Hive observation box, hive overwintering monitoring method and device

Abstract

The application discloses a bee colony observation box, a bee colony overwintering monitoring method and device, and the method comprises the following steps: receiving image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed in the bee colony observation box and collect bee gut images from different angles respectively; image data is processed by using a pre-trained bee colony intelligent monitoring model, and the bee colony area size and the bee colony tightness value are output; the bee colony intelligent monitoring model is trained by using image data transmitted by historical infrared video acquisition devices, and a YOLOv9 model is obtained; the bee colony intelligent monitoring model is used for measuring bee position information, determining the bee colony area size and the bee colony tightness value according to the bee position information; and whether the bee colony overwintering state is abnormal is determined according to the bee colony area size and the bee colony tightness value. The application can collect complete bee gut images and intelligently monitor the overwintering behavior of bees.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of beekeeping observation boxes, and in particular to a bee colony observation box, a bee colony overwintering monitoring method and device. BACKGROUND

[0002] This section is intended to provide background or context to the embodiments of the application recited in the claims. The description herein does not constitute admission that the prior art is prior art nor, that anything in this section is "prior art" with respect to the application.

[0003] The process of overwintering is crucial for the survival of a bee colony, especially in temperate and cold climates. Overwintering of domesticated bees is a key challenge in beekeeping. In the prior art, overwintering observation of domesticated bees is usually based on observation of bee status in a bee observation box, and a large amount of white sugar is supplemented to the bees and the heat preservation is strengthened to assist the overwintering of the bee colony.

[0004] However, the existing bee observation box cannot collect bee behavior data in the real living environment of bees. The traditional beekeeping hive is a "black box" for a long time. Bees reproduce, produce honey, overwinter, etc. in a dark environment, and frequent opening of the box affects the normal operation of the bees. It is difficult to know the state of the bees without opening the box. After opening the box, the environment of the whole beehive is destroyed, and the behavior of the bees will be affected to varying degrees. There is a scheme in the prior art that shoots the bee colony through the beehive, but this scheme destroys the living habits of the bee colony in the dark beehive conditions and cannot monitor the entire interior of the bee colony. The image collection is incomplete, and it is difficult to accurately determine whether the overwintering behavior of the bee colony is normal.

[0005] Currently, for observation and analysis of the overwintering behavior of the bee colony, the prior art mainly manually operates the bee observation box by artificial operation, and determines whether the overwintering of the bee colony is abnormal according to the experience of experts, and lacks reasonable and accurate algorithm determination. SUMMARY

[0006] The embodiments of the present application also provide a bee colony overwintering monitoring method for intelligently monitoring the overwintering behavior of bees based on bee behavior data in the real living environment of bees, and improving the analysis accuracy of the overwintering behavior of bees. The method comprises:

[0007] receiving image data transmitted by a plurality of infrared video collection devices; the plurality of infrared video collection devices are installed inside a bee observation box and collect images of the bee colony from different angles, respectively;

[0008] The pre-trained swarm intelligence monitoring model is used for processing the image data, and the area size of the bee cluster and the tightness value of the bee cluster are output; the tightness value of the bee cluster reflects the aggregation state of the bees; wherein, the swarm intelligence monitoring model is trained by using the image data transmitted by the historical infrared video acquisition device, and is obtained by training the YOLOv9 model; the swarm intelligence monitoring model is used for measuring the position information of the bees, and the area size of the bee cluster and the tightness value of the bee cluster are determined according to the position information of the bees.

[0009] According to the difference between the area size of the bee cluster and the preset area threshold of the bee cluster, and the difference between the tightness value of the bee cluster and the tightness threshold of the bee cluster, whether the overwintering state of the bee swarm is abnormal is determined.

[0010] The embodiment of the present application also provides a bee swarm overwintering monitoring device for intelligently monitoring the overwintering behavior of bees based on the behavior data of bees in the real living environment of bees, and improving the analysis accuracy of the overwintering behavior of bees, which comprises:

[0011] The data receiving module is used for receiving the image data transmitted by the plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed in the inside of the bee observation box and respectively collect bee cluster images from different angles;

[0012] The data intelligent processing module is used for processing the image data by using the pre-trained swarm intelligence monitoring model, and outputting the area size of the bee cluster and the tightness value of the bee cluster; the tightness value of the bee cluster reflects the aggregation state of the bees; wherein, the swarm intelligence monitoring model is trained by using the image data transmitted by the historical infrared video acquisition device, and is obtained by training the YOLOv9 model; the swarm intelligence monitoring model is used for measuring the position information of the bees, and the area size of the bee cluster and the tightness value of the bee cluster are determined according to the position information of the bees;

[0013] The result output module is used for determining whether the overwintering state of the bee swarm is abnormal according to the difference between the area size of the bee cluster and the preset area threshold of the bee cluster, and the difference between the tightness value of the bee cluster and the tightness threshold of the bee cluster.

[0014] The embodiment of the present application also provides a bee swarm observation box for collecting the behavior data of bees in the real living environment of bees, and collecting complete bee cluster images for bee swarm overwintering monitoring, which comprises: a wooden box body, a plurality of infrared video acquisition devices, and an intelligent host;

[0015] The bee cluster is located in the middle region inside the wooden box body, and the plurality of infrared video acquisition devices are respectively installed in the inside of the wooden box body, and the angle of view of each infrared video acquisition device is directed to the bee cluster; wherein, the inside of the wooden box body is divided into a middle bee breeding area and two side bee swarm observation areas along the long side direction of the box body by two transparent acrylic plates;

[0016] A plurality of infrared video acquisition devices are respectively connected to the intelligent host, the intelligent host is arranged on the outer wall of the wooden box body, and the intelligent host is used for receiving and caching image data transmitted by the infrared video acquisition device, transmitting the image data to a remote server, so that the remote server: based on the image data, using an artificial intelligence algorithm to monitor the overwintering behavior of the bee colony;

[0017] A ventilation opening is formed on one side of the wooden box body.

