Bee colony observation box and bee colony overwintering monitoring method and device

Abstract

The invention discloses a bee colony observation box and a bee colony overwintering monitoring method and device. 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 arranged in the bee colony observation box and are used for respectively acquiring honeycomb images from different angles; processing the image data by using a pre-trained bee colony intelligent monitoring model, and outputting a bee colony area size and a bee colony closeness value; the bee colony intelligent monitoring model is obtained by training a YOLOv9 model by using image data transmitted by historical infrared video acquisition equipment; the bee colony intelligent monitoring model is used for measuring bee position information and determining the area size and the closeness value of a bee colony according to the bee position information; and determining whether the overwintering state of the bee colony is abnormal or not according to the bee colony area and the bee colony closeness value. According to the invention, complete honeycomb images can be acquired, and the overwintering behavior of bees can be intelligently monitored.

Description

Technical Field

[0001] The invention relates to the technical field of beehives for honeybee biology research, and in particular to a bee colony observation box, a bee colony overwintering monitoring method and a bee colony overwintering monitoring device. Background Art

[0002] This section is intended to provide a background or context to the embodiments of the invention recited in the claims. No admission is made that the description herein is prior art by inclusion in this section.

[0003] The process of bees overwintering is crucial to the survival of the bee colony, especially in temperate and cold climates. The overwintering of farmed bees is a key challenge in the beekeeping process. In the prior art, the observation of farmed bees overwintering is usually based on observing the status of the bees in a bee observation box, and according to the status of the bees, a large amount of white sugar is added to the bees, and insulation is strengthened to assist the bee colony to overwinter.

[0004] However, the existing bee observation boxes cannot collect data on bee behavior in the real living environment of bees. Traditional beehives are kept in a "black box" for a long time. Bees reproduce, produce honey, and hibernate in a dark environment. Frequent opening of the box to check affects the normal operation of the bees. It is difficult to know the status of the bees without opening the box for inspection. After opening the box, the entire beehive environment is destroyed, and the behavior of the bees will also be affected to varying degrees. There is a solution in the prior art to photograph the bee colony through the beehive, but this solution not only destroys the living habits of the bee colony under dark beehive conditions, but also fails to monitor the entire bee colony. The image collection is incomplete, and it is difficult to accurately judge whether the bee colony's wintering behavior is normal.

[0005] Currently, the existing technology for observing and analyzing the overwintering behavior of bee colonies mainly relies on manual operation of bee observation boxes, and judges whether the overwintering of the bee colony is abnormal based on expert experience, but lacks reasonable and accurate algorithm judgment. Summary of the invention

[0006] The embodiment of the present invention also provides a bee colony overwintering monitoring method, which is used to intelligently monitor the overwintering behavior of bees based on the bee behavior data in the real living environment of the bees, so as to improve the accuracy of the analysis of the overwintering behavior of the bees. The method comprises:

[0007] 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 to respectively acquire honeycomb images from different angles;

[0008] The image data is processed using a pre-trained bee colony intelligent monitoring model to output the bee cluster area size and bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; the bee colony intelligent monitoring model uses the image data transmitted by the historical infrared video acquisition device to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the bee cluster area size and bee cluster compactness value based on the bee position information;

[0009] Whether the wintering status of the bee colony is abnormal is determined based on the difference between the size of the bee cluster area and the preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold.

[0010] The embodiment of the present invention also provides a bee colony overwintering monitoring device, which is used to intelligently monitor the overwintering behavior of bees based on the bee behavior data in the real living environment of the bees, so as to improve the accuracy of the analysis of the overwintering behavior of the bees. The device includes:

[0011] A data receiving module is used to receive image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed inside the bee observation box to respectively acquire honeycomb images from different angles;

[0012] The data intelligent processing module is used to process the image data using the pre-trained bee colony intelligent monitoring model, and output the bee cluster area size and bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; the bee colony intelligent monitoring model uses the image data transmitted by the historical infrared video acquisition device to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the bee cluster area size and bee cluster compactness value according to the bee position information;

[0013] The result output module is used to determine whether the wintering state of the bee colony is abnormal based on the difference between the bee cluster area size and the preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold.

[0014] The embodiment of the present invention also provides a bee colony observation box, which is used to collect bee behavior data in the real living environment of bees and collect complete honeycomb images for bee colony overwintering monitoring. The bee colony observation box includes: a wooden box, multiple infrared video acquisition devices, and an intelligent host;

[0015] The beehive is located in the middle area of ​​the wooden box. Multiple infrared video acquisition devices are installed inside the wooden box. The viewing angle of each infrared video acquisition device is toward the beehive. The inside of the wooden box is divided into the middle bee breeding area and the bee colony observation area on both sides by two transparent acrylic panels along the long side of the box.

