter monitoring results can provide intelligent decision support for breeding managers. For example, by analyzing the relationship between animal distribution and environmental parameters, breeding density and feed placement strategies can be optimized; by predicting animal behavior, preventive measures can be taken in advance to avoid problems such as animal escape or disease spread.

[0098] More preferably, the breeding management method further comprises:

[0099] Receiving motion sensing signals collected by motion sensors installed at each animal;

[0100] The animal's motion state is identified by the motion sensing signal, and the animal's health state is determined based on the animal's motion state.

[0101] In the above-mentioned more preferred breeding management method, a function of setting a motion sensor on the animal to collect motion sensor signals, and identifying the animal's motion state through these signals, and then determining the animal's health status is introduced. By setting a motion sensor on the animal, the animal's motion state can be monitored in real time and continuously. This real-time data acquisition method enables breeding managers to promptly discover possible health problems of the animal and take quick measures. The data collected by the motion sensor can be accurately processed and analyzed to identify the animal's motion state. These status information, such as activity, movement mode, etc., are important indicators for assessing the health status of the animal. Therefore, this method can more accurately assess the health status of the animal.

[0102] Traditional breeding management methods often require breeders to regularly observe the behavior and physical signs of animals, which is not only time-consuming and labor-intensive, but may also lead to inaccurate evaluation results due to human factors. However, by collecting and analyzing data through motion sensors, management efficiency can be greatly improved and manpower input can be reduced. This breeding management method based on motion sensor data embodies the concept of intelligent breeding. Monitoring and managing animals through automated and data-based means makes the breeding process more scientific and standardized, while also improving the efficiency and benefits of breeding. Through real-time monitoring and analysis of the animal's movement status, possible health problems of animals can be discovered in a timely manner and early warnings can be issued. This helps breeding managers take preventive measures in advance to avoid the deterioration of problems.

[0103] To sum up, the above-mentioned more preferred breeding management method realizes real-time and accurate monitoring and evaluation of animal health status by introducing motion sensors to collect animal motion sensing signals, improves the efficiency and intelligence level of breeding management, and also provides better protection for animal health.

[0104] In the breeding process, animal behavior parameters refer to quantitative indicators used to describe and monitor animal behavior characteristics. These parameters can help breeders understand the health status and welfare level of animals and whether the breeding environment is suitable. The following are some common animal behavior parameters that can be selected and monitored according to different breeding animals and specific needs:

[0105] Activity level: The frequency, speed, and distance of an animal's movement reflect the animal's activity level and energy consumption. Animal activity data is recorded and analyzed through motion sensors, video monitoring systems, and other equipment. Changes in activity levels may be related to the animal's health status, disease state, or discomfort in the breeding environment.

[0106] Eating behavior: The frequency of eating, food intake, and water intake of animals reflect their appetite and nutritional intake. Using smart feeders and waterers, animals’ eating data can be automatically recorded. Abnormal eating behavior may be a sign that the animal is sick or in a poor breeding environment.

[0107] Sleep and rest behavior: The duration, quality, and resting posture of animals reflect their comfort and physiological state. Video monitoring and animal activity sensors are used to observe and analyze the sleeping and resting behaviors of animals. Insufficient sleep and rest may affect the growth, immunity, and reproductive capacity of animals.

[0108] Social behavior: The interactions, communication, and interactions between animals, reflecting the social needs and group structure of the animals. Use video surveillance systems to observe and analyze the social behavior of animals. Abnormal social behavior may be a sign of stress, aggression, or discomfort between animals.

[0109] Self-care behavior: The cleaning, grooming, and grooming of an animal’s body surface reflects the animal’s self-care ability and health status. Assess the animal’s self-care behavior through video monitoring and direct observation. A decrease in self-care behavior may mean that the animal has health problems or is in a poor housing environment.

[0110] Abnormal behavior: unusual or abnormal behavior that is inconsistent with normal behavior patterns, such as excessive anxiety, increased aggression, excessive vocalization, etc. Abnormal behavior is identified and recorded through video monitoring, sound analysis and direct observation. Abnormal behavior may be a direct manifestation of animal illness, injury or an unsuitable breeding environment, requiring timely intervention and treatment.

[0111] When monitoring and analyzing these animal behavior parameters, motion sensors (also known as gyroscope sensors) are used to detect data in order to better understand the animal's behavior patterns and health status. These parameters can provide important decision support for farmers, helping them optimize the breeding environment, improve animal welfare, and reduce disease risks. In specific implementation, proximity sensors can even be set at the above-mentioned important points on the ranch (such as water sources and grasslands) to determine the actual movement intentions of the corresponding grazing animals.