[0018] The embodiment of the present application also provides a computer device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the above-mentioned bee colony overwintering monitoring method when executing the computer program.

[0019] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-mentioned bee colony overwintering monitoring method.

[0020] The embodiment of the present application also provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the above-mentioned bee colony overwintering monitoring method.

[0021] In the embodiment of the present application, the image data transmitted by the plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices: installed inside the bee observation box, respectively collect bee nest images from different angles; the image data is processed by using a pre-trained bee colony intelligent monitoring model, and the bee cluster area size and the bee cluster tightness value are output; the bee cluster tightness value reflects the aggregation state of bees; wherein the bee colony intelligent monitoring model is obtained by training the YOLOv9 model using the image data transmitted by the historical infrared video acquisition device; the bee colony intelligent monitoring model is used to measure the position information of the bees, and the bee cluster area size and the bee cluster tightness value are determined according to the position information of the bees; according to the difference between the bee cluster area size and the preset bee cluster area threshold value, the difference between the bee cluster tightness value and the bee cluster tightness threshold value, whether the overwintering state of the bee colony is abnormal is determined. In the embodiment of the present application, a bee colony overwintering monitoring method is provided, which can receive the bee nest images collected from different angles inside the bee observation box, the received bee nest images do not need to open the beehive for processing, which conforms to the real living environment and habits of bees, and then the YOLOv9 model is used to calculate the bee cluster area size and the bee cluster tightness value of the bee colony, and the two data are used to quantitatively judge whether the overwintering state of the bee colony is abnormal, compared with the artificial judgment of the prior art, the intelligent monitoring of the overwintering behavior of the bees is realized, and the analysis accuracy of the overwintering behavior of the bees is improved.

[0022] In the embodiment of the present application, the bee colony observation box comprises: a wooden box body, a plurality of infrared video acquisition devices, and a smart host; the bee stomach is located in the middle region inside the wooden box body, and the plurality of infrared video acquisition devices are respectively installed inside the wooden box body, and the angle of view of each infrared video acquisition device is directed to the bee stomach; wherein, the inside of the wooden box body is divided into a middle bee breeding area and two side bee colony observation areas along the long side direction of the box body by two transparent acrylic plates; the plurality of infrared video acquisition devices are respectively connected to the smart host, the smart host is arranged on the outer wall of the wooden box body, and the smart host is used for receiving and caching image data transmitted by the infrared video acquisition devices, transmitting the image data to a remote server, so that the remote server: based on the image data, using an artificial intelligence algorithm to monitor the overwintering behavior of the bee colony; a ventilation opening is arranged on one side of the wooden box body. In the embodiment of the present application, the inside of the wooden box body is divided into a middle bee breeding area and two side bee colony observation areas along the long side direction of the box body by two transparent acrylic plates, the bee stomach is located in the middle bee breeding area inside the wooden box body, the plurality of infrared video acquisition devices are respectively installed inside the wooden box body, and the angle of view of each infrared video acquisition device is directed to the bee stomach, which can solve the problem that the existing bee observation box cannot shoot completely, and can collect behavior data and complete bee stomach images of bees in a real living environment in all directions and without dead angles even in a dark environment, for overwintering monitoring of the bee colony. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. In the drawings:

[0024] Figure 1 It is a structure schematic diagram of the bee colony observation box in the embodiment of the present application;

[0025] Figure 2 It is a specific example diagram of the bee colony observation box in the embodiment of the present application;

[0026] Figure 3 It is a flowchart of the bee colony overwintering monitoring method in the embodiment of the present application;

[0027] Figure 4 It is a specific example diagram of the bee colony overwintering monitoring method in the embodiment of the present application;

[0028] Figure 5 It is a structure schematic diagram of the bee colony overwintering monitoring device in the embodiment of the present application;

[0029] Figure 6 It is a schematic diagram of the computer device in the embodiment of the present application. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0031] Existing bee observation boxes cannot collect data on bee behavior in their actual living environment. Traditional beekeeping hives are essentially "black boxes," with bees breeding, producing honey, and overwintering in darkness. Frequent opening of the hive disrupts normal bee activities, while not opening it makes it difficult to know the bees' condition. Opening the hive disrupts the entire hive environment, affecting bee behavior to varying degrees. Even with cameras, existing bee observation boxes cannot monitor the entire bee colony. Current technology for observing and analyzing bee colony overwintering behavior mainly relies on manual operation of the observation box, with expert judgment based on overwintering abnormalities, lacking reasonable and accurate algorithms for judgment. Furthermore, monitoring bee colony overwintering behavior based on image data from existing observation boxes will affect the accuracy of bee behavior analysis, potentially leading to overwintering failure.

[0032] Therefore, this invention first proposes a bee colony observation box, which can collect real, dark-environment bee behavior data and complete bee comb images. This invention also proposes a bee colony overwintering monitoring method, which can intelligently monitor bee colony overwintering behavior based on real, dark-environment bee behavior data and complete bee comb images.