[0016] Multiple infrared video acquisition devices are respectively connected to an intelligent host, which is arranged on the outer wall of the wooden box. The intelligent host is used 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: uses an artificial intelligence algorithm to monitor the wintering behavior of the bee colony based on the image data;

[0017] A ventilation hole is provided on one side of the wooden box.

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

[0019] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned bee colony overwintering monitoring method is implemented.

[0020] An embodiment of the present invention further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the above-mentioned bee colony overwintering monitoring method is implemented.

[0021] In an embodiment of the present invention, image data transmitted by multiple infrared video acquisition devices; the multiple infrared video acquisition devices: are installed inside a bee observation box to collect bee comb images from different angles; the image data is processed using a pre-trained bee colony intelligent monitoring model to output the size of a bee cluster and the bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; wherein the bee colony intelligent monitoring model uses the image data transmitted by historical infrared video acquisition devices to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the size of a bee cluster area and the bee cluster compactness value based on the bee position information; based on the difference between the bee cluster area and a preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold, it is determined whether the bee colony overwintering state is abnormal. A bee colony overwintering monitoring method is proposed in an embodiment of the present invention. The method can receive honeycomb images collected from different angles and installed inside a bee observation box. The received honeycomb images do not require the beehive to be opened for processing, which conforms to the real living environment and habits of the bees. Then, the YOLOv9 model is used to calculate the bee cluster area size and bee cluster compactness value of the bee colony. These two data are used to quantitatively judge whether the overwintering state of the bee colony is abnormal. Compared with the manual judgment of the prior art, the method realizes the intelligent monitoring of the overwintering behavior of the bees and improves the accuracy of the analysis of the overwintering behavior of the bees.

[0022] In an embodiment of the present invention, a bee colony observation box includes: a wooden box, multiple infrared video acquisition devices, and an intelligent host; the bee comb is located in the middle area inside the wooden box, and multiple infrared video acquisition devices are respectively installed inside the wooden box, and the viewing angle of each infrared video acquisition device is toward the bee comb; wherein, the interior of the wooden box is divided into: a bee breeding area in the middle and bee colony observation areas on both sides by two transparent acrylic panels along the long side of the box; multiple infrared video acquisition devices are respectively connected to the intelligent host, and the intelligent host is arranged on the outer wall of the wooden box, and the intelligent 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: uses an artificial intelligence algorithm based on the image data to monitor the overwintering behavior of the bee colony; and opens a ventilation hole on one side of the wooden box. The bee colony observation box in the embodiment of the present invention has a wooden box body whose interior is divided into a bee breeding area in the middle and bee colony observation areas on both sides along the long side direction of the box body by two transparent acrylic panels. The bee spleen is located in the bee breeding area in the middle of the wooden box body. Multiple infrared video acquisition devices are respectively installed in the wooden box body. The viewing angle of each infrared video acquisition device is toward the bee spleen, which can solve the problem of incomplete shooting of the existing bee observation box. Even in a dark environment, the behavioral data of the bees in the real living environment and the complete bee spleen images can be collected in an all-round and no-dead-angle manner for the bee colony overwintering monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. 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 creative work. In the drawings:

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

[0025] Figure 2 A specific example diagram of a bee colony observation box in an embodiment of the present invention;

[0026] Figure 3 Schematic diagram of the process of monitoring bee colony overwintering in an embodiment of the present invention;

[0027] Figure 4 A diagram showing a specific example of a bee colony overwintering monitoring method according to an embodiment of the present invention;

[0028] Figure 5 Schematic diagram of the structure of a bee colony overwintering monitoring device in an embodiment of the present invention;

[0029] Figure 6 Schematic diagram of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Here, the exemplary 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 are unable to collect data on bee behavior in the real living environment of bees. Traditional beehives have long been in a "black box". Bees reproduce, produce honey, and hibernate in a dark environment. Frequent opening of the box to check affects the normal operation of the bees. It is difficult to know the status of the bees without opening the box for inspection. After opening the box, the entire beehive environment is destroyed, and the behavior of the bees will be affected to varying degrees. Even if a camera is installed, the existing bee observation box cannot monitor the entire bee colony. In the prior art, the observation and analysis of the bee colony's overwintering behavior is mainly carried out by manually operating the bee observation box, and judging whether the bee colony's overwintering is abnormal based on expert experience, lacking a reasonable and accurate algorithm judgment. In addition, if the bee colony's overwintering behavior is monitored based on the image data in the observation box in the prior art, it will affect the accuracy of the bee behavior analysis, and in severe cases, it will cause the bees to fail to overwinter.