[0112] Animal behavior parameters are quantitative indicators used to describe and monitor animal behavior characteristics during farming. The following are some common animal behavior parameters that help farmers understand the health status, welfare level and whether the breeding environment is suitable for the animals.

[0113] Activity level parameters: Movement distance: The distance an animal moves in a certain period of time is measured through motion sensors and other equipment to reflect the animal's activity level and energy consumption. Behavior frequency: The frequency of an animal performing a certain behavior (such as standing, lying down, walking, etc.) is recorded to evaluate its daily activity pattern.

[0114] Eating behavior parameters: Eating frequency: Record the number of times an animal eats per day or per meal to understand the animal's appetite. Food intake: Use equipment such as smart feeders to measure the animal's food intake to assess its nutritional needs and health. Water intake: Monitor the animal's water intake to help determine its water needs and health.

[0115] Social behavior parameters: Interaction frequency: Detect interactive behaviors between animals (such as mutual grooming, games, etc.), record their frequency and duration, and evaluate the social needs and group structure of animals.

[0116] Self-care behavior parameters: Grooming behavior: Record the frequency and duration of the animal's grooming to assess its self-care ability and health status. Since the grooming behavior response has its own fluctuations in the motion sensor, the corresponding animal's behavior parameters can be determined by determining the detection signal of the motion sensor.

[0117] Cleaning behavior: Detect animals’ cleaning behaviors (such as washing their faces, taking sand baths, etc.) to understand their hygiene habits and health status.

[0118] More preferably, after determining the corresponding flight trajectory information according to the cluster center point, the method further includes:

[0119] Performing signal disassembly on the flight trajectory information to determine the position information of each collection point in the flight trajectory information, and generating corresponding delivery information according to the position information of each collection point;

[0120] The flight trajectory information and the delivery information are sent to the second drone so that the second drone delivers the medicine or feed at the corresponding collection point;

[0121] The above solution can also be applied before clustering, that is, by placing food to make animals gather in a certain area, so that the subsequent image acquisition can be more stable and accurate, and the possibility of deviation can be reduced. When implementing it, the grassland status can also be identified, and animals can be moved to the area desired by the user by placing food or other items that attract animals. For example, animals can be moved from area A to area B, where water and grass are relatively poor and water and grass are relatively lush in area B; the above guidance makes breeding more scientific and facilitates the maintenance of pastures.

[0122] The above-mentioned more preferred breeding management method introduces the functions of signal decomposition of flight trajectory information, generation of delivery information, and delivery of medicine or feed through a second drone. The technical effects of this method are mainly reflected in the following aspects:

[0123] By decomposing the flight trajectory information, the precise location information of each collection point can be determined. The delivery information generated based on this location information can ensure that the medicine or feed is accurately delivered to the target location, avoiding waste of resources and misdelivery. Traditional drug or feed delivery often requires manual operation, which is not only inefficient but also may pose safety risks. Automatic delivery through the second drone can greatly improve the efficiency and safety of delivery and reduce the need for manual intervention. Combining flight trajectory information and delivery information can provide intelligent decision support for breeding managers. For example, the location and delivery amount of the delivery point can be dynamically adjusted according to the distribution of animals and environmental parameter information to achieve the best management effect. Through precise delivery, it can ensure that the medicine or feed is fully utilized and reduce the waste of resources. At the same time, since the selection of delivery points is more reasonable, the impact on the environment can also be reduced.

[0124] The smart breeding management method further includes:

[0125] In the client management module, a corresponding key pair is generated according to the identity of the client, the key pair includes an encryption key and a decryption key, the encryption key is extracted from the key pair and sent to the client corresponding to the identity, and the decryption key is extracted from the key pair, the decryption key is bound to the identity to generate decryption information, and the decryption information is sent to the client management node at the blockchain system;

[0126] In the data management module, the collected breeding management information is split into multiple public data packets and private data packets, and the private data packets are encrypted using the encryption key, wherein the public data packets include an identity identifier and timestamp information of the breeding management information, and the private data packets include breeding quantity, animal health status and animal number information; the public data packets and the private data packets are distributed and stored in each breeding data block of the blockchain system;

[0127] When acquiring the breeding management information of the data management module, the breeding management node extracts the public data packet and the private data packet from each breeding data block according to the corresponding timestamp information and the identity of the client, queries the decryption information according to the identity, extracts the corresponding decryption key to decrypt the private data packet, and restores the breeding management information based on the public data packet and the private data packet using a preset data template.