[0033] The bee colony observation box in the embodiments of the present invention will be introduced first below.

[0034] Figure 1 This is a schematic diagram of the bee colony observation box in an embodiment of the present invention, as shown below. Figure 1 As shown, the bee colony observation box in this embodiment of the invention includes: a wooden box, multiple infrared video acquisition devices, and an intelligent host;

[0035] The beehives are located in the middle of the wooden box. Multiple infrared video acquisition devices are installed inside the wooden box, with each infrared video acquisition device facing the beehives.

[0036] Multiple infrared video acquisition devices are connected to the intelligent host, which is set on the outer wall of the wooden box. The intelligent host is used to receive and buffer the image data transmitted by the infrared video acquisition devices and transmit the image data to the remote server, so that the remote server can monitor the overwintering behavior of bee colonies based on the image data and using artificial intelligence algorithms.

[0037] A ventilation opening is formed on one side of the wooden box.

[0038] The shape and structure of the wooden box can be customized, and the infrared video acquisition device can monitor the bee breeding area and the bee stomach area.

[0039] For example, the wooden box is in a cylindrical structure, and the wooden box can be divided into a middle bee breeding area and a surrounding bee group observation area by a transparent acrylic plate; or the wooden box is in a cuboid structure, and the wooden box can be divided into a middle bee breeding area and two side bee group observation areas by a transparent acrylic plate.

[0040] Meanwhile, feeding areas and various sensors for observing bees can be set based on the habits of the bee group and observation requirements.

[0041] In order to further improve the effectiveness of the obtained data, avoid the camera being damaged by the bees, or avoid the bee group activity being disturbed by the camera, Figure 2 A specific example of the bee group observation box in the embodiment of the present application is shown in Figure 2 The bee group observation box includes a wooden box, at least two infrared video acquisition devices 5, 6, and an intelligent host 8; the wooden box is composed of a box body 2, a box cover 1, and a box bottom 10;

[0042] The inside of the wooden box is divided into a middle bee breeding area and two side bee group observation areas along the long side direction of the box body by two transparent acrylic plates 3, 4; the two infrared video acquisition devices 5, 6 are respectively arranged in the two side bee group observation areas, and the viewing angle of each infrared video acquisition device 5, 6 is directed to the bee stomach of the middle bee breeding area.

[0043] The two infrared video acquisition devices 5, 6 are respectively connected to the intelligent host 8, the intelligent host 8 is arranged on the outer wall of the wooden box, and the intelligent host 8 is used for receiving and caching the image data transmitted by the infrared video acquisition devices 5, 6, and transmitting the image data to a remote server, so that the remote server: based on the image data, monitors the overwintering behavior of the bee group by using an artificial intelligence algorithm;

[0044] A ventilation opening 9 is formed on one side of the box in the bee breeding area.

[0045] As shown in Figure 2 It can be seen from the bee group observation box that the bee group observation box in the embodiment of the present application uses two transparent acrylic plates as a partition screen to construct an interference-free space that can observe the activities of bees without interfering with the natural behavior of bees, and sets infrared cameras in the two side bee group observation areas, which can solve the problem of incomplete shooting of the existing bee observation box. Even in a dark environment, the behavior data and complete bee stomach images of bees in a real living environment can be collected in all directions and without dead angles for overwintering monitoring of the bee group.

[0046] In order to improve the high definition of the image data collection of the bee colony, a high-transmittance and low-reflection film is attached to the transparent acrylic plate.

[0047] In one embodiment, an adjustable nest door 7 is arranged on one side of the wooden box of the bee breeding area.

[0048] In one embodiment, a feeding component is arranged on one side of the wooden box of the bee breeding area for feeding bees.

[0049] In one embodiment, a temperature and humidity sensor is arranged inside the wooden box, and the temperature and humidity sensor is connected to the intelligent host 8, and the intelligent host 8 is used to: receive the temperature and humidity data transmitted by the temperature and humidity sensor, and issue a warning of whether the temperature and humidity of the bee colony observation box is abnormal according to the temperature and humidity data.

[0050] Further, a temperature and humidity system can be arranged inside the bee colony observation box, which can adjust the temperature and humidity inside the bee colony observation box according to the temperature and humidity data and the warning of whether the temperature and humidity of the bee colony observation box is abnormal issued by the intelligent host 8.

[0051] In one embodiment, a weighing sensor 11 is arranged in a designated area of the wooden box, and the weighing sensor 11 is connected to the intelligent host 8, and the intelligent host 8 is used to receive, store and display the weight data transmitted by the weighing sensor. For example, the weight detection range of the weighing sensor is 0-60Kg±0.1%, and 0.1% represents the detection accuracy. Those skilled in the art can use appropriate sensors according to actual needs.

[0052] Further, the intelligent host is specifically used for: transmitting the image data, the temperature and humidity data, and the weight data to a remote server by one of wired, wireless, and satellite communication, so that the remote server: based on the image data, the temperature and humidity data, or the weight data, monitors the overwintering behavior of the bee colony by using an artificial intelligence algorithm.

[0053] In order to obtain more complete bee colony image data, the infrared video acquisition device is an infrared wide-angle camera. The spherical infrared camera ensures the accuracy of the image, avoids deformation and visual blind area. Through high-definition and distortion-free images, researchers can more accurately evaluate the state of the nest, the activity of the bees and other key indicators. Improve the accuracy and reliability of data analysis.