[0032] To this end, an embodiment of the present invention first proposes a bee colony observation box, based on which real bee behavior data and complete bee comb images in a dark environment can be collected. An embodiment of the present invention then proposes a bee colony overwintering monitoring method, which can perform intelligent bee colony overwintering behavior monitoring based on real bee behavior data and complete bee comb images in a dark environment.

[0033] The following first introduces the bee colony observation box in an embodiment of the present invention.

[0034] Figure 1 Schematic diagram of the structure of the bee colony observation box in the embodiment of the present invention. Figure 1 As shown, the bee colony observation box in the embodiment of the present invention includes: a wooden box, a plurality of infrared video acquisition devices, and an intelligent host;

[0035] The beehive is located in the middle area of ​​the wooden box, and multiple infrared video acquisition devices are installed inside the wooden box, with the viewing angle of each infrared video acquisition device facing the beehive;

[0036] Multiple infrared video acquisition devices are respectively connected to an intelligent host, which is arranged on the outer wall of the wooden box. The intelligent host is used 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: uses an artificial intelligence algorithm to monitor the wintering behavior of the bee colony based on the image data;

[0037] A ventilation hole is provided on one side of the wooden box.

[0038] The shape and structure of the wooden box can be customized at will to ensure that the infrared video acquisition equipment can monitor the bee breeding area, beehives and other areas.

[0039] For example, the wooden box has a cylindrical structure, and the wooden box can be divided into a bee breeding area in the middle and a bee colony observation area around it by a transparent acrylic plate; or the wooden box has a rectangular structure, and the wooden box can be divided into a bee breeding area in the middle and bee colony observation areas on both sides by a transparent acrylic plate.

[0040] At the same time, feeding areas and multiple sensors can be set up to observe bees based on the habits of the bee colony and observation needs.

[0041] In order to further improve the effectiveness of the acquired data and avoid the camera being damaged by bees or the bee colony activities being disturbed by the camera, Figure 2 FIG. 1 is a specific example diagram of a bee colony observation box in an embodiment of the present invention. Figure 2 As shown, the bee colony observation box includes: a wooden box body, at least two infrared video acquisition devices 5, 6, and an intelligent host 8; the wooden box body is composed of a box body 2, a box cover 1 and a box bottom 10;

[0042] The interior of the wooden box is divided into a bee breeding area in the middle and bee colony observation areas on both sides by two transparent acrylic panels 3 and 4 along the long side of the box; two infrared video acquisition devices 5 and 6 are respectively arranged inside the bee colony observation areas on both sides, and the viewing angle of each infrared video acquisition device 5 and 6 is toward the honeycomb in the bee breeding area in the middle;

[0043] The two infrared video acquisition devices 5 and 6 are respectively connected to the intelligent host 8, which is arranged on the outer wall of the wooden box. The intelligent host 8 is used to receive and cache the image data transmitted by the infrared video acquisition devices 5 and 6, and transmit the image data to the remote server, so that the remote server: based on the image data, uses artificial intelligence algorithm to monitor the wintering behavior of the bee colony;

[0044] A vent 9 is provided on one side of the box in the bee breeding area.

[0045] from Figure 2 It can be seen from the bee colony observation box in the embodiment of the present invention that two transparent acrylic panels are used as partition screens to construct an interference-free space that allows observation of bee activities without interfering with the natural behavior of the bees. Infrared cameras are set in the bee colony observation areas on both sides to solve the problem of incomplete shooting in existing bee observation boxes. Even in a dark environment, the behavioral data of the bees in their real living environment and complete honeycomb images can be collected in all directions and without blind spots for wintering monitoring of the bee colony.

[0046] In order to improve the high definition of swarm image data collection, a highly transparent reflective film is attached to the transparent acrylic board.

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

[0048] In one embodiment, a feeding component is provided on one side of the wooden box in the bee breeding area for feeding the 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 smart host 8. The smart host 8 is used to: receive the temperature and humidity data transmitted by the temperature and humidity sensor, and issue a warning whether the temperature and humidity of the bee colony observation box are abnormal based on the temperature and humidity data.

[0050] Furthermore, a temperature and humidity system may be provided inside the bee colony observation box, which may adjust the temperature and humidity inside the bee colony observation box according to the temperature and humidity data and the warning issued by the intelligent host 8 as to whether the temperature and humidity of the bee colony observation box are abnormal.

[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 an intelligent host 8, and the intelligent host 8 is used to receive, store, and display 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% indicates the detection accuracy. Those skilled in the art can use a suitable sensor according to actual needs.

[0052] Furthermore, the intelligent host is specifically used to: transmit image data, temperature and humidity data, and weight data to a remote server via one of wired, wireless, and satellite communications, so that the remote server: uses an artificial intelligence algorithm to monitor the wintering behavior of the bee colony based on the image data, temperature and humidity data or weight data.