[0128] Since the information of each ranch and the status of the animals raised on the ranch are relatively important confidential information during the smart farming process, it is necessary to encrypt the above information to prevent others from obtaining the above information. The above smart farming management method achieves safe, reliable and efficient management of farming management information by introducing blockchain technology, customer management module and data management module. The following are the corresponding technical effects:

[0129] The client management module generates a corresponding key pair (encryption key and decryption key) based on the client's identity, ensuring the security of data transmission. The encryption key is sent to the client, and the decryption key is stored in the client management node of the blockchain system after being bound to the identity, which enhances the confidentiality of the data. The distributed storage of the blockchain system makes the data less likely to be tampered with, improving the credibility of the data.

[0130] Flexibility and efficiency of data management: The breeding management information is split into public and private data packets. The public data packets contain identity and timestamp information for quick location and retrieval. The private data packets contain key information such as breeding quantity, animal health status and animal number, which are encrypted with encryption keys to ensure data security. Distributed storage in each breeding data block of the blockchain system makes data access and acquisition more flexible and efficient.

[0131] Intelligent management and decision support: Through the pre-set data templates, the breeding management information is restored, so that the management personnel can easily obtain the required data, monitor and make decisions on breeding activities. Combined with the timestamp information and identity identification, the relevant breeding data can be obtained quickly and accurately, providing strong support for management decisions.

[0132] Data traceability and credibility: Due to the characteristics of blockchain technology, the data stored in it is tamper-proof and traceable, ensuring the authenticity and credibility of aquaculture management information. When disputes arise or verification is required, the source and authenticity of the data can be easily traced and verified. In summary, the smart aquaculture management method achieves safe, reliable and efficient management of aquaculture management information by introducing blockchain technology, customer management module and data management module. This not only improves data security, but also improves the flexibility and efficiency of data management, providing strong support for the development of the aquaculture industry.

[0133] The drone-assisted intelligent farming management method in the embodiment of the present invention can intelligently plan the flight trajectory of the drone based on the density information of animal distribution. This not only improves the working efficiency of the drone, but also ensures the comprehensiveness and accuracy of data collection.

[0134] Example 2: Please refer to Figure 4 , Figure 4 Schematic diagram of the structure of the drone-assisted intelligent farming management system disclosed in the embodiment of the present invention. Figure 4 As shown, the drone-assisted intelligent farming management system may include:

[0135] Receiving module 21: used to receive first image information collected when the first UAV is at a first flight altitude, and perform image preprocessing operations on the first image information;

[0136] The center point determination module 22 is used to input the pre-processed first image information into a pre-built animal recognition model for recognition to determine the image information of each animal in the first image information, and determine the center point position of each animal in the first image information according to the animal image information;

[0137] The flight trajectory determination module 23 determines the density distribution image information of each animal in the first image information according to the center point position of each animal in the first image information and the first image information, and performs a clustering operation on the density distribution image information to determine a plurality of cluster center points in the density distribution image information, and determines the corresponding flight trajectory information according to the cluster center points;

[0138] Image acquisition module 24: used for sending the flight trajectory information to the first UAV to control the first UAV to collect the second image information group and the infrared image information group at the second flight altitude along the set flight trajectory, and send the acquired second image information group and infrared image information group to the background server; wherein, the second image information group includes a plurality of second image information, the infrared image information group includes a plurality of infrared image information, and the second image information corresponds to the infrared image information one by one.

[0139] The drone-assisted intelligent farming management method in the embodiment of the present invention can intelligently plan the flight trajectory of the drone based on the density information of animal distribution. This not only improves the working efficiency of the drone, but also ensures the comprehensiveness and accuracy of data collection.

[0140] Example 3: Please refer to Figure 5 , Figure 5 Schematic diagram of the structure of an electronic device disclosed in an embodiment of the present invention. The electronic device may be a computer, a server, etc. Of course, in certain circumstances, it may also be a smart device such as a mobile phone, a tablet computer, a monitoring terminal, and an image acquisition device with processing functions. Figure 5 As shown, the electronic device may include:

[0141] A memory 510 storing executable program codes;

[0142] a processor 520 coupled to the memory 510;

[0143] Among them, the processor 520 calls the executable program code stored in the memory 510 to execute some or all steps of the drone-assisted smart farming management method in Example 1.