[0054] The bee colony observation box in the embodiment of the application adopts the structure of a three-box beehive, which ensures the integrity and flow of the bee activity area. The high-definition and low-noise camera is installed at both ends of the beehive, separated from the beehive by an acrylic plate, which not only realizes comprehensive capture of the bee activity, but also avoids hindering the flight path and activity area of the bees.

[0055] The bee colony observation box in the embodiment of the application integrates multiple advanced technologies, mainly composed of a specially-made wooden multi-box type beehive body, a temperature and humidity sensor, a weighing sensor, and two high-definition infrared cameras, etc. The multi-box partition structure is composed of a middle bee breeding area and two side bee colony observation areas. The specially-made multi-box partition structure can effectively solve the problems of poor observation effect inside the beehive and the influence on bee colony breeding. The specific video acquisition device installed in the two side bee colony observation areas can effectively solve the problems of incomplete observation image and unreasonable angle inside the beehive, and provide a solid foundation for the bee identification model in the hive. According to the dark environmental conditions inside the beehive and the wax secretion habit of bees, the internal structure area of the beehive is specifically segmented, and a bee colony observation window is constructed by two specially-made transparent acrylic plates with specific size and thickness. The bees in the hive can be observed through the observation window without disturbing the normal activities of the bees, and the actual beekeeping environment can be adapted. In addition, the temperature and humidity sensor is installed inside the equipment to collect real-time data of the temperature and humidity in the beehive, and the weighing sensor collects real-time data of the weight change of the beehive. The image, temperature and humidity, and weight information in the beehive can be collected to analyze the state change of the bee colony from multiple angles and comprehensively. The behavior change of the bees in the hive is recorded, and the environmental change of the beehive at the moment is also recorded.

[0056] The bee colony overwintering monitoring method in the embodiment of the application is introduced as follows, Figure 3 The flowchart of the bee colony overwintering monitoring method in the embodiment of the application is shown in the figure, and the method comprises the following steps:

[0057] Step 301, receiving image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed inside a bee observation box and collect beehive images from different angles respectively;

[0058] Step 302, processing the image data by using a pre-trained bee colony intelligent monitoring model to output a bee cluster area size and a bee cluster tightness value; the bee cluster tightness value reflects the aggregation state of the bees; wherein the bee colony intelligent monitoring model is trained by using image data transmitted by a historical infrared video acquisition device, and is obtained by training a YOLOv9 model; the bee colony intelligent monitoring model is used to measure the position information of the bees, and the bee cluster area size and the bee cluster tightness value are determined according to the position information of the bees;

[0059] Step 303, determining whether the overwintering state of the bee colony is abnormal according to the difference between the bee cluster area size and a preset bee cluster area threshold value, and the difference between the bee cluster tightness value and a bee cluster tightness threshold value.

[0060] From Figure 3As shown in the flow, in the embodiment of the application, image data transmitted by a plurality of infrared video acquisition devices is received; the plurality of infrared video acquisition devices are installed inside a bee colony observation box and collect images of bee stomachs from different angles respectively; a pre-trained bee colony intelligent monitoring model is used to process the image data, and the size of a bee cluster area and a bee cluster tightness value are output; the bee cluster tightness value reflects the aggregation state of bees; wherein the bee colony intelligent monitoring model is trained by using image data transmitted by historical infrared video acquisition devices, and is obtained by using a YOLOv9 model; the bee colony intelligent monitoring model is used to measure bee position information, determine the size of the bee cluster area and the bee cluster tightness value according to the bee position information; according to the difference between the size of the bee cluster area and a preset bee cluster area threshold value, the difference between the bee cluster tightness value and a bee cluster tightness threshold value, it is determined whether the overwintering state of the bee colony is abnormal. In the embodiment of the application, a bee colony overwintering monitoring method is proposed, which can receive images of bee stomachs collected from different angles inside a bee observation box, the received images of the bee stomachs do not need to be opened for processing, and conform to the real living environment and habits of bees, and then a YOLOv9 model is used to calculate the size of the bee cluster area and the bee cluster tightness value of the bee colony, and the two data are used to quantitatively determine whether the overwintering state of the bee colony is abnormal. Compared with the manual judgment of the prior art, the intelligent monitoring of the overwintering behavior of bees is realized, and the analysis accuracy of the overwintering behavior of bees is improved.

[0061] The bee colony overwintering monitoring method in the embodiment of the application can be implemented based on the remote server mentioned in the foregoing bee colony observation box. When implemented, a trained bee colony intelligent monitoring model is deployed on the server in advance, and intelligent judgment is performed based on received data to determine whether the overwintering state of the bee colony is abnormal.

[0062] In one embodiment, the bee colony intelligent monitoring model can be trained as follows:

[0063] The image data transmitted by the historical infrared video acquisition device is divided into a training set and a test set; wherein the image data in the training set is labeled by using an image labeling tool;

[0064] The YOLOv9 model is trained by using the training set, and the YOLOv9 model is tested by using the test set, and the bee colony intelligent monitoring model is obtained.

[0065] When implemented, the image data transmitted by the historical infrared video acquisition device can be divided into a training set, a test set and a validation set. The labelImg tool is used to label all ordinary bees in the training set as a training set, which is used to train the identification of the position information of bees in the nest.

[0066] In view of the small size of the individual bees, the large and dense number of bees in the honeycomb, the embodiment of the present application selects a YOLOv9 model for training, which has better detection effect on small targets, faster test speed, better multi-target recognition effect, and supports various data enhancement technologies such as greying and filtering, and can adapt to honeycomb image recognition. The YOLOv9 model is trained using the training set data, and the performance of the YOLOv9 model is verified using the validation set.