[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 and avoids deformation and visual blind spots. Through high-definition, distortion-free images, researchers can more accurately evaluate the state of the honeycomb, 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 present invention adopts a three-box beehive structure to ensure the integrity and flow of the bee activity area. High-definition and low-noise cameras are installed at both ends of the beehive and separated from the beehive by acrylic plates, which not only realizes the comprehensive capture of bee activities, but also avoids obstruction to the flight path and activity area of ​​the bees.

[0055] The bee colony observation box in the embodiment of the present invention integrates a number of advanced technologies, and is mainly composed of a special wooden multi-box beehive, a temperature and humidity sensor, a weighing sensor, and two high-definition infrared cameras. The multi-box partition structure consists of a bee breeding area in the middle and bee colony observation areas on both sides. The special multi-box partition structure can effectively solve the difficulties such as poor observation effect inside the hive and affecting bee colony breeding. By installing specific video acquisition devices in the bee colony observation areas on both sides, the problems of incomplete observation images inside the hive and unreasonable angles are effectively solved, providing a solid foundation for the recognition model of bees in the hive. In view of the dark environmental conditions inside the hive and the biological habits of bees such as wax secretion, the internal structure area of ​​the beehive is specifically divided, and a bee colony observation window is constructed through two transparent acrylic plates of special size and thickness. Observing the bees in the hive through the observation window can not disturb the normal activities of the bees, and can adapt to the actual beekeeping working environment. In addition, temperature and humidity sensors are installed inside the device to collect real-time data on temperature and humidity inside the beehive, and weighing sensors collect real-time changes in the weight of the beehive. The device can collect internal images, temperature, humidity and weight information of the beehive to conduct multi-angle and comprehensive analysis of changes in the status of the bee colony. While recording changes in the behavior of the bees in the hive, it also records changes in the current environment of the hive.

[0056] The following describes a method for monitoring bee colony wintering in an embodiment of the present invention. Figure 3 The present invention is a flowchart of a bee colony overwintering monitoring method according to an embodiment of the present invention, the method comprising:

[0057] Step 301, receiving image data transmitted by multiple infrared video acquisition devices; the multiple infrared video acquisition devices are installed inside the bee observation box to respectively capture honeycomb images from different angles;

[0058] Step 302: Process the image data using a pre-trained bee colony intelligent monitoring model, and output the bee cluster area size and bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; wherein the bee colony intelligent monitoring model uses the image data transmitted by the historical infrared video acquisition device to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the bee cluster area size and bee cluster compactness value according to the bee position information;

[0059] Step 303: Determine whether the wintering state of the bee colony is abnormal based on the difference between the bee cluster area size and the preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold.

[0060] from Figure 3It can be seen from the shown process that in the embodiment of the present invention, image data transmitted by multiple infrared video acquisition devices are received; the multiple infrared video acquisition devices are installed inside the bee colony observation box to collect bee spleen images from different angles; the image data is processed using a pre-trained bee colony intelligent monitoring model to output the bee cluster area size and the bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; wherein the bee colony intelligent monitoring model uses the image data transmitted by the historical infrared video acquisition device to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the bee cluster area size and the bee cluster compactness value according to the bee position information; according to the difference between the bee cluster area size and the preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold, it is determined whether the bee colony overwintering state is abnormal. A bee colony overwintering monitoring method is proposed in an embodiment of the present invention. The method can receive honeycomb images collected from different angles and installed inside a bee observation box. The received honeycomb images do not require the beehive to be opened for processing, which conforms to the real living environment and habits of the bees. Then, the YOLOv9 model is used to calculate the bee cluster area size and bee cluster compactness value of the bee colony. These two data are used to quantitatively judge whether the overwintering state of the bee colony is abnormal. Compared with the manual judgment of the prior art, the method realizes the intelligent monitoring of the overwintering behavior of the bees and improves the accuracy of the analysis of the overwintering behavior of the bees.

[0061] The bee colony overwintering monitoring method in the embodiment of the present invention can be implemented based on the remote server mentioned in the aforementioned bee colony observation box. During implementation, a trained bee colony intelligent monitoring model is deployed in advance on the end server, and intelligent judgment is made based on the received data to determine whether the bee colony overwintering status is abnormal.

[0062] In one embodiment, the swarm 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; the image data in the training set is annotated using an image annotation tool;

[0064] The YOLOv9 model is trained using the training set, and tested using the test set to obtain the swarm intelligent monitoring model.

[0065] During implementation, 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 common bees in the training set as a training set for training the location information recognition of bees in the nest.