[0144] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute some or all of the steps in the drone-assisted smart farming management method in Embodiment 1.

[0145] An embodiment of the present invention further discloses a computer program product, wherein when the computer program product is run on a computer, the computer executes part or all of the steps in the drone-assisted smart farming management method in embodiment one.

[0146] An embodiment of the present invention further discloses an application publishing platform, wherein the application publishing platform is used to publish a computer program product, wherein when the computer program product runs on a computer, the computer executes some or all of the steps in the drone-assisted smart farming management method in embodiment one.

[0147] In various embodiments of the present invention, it should be understood that the size of the serial numbers of the processes does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0148] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place or distributed over multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0149] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0150] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-accessible memory. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a memory and includes several requests to enable a computer device (which can be a personal computer, server or network device, etc., specifically a processor in a computer device) to perform some or all steps of the method described in each embodiment of the present invention.

[0151] In the embodiments provided by the present invention, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that determining B according to A does not mean determining B only according to A, and B can also be determined according to A and / or other information.

[0152] A person of ordinary skill in the art may understand that part or all of the steps in the various methods of the embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically-erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0153] The above is a detailed introduction to the drone-assisted smart farming management method, system, electronic device and storage medium disclosed in the embodiments of the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A drone-assisted smart farming management method, characterized in that: include: receiving first image information collected when the first unmanned aerial vehicle is at a first flight altitude, and performing an image preprocessing operation on the first image information; Inputting the preprocessed first image information into a pre-built animal recognition model for recognition to determine image information of each animal in the first image information, and determining the center point position of each animal in the first image information according to the animal image information; Determine density distribution image information of each animal in the first image information according to the center point position of each animal in the first image information and the first image information, perform clustering operation on the density distribution image information to determine multiple cluster center points in the density distribution image information, and determine corresponding flight trajectory information according to the cluster center points; Performing signal disassembly on the flight trajectory information to determine the position information of each collection point in the flight trajectory information, and generating corresponding delivery information according to the position information of each collection point; The flight trajectory information and the delivery information are sent to the second drone so that the second drone delivers the medicine or feed at the corresponding collection point; The flight trajectory information is sent to the first drone to control the first drone to collect the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude, and the acquired second image information group and infrared image information group are sent to the backend server; wherein the second image information group includes a plurality of second image information, the infrared image information group includes a plurality of infrared image information, and the second image information corresponds to the infrared image information one by one; Performing feature recognition on each second image information in the second image information group according to a pre-built target object recognition model to determine corresponding target object information and target position information corresponding to the target object information; Matching the determined target object information with the constructed target object-animal number relationship table to determine the corresponding animal number information; The infrared image information is subjected to temperature analysis to determine the temperature information corresponding to the corresponding target object, and the animal number information is data-associated with the temperature information according to the target position information; if the temperature information corresponding to the target object does not match the set threshold range, the animal number information with the temperature abnormality will be sent to the breeding management personnel.

2. The drone-assisted smart farming management method according to claim 1, characterized in that: The receiving first image information collected when the first drone is at the first flight altitude, and performing an image preprocessing operation on the first image information, includes: A first image information group collected by a first unmanned aerial vehicle at a first flight altitude is received, wherein the first image information group includes a plurality of first image information; the first image information group is obtained by the following steps: Acquire first image information captured at a corresponding collection point through an onboard camera of the first unmanned aerial vehicle, and acquire attitude information corresponding to the first image information through an attitude sensor; Determining, according to the posture data, whether a shooting state of the onboard camera when shooting the first image information satisfies a preset shaking condition; When it is determined that the shooting state meets a preset shaking condition, marking the first image information, and performing a grayscale operation on the first image information meeting the shaking condition to obtain a first grayscale image; The contour feature recognition is performed on the first grayscale image by using a background segmentation model to determine the contour image information of each animal in the first image information.

3. The method for intelligent farming management based on drone assistance as claimed in claim 2, characterized in that: The animal recognition model is constructed by the following steps: Acquire sample image information, where the sample image information is a sample image taken by the drone at a first set height, and manually mark the animals that appear; Dividing the sample image information to obtain a model training set and a model test set; Inputting the model training set into a pre-built animal recognition model for recognition, and continuously adjusting network parameters until the animal recognition model meets the set conditions; The model test set is input into an animal recognition model that meets the set conditions for testing and verification to determine whether the set requirements are met. If so, the corresponding animal recognition model is output; if not, it returns to continue training.