[0067] Figure 4 A specific example of the overwintering monitoring method of the bee colony in the embodiment of the present application is shown in the figure Figure 4 As shown, the intelligent monitoring model of the bee colony comprises a plurality of processing layer combinations, which are convolution layers and pooling layers. The convolution layers are used to extract local features of image data, and the pooling layers are used to reduce the dimension of image data while retaining local features. For example, for an input image with a resolution of 416x416 pixels and a color channel of x16, after the first processing layer combination, an image with a resolution of 208x208 pixels and a color channel of x32 is output, and then after the second, third, fourth and fifth processing layer combinations, an image with a resolution of 13x13 pixels and a color channel of x512 is output while retaining the effective local features of the image. Then, the bee image coordinate information is obtained, followed by bee clustering behavior recognition, calculation of the bee cluster area and the bee cluster tightness value, and finally identification of the overwintering condition of the bee colony. Figure 4 x16, x32, x64, x128, x256, x512 respectively represent the number of color channels after processing by different convolution layers and pooling layers.

[0068] In specific implementation, the input bee colony image data is extracted by the convolution layer to extract the local features of the bees, such as specific textures, colors or edges of the bees, so as to identify the area of the bees in the image. The pooling layer is used to retain the main features while reducing the dimension and calculation amount of the data, and improving the generalization ability of the model. The target detection model loss function of the model YOLOv9 is composed of two parts of classification loss and regression loss, and the formula composition is as follows:

[0069]

[0070] In the formula, AP is the average precision of pattern recognition, q is the label of the target bee detection, p is the distance between the center points of the frames, a is the loss function adjustment weight, g is the loss function adjustment parameter, CIoU loss is the intersection over union of the predicted frame and the true frame, IOU is the overlap degree of the predicted bounding box and the true bounding box, c is the diagonal length of the minimum enclosing rectangle between the frames, v is the width-height ratio similarity between the frames, DFL represents the bee bounding box accuracy estimation value, and S iThe representative bounding box prediction probability at position i, y is a general distribution value, and i is the serial number of its target box.

[0071] The trained model is used to evaluate the performance of the validation set. The evaluation index of the model uses the accuracy and recall rate of the model. The accuracy refers to the ability of the model to accurately identify bees, and the recall rate refers to the ability of the model to identify the number of all bees. The specific formula is as follows:

[0072]

[0073] TP refers to the number of targets that the model correctly predicts as positive samples, FP refers to the number of targets that the model incorrectly predicts as positive samples, and FN refers to the number of targets that the model incorrectly predicts as negative samples.

[0074] The swarm intelligence monitoring model in the embodiment of the application is mainly used to measure the position information of bees, and determines the swarm area size and swarm tightness value according to the position information of the bees.

[0075] In one embodiment, the swarm intelligence monitoring model can determine the swarm area size and swarm tightness value according to the position information of the bees in the following manner:

[0076] According to the position information of the bees, the sum of the maximum coordinate distance difference between all bees in the same coordinate system is calculated, and the product of the sum and the body length of the bees is taken as the swarm area size.

[0077] The minimum value of the distance between each bee and other bees is calculated, and the average value of the plurality of minimum values is taken as the swarm tightness value.

[0078] According to the swarm area size, the group size and abnormal situation of the bees during the overwintering period can be determined. The swarm refers to an oval spherical body formed by the bees gathering around the brood in winter. The video collection angle of the overwintering swarm can be perpendicular to the brood, and the projection area information of the swarm perpendicular to the brood can be completely collected, so that the swarm area size can be effectively measured. Generally, the area of a single brood single side is about 1035cm 2 (23cm x 45cm), and the overwintering swarm generally gathers in the central area of the brood. When the projection area of the overwintering swarm exceeds 2 / 3 of the area of the brood (about 690cm 2 ), it is determined that the swarm has a scattering abnormal situation, and when the projection area of the overwintering swarm is less than 1 / 4 of the area of the brood (about 259cm 2 ), it is determined that the overwintering swarm group is too small. When the shooting angle of the infrared video collection device is not perpendicular to the brood, the swarm area size can be estimated according to the projection area of the swarm and the shooting angle of the infrared video collection device.

[0079] The size of the aggregation tightness value can determine the abnormal situation such as overheat or overcooling of the bee colony during overwintering. Generally, the body width of a single Italian bee is about 0.5 cm, and the bees are closely attached to resist cold during overwintering. When the aggregation tightness value is greater than 1 cm, it is determined that the overwintering bee colony is overheated, and when the aggregation tightness value is less than 0.5 cm, it is determined that the overwintering bee colony is overcooled. The intelligent monitoring model of the bee colony can determine the occurrence of abnormal situations such as overheat, overcooling, and small colony and scattered colony of the overwintering bee colony, and remind the beekeeper to take reasonable temperature increasing or decreasing measures in time, so as to improve the overwintering success rate and avoid overwintering failure of the bee colony.

[0080] The specific calculation formula is as follows.

[0081] B = ((x1, y1), (x2, y2), …, (x i , y i ) (4)

[0082]

[0083] In the formula, B is the bee coordinate information data set, x i and x j represent the position information of the bees in the image, A represents the projection area value of the bee cluster, L represents the body length of the bee, d i represents the shortest distance between each bee and other bees, n represents the number of bees in the bee cluster, and D represents the aggregation tightness value.

