Bird recognition method and device based on bird posture and image quality

Abstract

The embodiment of the invention discloses a bird recognition method and device based on bird postures and image quality. A specific embodiment of the method comprises the following steps: in response to detecting that bird activities exist in a feeding area, collecting a real-time monitoring video corresponding to the feeding area; performing video sampling on the real-time monitoring video; performing the following processing steps on the to-be-detected image: performing bird posture recognition on the to-be-detected image; determining an image quality result corresponding to the to-be-detected image; bird identification is carried out on the to-be-detected image; performing recognition result optimization on the bird recognition result according to the bird posture recognition result and the image quality result; and generating a global bird recognition result for the real-time monitoring video according to the obtained optimized bird recognition result sequence. According to the embodiment, the bird recognition accuracy is greatly improved.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and more particularly to a method and device for bird recognition based on bird posture and image quality. Background Art

[0002] With the rapid development of artificial intelligence technology, artificial intelligence technology has enabled the fields of biometrics, species protection, and especially animal behavior analysis and protection. At present, it is difficult to accurately identify birds, especially in bird identification, due to the interference of bird posture and image quality degradation.

[0003] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the invention

[0004] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0005] Some embodiments of the present disclosure propose a bird recognition method and device based on bird posture and image quality to solve one or more of the technical problems mentioned in the above background technology section.

[0006] In a first aspect, some embodiments of the present disclosure provide a bird recognition method based on bird posture and image quality, the method comprising: in response to detecting the presence of bird activities in a feeding area, collecting real-time monitoring video corresponding to the above-mentioned feeding area; performing video sampling on the above-mentioned real-time monitoring video to obtain a sequence of images to be detected; for each image to be detected in the above-mentioned sequence of images to be detected, performing the following processing steps: performing bird posture recognition on the above-mentioned image to be detected to generate a bird posture recognition result; determining the image quality result corresponding to the above-mentioned image to be detected; performing bird recognition on the above-mentioned image to be detected to determine the bird recognition result; based on the above-mentioned bird posture recognition result and the above-mentioned image quality result, optimizing the above-mentioned bird recognition result to obtain an optimized bird recognition result; based on the obtained optimized bird recognition result sequence, generating a global bird recognition result for the above-mentioned real-time monitoring video.

[0007] Optionally, in response to detecting the presence of bird activity in the feeding area and before collecting the real-time monitoring video corresponding to the feeding area, the method further includes: detecting heat source changes and / or pressure changes in the feeding area through infrared sensors and / or pressure sensors to determine whether there is bird activity in the feeding area.

[0008] Optionally, the above-mentioned video sampling of the real-time monitoring video to obtain the image sequence to be detected includes: performing video downsampling on the real-time monitoring video at preset time intervals to obtain the image sequence to be detected.

[0009] Optionally, the video sampling of the real-time monitoring video to obtain the image sequence to be detected includes: performing video downsampling on the real-time monitoring video at preset time intervals to obtain a candidate image sequence; for each candidate image in the candidate image sequence, performing the following screening steps: performing image binarization processing on the candidate image to obtain a binarized image; performing dilation processing on the binarized image to obtain a dilated image; determining the image difference between the dilated image and the background image to obtain at least one difference set area, wherein the background image is an image when there is no obstruction in the feeding area; in response to the presence of a difference set area whose corresponding area area is greater than a preset area area in at least one of the difference set areas, determining the candidate image as the image to be detected.

[0010] Optionally, the above-mentioned bird posture recognition is performed on the above-mentioned image to be detected to generate a bird posture recognition result, including: determining the bird posture probability distribution corresponding to the above-mentioned image to be detected through a pre-trained bird posture recognition model, wherein the above-mentioned bird posture probability distribution represents the confidence level of different bird postures corresponding to the birds contained in the image to be detected; and determining the above-mentioned bird posture recognition result based on the above-mentioned bird posture distribution.

[0011] Optionally, the determining of the image quality result corresponding to the image to be detected includes: performing image quality scoring on the image to be detected by using a pre-trained image quality scoring model to obtain the image quality result.

[0012] Optionally, the performing bird recognition on the image to be detected to determine the bird recognition result includes: performing bird recognition on the image to be detected by using a pre-trained fine-grained bird recognition model to determine the bird recognition result.

[0013] Optionally, the above-mentioned bird recognition results are optimized based on the above-mentioned bird posture recognition results and the above-mentioned image quality results to obtain the optimized bird recognition results, including: taking the above-mentioned bird posture recognition results and the above-mentioned image quality results as optimization weights, performing confidence-weighted fusion on the above-mentioned bird recognition results to obtain the above-mentioned optimized bird recognition results.