[0066] In view of the fact that individual bees are small and the number of bees inside the hive is large and dense, the embodiment of the present invention selects the YOLOv9 model for training. Compared with other traditional deep learning algorithms, the YOLOv9 model has better detection effect for small targets, faster inspection speed, better multi-target recognition effect, and supports multiple data enhancement technologies such as graying and filtering, which can adapt to the recognition of bee images inside the hive. 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 FIG. 1 is a specific example of a method for monitoring bee colony wintering in an embodiment of the present invention, as shown in FIG. Figure 4 As shown, the bee colony intelligent monitoring model includes a combination of multiple processing layers, wherein the processing layer combination is a convolution layer and a pooling layer. The convolution layer is used to extract local features of image data, and the pooling layer is used to reduce the dimension of image data while retaining local features. For example, for an input image with a resolution of 416×416 pixels and a color channel of ×16, after the first processing layer combination, an image with a resolution of 208×208 pixels and a channel of ×32 is output, and after the second, third, fourth and fifth processing layer combinations, an image with a resolution of 13×13 pixels and a color channel of ×512 is output, while retaining the image with effective local features, and then the bee image coordinate information is obtained, followed by bee clustering behavior recognition, calculation of the bee cluster area and the bee cluster compactness value, and finally identification of the bee colony overwintering situation. Figure 4 In the figure, ×16, ×32, ×64, ×128, ×256, and ×512 represent the number of color channels after processing by different convolutional layers and pooling layers, respectively.

[0068] In the specific implementation, the bee colony image data is input, and the local features of the bees are extracted through the convolution layer, which are usually bee-specific textures, colors or edges, so as to identify the area of ​​the bees in the image; the role of the pooling layer is to reduce the dimension and amount of data while retaining the main features, thereby improving the generalization ability of the model. The target detection model loss function of the model YOLOv9 consists of two parts: classification loss and regression loss, and its formula 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 box, α is the loss function adjustment weight, γ is the loss function adjustment parameter, and CIoU loss is the intersection-over-union ratio of the predicted box and the true box, IOU is the overlap between the predicted bounding box and the true bounding box, c is the diagonal length of the minimum enclosing rectangle between the boxes, v is the aspect ratio similarity between the boxes, DFL represents the estimated accuracy of the bee bounding box, S iRepresents the predicted probability of the bounding box at position i, y is the general distribution value, and i is the ordinal number of its target box.

[0071] The trained model is evaluated using the validation set. The model's evaluation indicators use the model's precision and recall rate for evaluation. Precision refers to the model's ability to accurately identify bees, and recall refers to the model's ability to identify the number of all bees. The specific formula is as follows:

[0072]

[0073] TP refers to the number of targets correctly predicted as positive samples by the model, FP refers to the number of targets that are actually negative samples but incorrectly predicted as positive samples by the model, and FN refers to the number of targets that are actually negative samples but incorrectly predicted as positive samples by the correct model.

[0074] The bee colony intelligent monitoring model in the embodiment of the present invention is mainly used to measure the location information of the bees, and determine the area size and the compactness value of the bee cluster according to the bee location information.

[0075] In one embodiment, the bee colony intelligent monitoring model can determine the bee cluster area size and bee cluster compactness value according to the bee position information in the following manner:

[0076] According to the bee position information, the sum of the maximum coordinate distance differences between two bees in the same coordinate system is calculated, and the product of the sum and the bee body length is used as the size of the bee cluster area;

[0077] The minimum distance between each bee and other bees is calculated to obtain multiple minimum values, and the average of the multiple minimum values ​​is used as the bee cluster compactness value.

[0078] The size of the bee colony and any abnormalities during the winter can be determined based on the size of the bee colony. A bee colony refers to an elliptical sphere formed by bees gathering around the honeycomb in winter. The video acquisition angle of a bird's-eye view or an elevation view will cause deviations in the bee colony area information. The video acquisition angle of this bee colony observation box can be perpendicular to the honeycomb, and the projection area information of the bee colony perpendicular to the honeycomb can be fully collected, thereby effectively measuring the size of the bee colony area. Usually, the single-sided area of ​​a single honeycomb is about 1035cm 2 (23cm×45cm), the wintering bee clusters usually gather in the center of the honeycomb. When the projected area of ​​the wintering bee cluster exceeds 2 / 3 of the honeycomb area (about 690cm 2 ), it is determined that the bee colony has broken up abnormally. When the projected area of ​​the overwintering bee colony is less than 1 / 4 of the honeycomb area (about 259cm 2 ), it is determined that the overwintering bee colony is too small. During implementation, when the shooting angle of the infrared video acquisition device is not perpendicular to the honeycomb, the size of the bee colony can be estimated based on the projection area of ​​the bee colony and the shooting angle of the infrared video acquisition device.