[0084] In implementation, the image data transmitted by the plurality of infrared video acquisition devices can include:

[0085] The image data transmitted by the plurality of infrared video acquisition devices is received in one of wired, wireless, and satellite communication modes.

[0086] In one embodiment, after determining whether the overwintering state of the bee colony is abnormal, the method can further include:

[0087] receiving temperature and humidity data and weight data;

[0088] outputting a bee colony overwintering behavior report according to the temperature and humidity data, the weight data, and whether the overwintering state of the bee colony is abnormal; the bee colony overwintering behavior report includes the temperature and humidity and weight change trend of the bee colony observation box, and bee colony maintenance measure suggestion information, which includes increasing or decreasing the temperature and humidity maintaining measure and increasing or decreasing the feeding frequency.

[0089] In implementation, the bee colony overwintering behavior report can be directly sent to the terminal of the beekeeper, and the information in the report is displayed in the form of charts and graphs, and is obviously distinguished and displayed in color to improve user experience.

[0090] In summary, the embodiment of the present application designs a multi-box intelligent observation box in a real environment, uses two high-transmittance transparent acrylic plates as a partition screen, constructs an interference-free space that can observe the activities of bees without interfering with the natural behavior of bees, and a built-in infrared high-definition camera can capture subtle behavior moments. In order to solve the problem that the traditional camera nest spleen part corner is not complete, an infrared wide-angle camera is used, which can capture the behavior of bees in all directions without dead angles even in dark environments. In order to solve the problem of the reflection of the infrared camera on the acrylic plate in the dark environment, a high-transmittance reflective film is selected to be pasted on the acrylic plate. In order to solve the problem of imaging blur caused by the dynamic target object in the shooting scene and the short distance between the camera, a 2.5mm short focal length camera is used, which can maintain good visual coverage while ensuring the integrity and detail performance of the shooting content. At the same time, a special terminal and the corresponding data acquisition interface are designed independently, and the collected data are seamlessly connected through intelligent host, terminal, switch and other network equipment, and the data are accurately collected and efficiently transmitted. After the data are successfully transmitted to the remote server, the intelligent software processing system (including the bee colony intelligent monitoring model, and can also include data preprocessing software, database, etc.) is used to deeply analyze and process the environmental data. A multi-level real-time monitoring and behavior recognition system is constructed, high-resolution cameras and various sensors are comprehensively used to monitor the behavior of bees in the beehive in detail, and the comprehensive identification of bee clustering, dancing, interaction, cooperation and overwintering behavior is realized. Based on video images, a bee behavior prediction model is constructed, the behavior change of bees under different environmental conditions is predicted through the identification and analysis of internal activity images, and the beekeeper is helped to take measures in advance to optimize the breeding environment and management strategy. The beehive is equipped with an advanced temperature control device. The high-precision temperature and humidity sensor and the corresponding intelligent control system are integrated, which can monitor the temperature change in the beehive in real time, and automatically adjust the temperature and humidity based on the preset temperature and humidity range. Through the intelligent temperature and humidity management device, the temperature and humidity in the beehive remain stable, creating a suitable and comfortable environment for bees.

[0091] The embodiment of the present application has the following beneficial technical effects:

[0092] (1) Deep understanding of bee behavior: With the help of advanced technical means such as video monitoring and various sensors, various behavior patterns of bees such as clustering behavior are observed and understood in real scenes. It is helpful to reveal the organizational structure and behavior habit of bee society and provide deeper scientific basis for beekeeping industry.

[0093] (2) Promote scientific research: The multi-box intelligent observation box also provides valuable real-time data and observation platform for scientific research of bee behavior, ecology, etc., which is helpful to promote the development of related disciplines.

[0094] (3) Through careful observation of the behavior of bees and data collection, it is found that the regularity of bee behavior, such as slow action and loss of appetite, can improve the overwintering success rate and avoid the collapse of the bee colony.

[0095] The application also provides a bee colony overwintering monitoring device, as described in the following embodiment. Since the principle of solving the problem of the device is similar to that of the bee colony overwintering monitoring method, the implementation of the device can be referred to the implementation of the bee colony overwintering monitoring method, and the repeated parts will not be described again.

[0096] Figure 5 The structure diagram of the bee colony overwintering monitoring device in the embodiment of the application is shown in FIG. 1. Figure 5 The device comprises:

[0097] The data receiving module 501 is configured to receive image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed inside a bee colony observation box and collect images of bee colonies from different angles, respectively.

[0098] The data intelligent processing module 502 is configured to process the image data by using a pre-trained bee colony intelligent monitoring model, and output a bee cluster area size and a bee cluster tightness value; the bee cluster tightness value reflects the aggregation state of bees; wherein the bee colony intelligent monitoring model is trained by using image data transmitted by historical infrared video acquisition devices, and is obtained by using a YOLOv9 model; the bee colony intelligent monitoring model is used to measure bee position information, and determine the bee cluster area size and the bee cluster tightness value according to the bee position information.

[0099] The result output module 503 is configured to determine whether the overwintering state of the bee colony is abnormal according to a difference between the bee cluster area size and a preset bee cluster area threshold value, and a difference between the bee cluster tightness value and a bee cluster tightness threshold value.

[0100] In one embodiment, the data receiving module 501 is specifically configured to receive the image data transmitted by the plurality of infrared video acquisition devices in one of the following modes: wired, wireless, and satellite communication.