4. The method for intelligent farming management based on drone assistance according to claim 1, characterized in that: The step of inputting the pre-processed first image information into a pre-built animal recognition model for recognition to determine the image information of each animal in the first image information includes: Divide the preprocessed first image information into a first recognition area, a second recognition area and a third recognition area; wherein the first recognition area, the second recognition area and the third recognition area have the same center point; and the distances from the edge of the first recognition area, the edge of the second recognition area and the edge of the third recognition area to the center point increase in sequence; Inputting the first recognition area, the second recognition area and the third recognition area into the pre-built first animal recognition model, the second animal recognition model and the third animal recognition model respectively for recognition to determine the image information of each animal in the first image information; When the first UAV collects the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude, the sensor parameters of each area are acquired by the sensor component arranged at the first UAV. The sensor component includes a gas sensor, a temperature sensor and a humidity sensor, and the environmental parameter information of the corresponding area is determined according to the sensor parameters of each area.

5. The drone-assisted smart farming management method according to claim 1, characterized in that: The breeding management method further comprises: Receiving motion sensing signals collected by motion sensors installed at each animal; The animal's motion state is identified by the motion sensing signal, and the animal's health state is determined based on the animal's motion state.

6. The drone-assisted smart farming management method according to claim 1, characterized in that: The smart breeding management method further includes: In the client management module, a corresponding key pair is generated according to the identity of the client, the key pair includes an encryption key and a decryption key, the encryption key is extracted from the key pair and sent to the client corresponding to the identity, and the decryption key is extracted from the key pair, the decryption key is bound to the identity to generate decryption information, and the decryption information is sent to the client management node at the blockchain system; In the data management module, the collected breeding management information is split into multiple public data packets and private data packets, and the private data packets are encrypted using the encryption key, wherein the public data packets include an identity identifier and timestamp information of the breeding management information, and the private data packets include breeding quantity, animal health status and animal number information; the public data packets and the private data packets are distributed and stored in each breeding data block of the blockchain system; When acquiring the breeding management information of the data management module, the breeding management node extracts the public data packet and the private data packet from each breeding data block according to the corresponding timestamp information and the identity of the client, queries the decryption information according to the identity, extracts the corresponding decryption key to decrypt the private data packet, and restores the breeding management information based on the public data packet and the private data packet using a preset data template.

7. A drone-assisted intelligent breeding management system, characterized in that: include: A receiving module: used for receiving first image information collected when the first UAV is at a first flight altitude, and performing image preprocessing operations on the first image information; A center point determination module: used for inputting the pre-processed first image information into a pre-built animal recognition model for recognition to determine the image information of each animal in the first image information, and determining the center point position of each animal in the first image information according to the animal image information; A flight trajectory determination module: determines the density distribution image information of each animal in the first image information according to the center point position of each animal in the first image information and the first image information, and performs a clustering operation on the density distribution image information to determine a plurality of cluster center points in the density distribution image information, and determines the corresponding flight trajectory information according to the cluster center points; Performing signal disassembly on the flight trajectory information to determine the position information of each collection point in the flight trajectory information, and generating corresponding delivery information according to the position information of each collection point; The flight trajectory information and the delivery information are sent to the second drone so that the second drone delivers the medicine or feed at the corresponding collection point; Image acquisition module: used for sending the flight trajectory information to the first UAV to control the first UAV to acquire the second image information group and the infrared image information group along the set flight trajectory at the second flight altitude, and sending the acquired second image information group and infrared image information group to the backend server; wherein the second image information group includes a plurality of second image information, the infrared image information group includes a plurality of infrared image information, and the second image information corresponds to the infrared image information one by one; Performing feature recognition on each second image information in the second image information group according to a pre-built target object recognition model to determine corresponding target object information and target position information corresponding to the target object information; Matching the determined target object information with the constructed target object-animal number relationship table to determine the corresponding animal number information; The infrared image information is subjected to temperature analysis to determine the temperature information corresponding to the corresponding target object, and the animal number information is data-associated with the temperature information according to the target position information; if the temperature information corresponding to the target object does not match the set threshold range, the animal number information with the temperature abnormality will be sent to the breeding management personnel.

8. An electronic device, characterized in that: include: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the drone-assisted smart farming management method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program enables a computer to execute the drone-assisted smart farming management method according to any one of claims 1 to 6.

Images