[0079] The size of the bee cluster density value can be used to determine abnormal conditions such as overheating or overcooling during the winter. Usually, a single Italian honey bee is about 0.5 cm wide, and bees stay close together during the winter to resist the cold. When the cluster density value is greater than 1 cm, the winter bee colony is judged to be overheated, and when the cluster density value is less than 0.5 cm, the winter bee colony is judged to be overcooled. The bee colony intelligent monitoring model can be used to determine the occurrence of abnormal conditions such as overheating, overcooling, too small and scattered bee colonies during the winter, and remind beekeepers to take reasonable temperature increase and reduction measures in time, thereby improving the success rate of wintering and avoiding the failure of wintering bee colonies.

[0080] The specific calculation formula is as follows.

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

[0082]

[0083] Where B is the bee coordinate information dataset, x i and x j They represent the horizontal and vertical position information of the bees in the image, A represents the projection area of ​​the bee cluster, L represents the length of the bee body, and d i represents the shortest distance between each bee and other bees, n represents the number of bees in the cluster, and D represents the cluster compactness value.

[0084] During implementation, receiving image data transmitted by multiple infrared video acquisition devices may include:

[0085] Image data transmitted by multiple infrared video acquisition devices are received through one of wired, wireless, and satellite communications.

[0086] In one embodiment, after determining whether the bee colony wintering state is abnormal, the following steps may also be performed:

[0087] Receive temperature, humidity and weight data;

[0088] Based on the temperature and humidity data, weight data and whether the overwintering status of the bee colony is abnormal, a bee colony overwintering behavior report is output; the bee colony overwintering behavior report includes the temperature, humidity and weight change trends of the bee colony observation box, and recommended information on bee colony maintenance measures. The recommended information on bee colony maintenance measures includes: increasing or decreasing insulation and humidity measures, and increasing or decreasing feeding frequency.

[0089] During implementation, the bee colony overwintering behavior report can be sent directly to the beekeeper's terminal. The report displays information in the form of charts, graphs, etc., and is clearly distinguished by color to improve the user experience.

[0090] In summary, the embodiment of the present invention designs a multi-box intelligent observation box in a real environment, using two highly transparent transparent acrylic plates as a partition screen to construct an interference-free space that allows observation of bee activities without interfering with the natural behavior of bees. The built-in infrared high-definition camera can capture subtle behavioral moments. In order to solve the problem that the traditional camera cannot shoot all corners of the honeycomb, an infrared wide-angle camera is used, which can capture the behavior of bees in all directions and without dead angles even in a dark environment. In order to solve the problem of reflection of the acrylic plate when the infrared camera illuminates it in a dark environment, a high-transmittance reflective film is selected to be attached to the acrylic plate. In order to solve the problem that the target object in the shooting scene is dynamic and the distance between it and the camera is close, resulting in blurred imaging, a 2.5mm short-focal-length camera is used to ensure the integrity and detail expression of the shooting content while maintaining good visual coverage. At the same time, a dedicated terminal and a corresponding data acquisition interface are independently designed, and the collected data are seamlessly connected through network devices such as intelligent hosts, terminals, switches, etc., so that the data can be accurately collected and efficiently transmitted. After the data is successfully transmitted to the remote server, the environmental data is deeply analyzed and processed using an intelligent software processing system (including a bee colony intelligent monitoring model, and may also include data preprocessing software, databases, etc.). A multi-level real-time monitoring and behavior recognition system is constructed, and high-resolution cameras and multiple sensors are used to monitor the behavior of bees inside the beehive in detail, so as to achieve comprehensive recognition of bee clustering, dancing, interaction, collaboration, and overwintering behaviors. A bee behavior prediction model is constructed based on video images. Through the recognition and analysis of internal activity images, the behavior changes of bees under different environmental conditions are predicted, helping beekeepers to take measures in advance to optimize the breeding environment and management strategies. The beehive is equipped with advanced temperature control devices. It integrates high-precision temperature and humidity sensors and corresponding intelligent control systems, which can monitor the temperature changes in the beehive in real time and make accurate and automatic adjustments 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 an environment that is both suitable for survival and comfortable for the bees. Whether it is a cold winter or a hot summer, the bees can ensure the quality and yield of honey under constant temperature conditions.

[0091] The embodiments of the present invention have the following beneficial technical effects:

[0092] (1) In-depth understanding of bee behavior: With the help of advanced technical means, such as video surveillance and various sensors, various behavioral patterns of bees, such as aggregation behavior, can be observed and understood in real scenarios. This helps to reveal the organizational structure and behavioral habits of the bee society and provide a deeper scientific basis for the beekeeping industry.

[0093] (2) Promote scientific research: Multi-chamber intelligent observation boxes also provide valuable real-time data and observation platforms for scientific research such as bee behavior and ecology, helping to promote the development of related disciplines.