[0101] In one embodiment, the bee colony intelligent monitoring model comprises a plurality of processing layer combinations, and the processing layer combinations are convolution layers and pooling layers; the convolution layers are used to extract local features of the image data, and the pooling layers are used to reduce the dimension of the image data while retaining the local features.

[0102] In one embodiment, the bee colony intelligent monitoring model is trained in the following manner:

[0103] The image data transmitted by the historical infrared video acquisition devices is divided into a training set and a test set; wherein the image data in the training set is labeled by using an image labeling tool;

[0104] The YOLOv9 model is trained by using the training set, and the YOLOv9 model is tested by using the test set, so as to obtain the swarm intelligence monitoring model.

[0105] In one embodiment, the swarm intelligence monitoring model determines the bee cluster area size and the bee cluster tightness value according to the bee position information in the following manner:

[0106] According to the bee position information, the sum of the maximum coordinate distance difference between two bees in the same coordinate system is calculated, and the product of the sum and the body length of the bee is taken as the bee cluster area size.

[0107] The minimum value of the distance between each bee and other bees is calculated, and the average value of the plurality of minimum values is taken as the bee cluster tightness value.

[0108] In one embodiment, the method further comprises:

[0109] The report processing module is configured to receive the temperature and humidity data and the weight data after the result output module 503 determines whether the overwintering state of the bee colony is abnormal, and output a bee colony overwintering behavior report according to the temperature and humidity data, the weight data, and whether the overwintering state of the bee colony is abnormal; the bee colony overwintering behavior report includes the temperature and humidity and weight change trend of the bee colony observation box, and bee colony maintenance measure suggestion information, and the bee colony maintenance measure suggestion information includes increasing or decreasing the heat and humidity measures, and increasing or decreasing the feeding frequency.

[0110] Figure 6 A schematic diagram of a computer device in an embodiment of the present application is shown in Figure 6 The present application also provides a computer device 600, which includes a processor 601, a memory 602, and a computer program 603 stored in the memory 602 and executable on the processor 601, wherein the processor 601 executes the computer program 603 to implement the above-described bee colony overwintering monitoring method.

[0111] The present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-described bee colony overwintering monitoring method.

[0112] The present application also provides a computer program product, which includes a computer program, and the computer program is executed by a processor to implement the above-described bee colony overwintering monitoring method.

[0113] In the embodiment of the present application, the bee colony observation box comprises a wooden box body, a plurality of infrared video acquisition devices and a smart host. The bee gut is located in the middle region inside the wooden box body, and the plurality of infrared video acquisition devices are respectively installed inside the wooden box body, with the visual angle of each infrared video acquisition device facing the bee gut. The plurality of infrared video acquisition devices are respectively connected to the smart host, and the smart host is arranged on the outer wall of the wooden box body. The smart host is used to receive and cache image data transmitted by the infrared video acquisition device, and transmit the image data to a remote server, so that the remote server: based on the image data, uses an artificial intelligence algorithm to monitor the overwintering behavior of the bee colony; and a ventilation opening is formed on one side of the wooden box body. In the embodiment of the present application, the bee gut is located in the middle region inside the wooden box body, and the plurality of infrared video acquisition devices are respectively installed inside the wooden box body, with the visual angle of each infrared video acquisition device facing the bee gut. This can solve the problem of incomplete shooting of the existing bee observation box, and can collect behavior data and complete bee gut images of bees in a real living environment from all directions and without dead angles even in a dark environment, for overwintering monitoring of the bee colony.

[0114] In the embodiment of the present application, image data transmitted by a plurality of infrared video acquisition devices is received. The plurality of infrared video acquisition devices are installed inside a bee colony observation box and respectively collect bee gut images from different angles. A pre-trained bee colony intelligent monitoring model is used to process the image data, and the area size of a bee cluster and the tightness value of the bee cluster are output. The tightness value of the bee cluster reflects the aggregation state of the bees. The bee colony intelligent monitoring model is trained using image data transmitted by historical infrared video acquisition devices and a YOLOv9 model. The bee colony intelligent monitoring model is used to measure the position information of the bees, determine the area size of the bee cluster and the tightness value of the bee cluster according to the position information of the bees, and determine whether the overwintering state of the bee colony is abnormal according to the difference between the area size of the bee cluster and a preset area threshold value, and the difference between the tightness value of the bee cluster and a tightness threshold value. In the embodiment of the present application, a bee colony overwintering monitoring method is provided, which can receive bee gut images collected from different angles inside a bee observation box. The received bee gut images do not need to be opened for processing, conform to the real living environment and habits of bees, and then the YOLOv9 model is used to calculate the area size of the bee cluster and the tightness value of the bee cluster, and the two data are used to quantitatively determine whether the overwintering state of the bee colony is abnormal. Compared with the artificial judgment of the prior art, the intelligent monitoring of the overwintering behavior of the bees is realized, and the analysis accuracy of the overwintering behavior of the bees is improved.

[0115] Those skilled in the art will appreciate that embodiments of the present application can be readily used as software, hardware, or a combination of software and hardware. In a software embodiment, the methods can be tangibly embodied in a machine-readable storage medium having stored thereon instructions that can be used to program a computer to perform any of the methods. The software implementation can be initialized by loading and executing a set of instructions arranged to perform one of the methods into the computer's memory. Alternatively, hard-wired circuitry can be used in place of, or in combination with, software instructions. Thus, the

[0116] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in one or more of the flowchart illustrations and / or block diagrams. Figure 1 means for performing one or more functions specified in one or more of the flowchart illustrations and / or block diagrams.