[0094] (3) Through careful observation of bee behavior and data collection, we discovered behavioral patterns of bees, such as slow movement and loss of appetite, which can improve the success rate of overwintering and avoid the collapse and death of bee colonies.

[0095] The present invention also provides a bee colony wintering monitoring device, as described in the following embodiments. Since the principle of the device to solve the problem is similar to that of the bee colony wintering monitoring method, the implementation of the device can refer to the implementation of the bee colony wintering monitoring method, and the repeated parts will not be repeated.

[0096] Figure 5 FIG. 1 is a schematic diagram of the structure of a bee colony overwintering monitoring device according to an embodiment of the present invention. Figure 5 As shown, the device comprises:

[0097] The data receiving module 501 is used to receive image data transmitted by multiple infrared video acquisition devices; the multiple infrared video acquisition devices are installed inside the bee colony observation box to collect honeycomb images from different angles;

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

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

[0100] In one embodiment, the data receiving module 501 is specifically used to receive image data transmitted by multiple infrared video acquisition devices through one of wired, wireless, and satellite communications.

[0101] In one embodiment, the swarm intelligent monitoring model includes a combination of multiple processing layers, wherein the processing layer combination is a convolution layer and a pooling layer. The convolution layer is used to extract local features of image data, and the pooling layer is used to reduce the dimension of image data while retaining the local features.

[0102] In one embodiment, the bee colony intelligent monitoring model is trained as follows:

[0103] The image data transmitted by the historical infrared video acquisition device is divided into a training set and a test set; the image data in the training set is annotated using an image annotation tool;

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

[0105] In one embodiment, the bee colony intelligent monitoring model determines the bee cluster area size and bee cluster compactness 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 differences between two bees in the same coordinate system is calculated, and the product of the sum and the bee body length is used as the size of the bee cluster area;

[0107] The minimum distance between each bee and other bees is calculated to obtain multiple minimum values, and the average of the multiple minimum values ​​is used as the bee cluster compactness value.

[0108] In one embodiment, it further includes:

[0109] The report processing module is used to receive temperature and humidity data and weight data after the result output module 503 determines whether the overwintering status of the bee colony is abnormal; output a bee colony overwintering behavior report based on the temperature and humidity data, weight data and whether the overwintering status of the bee colony is abnormal; the bee colony overwintering behavior report includes the temperature, humidity and weight change trends of the bee colony observation box, and the bee colony maintenance measures recommended information, and the bee colony maintenance measures recommended information includes: increasing or decreasing insulation and humidity measures, increasing or decreasing feeding frequency.

[0110] Figure 6 Schematic diagram of a computer device in an embodiment of the present invention. Figure 6 As shown, an embodiment of the present invention further provides a computer device 600, including 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 implements the above-mentioned bee colony overwintering monitoring method when executing the computer program 603.

[0111] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned bee colony overwintering monitoring method is implemented.

[0112] An embodiment of the present invention further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the above-mentioned bee colony overwintering monitoring method is implemented.

[0113] In the embodiment of the present invention, the bee colony observation box includes: a wooden box, a plurality of infrared video acquisition devices, and an intelligent host; the bee spleen is located in the middle area inside the wooden box, and a plurality of infrared video acquisition devices are respectively installed inside the wooden box, and the viewing angle of each infrared video acquisition device is toward the bee spleen; the plurality of infrared video acquisition devices are respectively connected to the intelligent host, and the intelligent host is arranged on the outer wall of the wooden box, and the intelligent host is used to receive and cache the image data transmitted by the infrared video acquisition device, and transmit the image data to the remote server, so that the remote server: based on the image data, uses an artificial intelligence algorithm to monitor the wintering behavior of the bee colony; and opens a ventilation hole on one side of the wooden box. In the bee colony observation box in the embodiment of the present invention, the bee spleen is located in the middle area inside the wooden box, and a plurality of infrared video acquisition devices are respectively installed inside the wooden box, and the viewing angle of each infrared video acquisition device is toward the bee spleen, which can solve the problem of incomplete shooting of the existing bee observation box, and can collect the behavior data and complete bee spleen images in the real living environment of the bees in all directions and without dead angles even in a dark environment, so as to provide the bee colony wintering monitoring.

[0114] In an embodiment of the present invention, image data transmitted by multiple infrared video acquisition devices are received; the multiple infrared video acquisition devices are installed inside a bee colony observation box to collect honeycomb images from different angles; the image data is processed using a pre-trained bee colony intelligent monitoring model to output the size of a bee cluster and a bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; wherein the bee colony intelligent monitoring model uses the image data transmitted by historical infrared video acquisition devices to train a YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the size of a bee cluster area and the bee cluster compactness value based on the bee position information; and determine whether the overwintering state of the bee colony is abnormal based on the difference between the bee cluster area and a preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold. A bee colony overwintering monitoring method is proposed in an embodiment of the present invention. The method can receive honeycomb images collected from different angles and installed inside a bee observation box. The received honeycomb images do not require the beehive to be opened for processing, which conforms to the real living environment and habits of the bees. Then, the YOLOv9 model is used to calculate the bee cluster area size and bee cluster compactness value of the bee colony. These two data are used to quantitatively judge whether the overwintering state of the bee colony is abnormal. Compared with the manual judgment of the prior art, the method realizes the intelligent monitoring of the overwintering behavior of the bees and improves the accuracy of the analysis of the overwintering behavior of the bees.

[0115] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0116] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0117] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0119] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A bee colony overwintering monitoring method, characterized in that: The method includes: 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 to respectively acquire honeycomb images from different angles; The image data is processed using a pre-trained bee colony intelligent monitoring model to output the bee cluster area size and bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; the bee colony intelligent monitoring model uses the image data transmitted by the historical infrared video acquisition device to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the bee cluster area size and bee cluster compactness value based on the bee position information; Whether the wintering status of the bee colony is abnormal is determined based on the difference between the size of the bee cluster area and the preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold.

2. The method according to claim 1, characterized in that Receive image data transmitted by multiple infrared video acquisition devices, including: Image data transmitted by multiple infrared video acquisition devices are received through one of wired, wireless, and satellite communications.

3. The method according to claim 1, characterized in that The bee colony intelligent monitoring model includes a combination of multiple processing layers, which are a convolution layer and a pooling layer. The convolution layer is used to extract local features of image data, and the pooling layer is used to reduce the dimension of image data while retaining the local features.

4. The method according to claim 3, characterized in that The bee colony intelligent monitoring model is trained as follows: The image data transmitted by the historical infrared video acquisition device is divided into a training set and a test set; the image data in the training set is annotated using an image annotation tool; The YOLOv9 model is trained using the training set, and tested using the test set to obtain the swarm intelligent monitoring model.

5. The method according to claim 1, characterized in that The bee colony intelligent monitoring model determines the bee cluster area size and bee cluster compactness value according to the bee location information in the following manner: According to the bee position information, the sum of the maximum coordinate distance differences between two bees in the same coordinate system is calculated, and the product of the sum and the bee body length is used as the size of the bee cluster area; The minimum distance between each bee and other bees is calculated to obtain multiple minimum values, and the average of the multiple minimum values ​​is used as the bee cluster compactness value.

6. The method according to claim 1, characterized in that After determining whether the bee colony's wintering status is abnormal, it also includes: Receive temperature, humidity and weight data; Based on the temperature and humidity data, weight data and whether the overwintering status of the bee colony is abnormal, a bee colony overwintering behavior report is output; the bee colony overwintering behavior report includes the temperature, humidity and weight change trends of the bee colony observation box, and recommended information on bee colony maintenance measures. The recommended information on bee colony maintenance measures includes: increasing or decreasing insulation and humidity measures, and increasing or decreasing feeding frequency.

7. A bee colony wintering monitoring device, characterized in that: The device includes: A data receiving module is used to receive image data transmitted by a plurality of infrared video acquisition devices; the plurality of infrared video acquisition devices are installed inside the bee colony observation box to respectively acquire honeycomb images from different angles; The data intelligent processing module is used to process the image data using the pre-trained bee colony intelligent monitoring model, and output the bee cluster area size and bee cluster compactness value; the bee cluster compactness value reflects the aggregation state of the bees; the bee colony intelligent monitoring model uses the image data transmitted by the historical infrared video acquisition device to train the YOLOv9 model; the bee colony intelligent monitoring model is used to measure the bee position information, and determine the bee cluster area size and bee cluster compactness value according to the bee position information; The result output module is used to determine whether the wintering state of the bee colony is abnormal based on the difference between the bee cluster area size and the preset bee cluster area threshold, and the difference between the bee cluster compactness value and the bee cluster compactness threshold.

8. A bee colony observation box, characterized in that: include: Wooden box, multiple infrared video acquisition devices, smart host; The beehive is located in the middle area of ​​the wooden box. Multiple infrared video acquisition devices are installed inside the wooden box. The viewing angle of each infrared video acquisition device is toward the beehive. The inside of the wooden box is divided into the middle bee breeding area and the bee colony observation area on both sides by two transparent acrylic panels along the long side of the box. Multiple infrared video acquisition devices are respectively connected to an intelligent host, which is arranged on the outer wall of the wooden box. The intelligent host is used 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: uses an artificial intelligence algorithm to monitor the wintering behavior of the bee colony based on the image data; A ventilation hole is provided on one side of the wooden box.

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

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

Images