[0117] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in one or more of the flowchart illustrations and / or block diagrams. Figure 1 means for performing one or more functions specified in one or more of the flowchart illustrations and / or block diagrams.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in one or more of the flowchart illustrations and / or block diagrams. Figure 1 means for performing one or more functions specified in one or more of the flowchart illustrations and / or block diagrams.

[0119] The specific embodiments described above have been disclosed by way of example and that, obviously, any modifications and / or alterations to the disclosed embodiment are conceivable to the skilled in the art falls within the scope of the present application.

Claims

1. A method of monitoring overwintering of a bee colony, characterized by, The method comprises: receiving image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed inside a bee colony observation box and respectively collect images of bee stomachs from different angles; processing the image data by using a pre-trained bee colony intelligent monitoring model to output a bee cluster area size and a bee cluster tightness value; the bee cluster tightness value reflects a honeybee aggregation state; wherein the bee colony intelligent monitoring model is obtained by training a YOLOv9 model by using historical image data transmitted by the infrared video acquisition devices; the bee colony intelligent monitoring model is used to measure honeybee position information and determine the bee cluster area size and the bee cluster tightness value according to the honeybee position information; determining whether the overwintering state of the bee colony is abnormal according to a difference between the bee cluster area size and a preset bee cluster area threshold value, and a difference between the bee cluster tightness value and a bee cluster tightness threshold value.

2. The method of claim 1, wherein, receiving image data transmitted by a plurality of infrared video acquisition devices, comprising: receiving image data transmitted by a plurality of infrared video acquisition devices in one of wired, wireless, and satellite communication modes.

3. The method of claim 1, wherein, The bee colony intelligent monitoring model comprises a plurality of processing layer combinations, and the processing layer combinations are convolution layers and pooling layers; the convolution layers are used to extract local features of image data, and the pooling layers are used to reduce the dimension of the image data while retaining the local features.

4. The method of claim 3, wherein, The bee colony intelligent monitoring model is obtained by training in the following manner: divide historical image data transmitted by the infrared video acquisition devices into a training set and a test set; wherein the image data in the training set is labeled by using an image labeling tool; train a YOLOv9 model by using the training set, test the YOLOv9 model by using the test set, and obtain the bee colony intelligent monitoring model.

5. The method of claim 1, wherein, The bee colony intelligent monitoring model determines the bee cluster area size and the bee cluster tightness value according to the honeybee position information in the following manner: calculate the sum of the maximum coordinate distance differences between two honeybees in the same coordinate system according to the honeybee position information, and take the product of the sum and the body length of the honeybee as the bee cluster area size; calculate the minimum value of the distance between each honeybee and other honeybees to obtain a plurality of minimum values, and take the average value of the plurality of minimum values as the bee cluster tightness value.

6. The method of claim 1, wherein, After determining whether the overwintering state of the bee colony is abnormal, the following steps are further included: receiving temperature and humidity data and weight data; outputting a bee colony overwintering behavior report according to the temperature and humidity data, the weight data, and whether the overwintering state of the bee colony is abnormal; the bee colony overwintering behavior report comprises temperature and humidity and weight change trends of the bee colony observation box and bee colony maintenance measure suggestion information, and the bee colony maintenance measure suggestion information comprises increasing or decreasing heat and humidity measures and increasing or decreasing feeding frequency.

7. A colony overwintering monitoring device, characterized in that, The device comprises: a data receiving module configured to receive image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed inside a bee colony observation box and respectively collect images of bee stomachs from different angles; The data intelligent processing module is configured to process the image data by using a pre-trained swarm intelligence monitoring model, and output a bee cluster area size and a bee cluster tightness value; the bee cluster tightness value reflects a honeybee aggregation state; the swarm intelligence monitoring model is obtained by training a YOLOv9 model by using image data transmitted by a historical infrared video acquisition device; the swarm intelligence monitoring model is configured to measure honeybee position information, and determine the bee cluster area size and the bee cluster tightness value according to the honeybee position information; The result output module is configured to determine whether the overwintering state of the bee swarm is abnormal according to a difference between the bee cluster area size and a preset bee cluster area threshold value, and a difference between the bee cluster tightness value and a bee cluster tightness threshold value.

8. A hive observation box, characterized in that The method comprises the following steps: The wooden box body, the plurality of infrared video acquisition devices, and the intelligent host; The bee gut is located in the middle region inside the wooden box body, and the plurality of infrared video acquisition devices are respectively installed inside the wooden box body, and the angle of view of each infrared video acquisition device is directed towards the bee gut; wherein the inside of the wooden box body is divided into a middle bee breeding area and two side bee swarm observation areas along the long side direction of the box body by two transparent acrylic plates; The plurality of infrared video acquisition devices are respectively connected to the intelligent host, the intelligent host is arranged on the outer wall of the wooden box body, and the intelligent host is configured to receive and cache image data transmitted by the infrared video acquisition devices, and transmit the image data to a remote server, so that the remote server: based on the image data, monitors the overwintering behavior of the bee swarm by using an artificial intelligence algorithm; The remote server implements the bee swarm overwintering monitoring method of claim 1; A ventilation opening is formed on one side of the wooden box body.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method of any one of claims 1 to 6.

10. A computer program product, characterised in that, The computer program product comprises a computer program, and the computer program is executed by the processor to implement the method of any one of claims 1 to 6.

Images