[0014] Optionally, the above-mentioned bird posture recognition result and the above-mentioned image quality result are used as optimization weights, and the above-mentioned bird recognition results are confidence-weighted fused to obtain the above-mentioned optimized bird recognition result, including: taking the above-mentioned bird posture recognition result and the above-mentioned image quality result as optimization weights, and performing confidence-weighted fusion on the above-mentioned bird recognition results to obtain the weighted fused bird recognition result; through a balance factor, the above-mentioned bird recognition result and the above-mentioned weighted fused bird recognition result are distributed balanced to obtain the above-mentioned optimized bird recognition result.

[0015] Optionally, the above-mentioned bird posture recognition result and the above-mentioned image quality result are used as optimization weights, and the above-mentioned bird recognition results are confidence-weighted fused to obtain the above-mentioned optimized bird recognition result, including: taking the above-mentioned bird posture recognition result and the above-mentioned image quality result as optimization weights, and performing confidence-weighted fusion on the above-mentioned bird recognition results to obtain the weighted fused bird recognition result; and normalizing the above-mentioned weighted fused bird recognition result through an activation layer to obtain the above-mentioned optimized bird recognition result.

[0016] In a second aspect, some embodiments of the present disclosure provide a bird identification device based on bird posture and image quality, the device comprising: an acquisition unit, configured to, in response to detecting the presence of bird activities in a feeding area, collect real-time monitoring video corresponding to the above-mentioned feeding area; a video sampling unit, configured to perform video sampling on the above-mentioned real-time monitoring video to obtain a sequence of images to be detected; an execution unit, configured to perform the following processing steps for each image to be detected in the above-mentioned sequence of images to be detected: perform bird posture recognition on the above-mentioned image to be detected to generate a bird posture recognition result; determine the image quality result corresponding to the above-mentioned image to be detected; perform bird recognition on the above-mentioned image to be detected to determine the bird recognition result; optimize the above-mentioned bird recognition result according to the above-mentioned bird posture recognition result and the above-mentioned image quality result to obtain an optimized bird recognition result; a generation unit, configured to generate a global bird recognition result for the above-mentioned real-time monitoring video according to the obtained optimized bird recognition result sequence.

[0017] Optionally, the bird identification device based on bird posture and image quality also includes: a detection unit, which is configured to, before collecting the real-time monitoring video corresponding to the above-mentioned feeding area in response to detecting the presence of bird activity in the feeding area, the above-mentioned method also includes: detecting heat source changes and / or pressure changes in the feeding area through infrared sensors and / or pressure sensors to determine whether there is bird activity in the feeding area.

[0018] Optionally, the video sampling unit is further configured to: perform video downsampling on the real-time monitoring video at preset time intervals to obtain a sequence of images to be detected.

[0019] Optionally, the video sampling unit is further configured to: perform video downsampling on the real-time monitoring video at preset time intervals to obtain a candidate image sequence; for each candidate image in the candidate image sequence, perform the following screening steps: perform image binarization on the candidate image to obtain a binarized image; perform dilation on the binarized image to obtain a dilated image; determine the image difference between the dilated image and the background image to obtain at least one difference set area, wherein the background image is an image when there is no obstruction in the feeding area; in response to the presence of a difference set area whose corresponding area area is greater than a preset area area in at least one of the difference set areas, determine the candidate image as an image to be detected.

[0020] Optionally, the execution unit is further configured to: determine the bird posture probability distribution corresponding to the image to be detected through a pre-trained bird posture recognition model, wherein the bird posture probability distribution represents the confidence level of different bird postures corresponding to the birds contained in the image to be detected; and determine the bird posture recognition result based on the bird posture distribution.

[0021] Optionally, the execution unit is further configured to: perform image quality scoring on the image to be detected using a pre-trained image quality scoring model to obtain the image quality result.

[0022] Optionally, the execution unit is further configured to: perform bird recognition on the image to be detected by using a pre-trained fine-grained bird recognition model to determine the bird recognition result.

[0023] Optionally, the execution unit is further configured to: use the bird posture recognition result and the image quality result as optimization weights, perform confidence-weighted fusion on the bird recognition result, and obtain the optimized bird recognition result.

[0024] Optionally, the execution unit is further configured to: perform confidence-weighted fusion on the bird recognition results using the bird posture recognition results and the image quality results as optimization weights to obtain the weighted fused bird recognition results; and perform distribution balancing on the bird recognition results and the weighted fused bird recognition results using a balancing factor to obtain the optimized bird recognition results.

[0025] Optionally, the execution unit is further configured to: perform confidence-weighted fusion on the bird recognition results using the bird posture recognition results and the image quality results as optimization weights to obtain weighted fused bird recognition results; and perform normalization on the weighted fused bird recognition results through an activation layer to obtain the optimized bird recognition results.

[0026] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the above-mentioned first aspect.

[0027] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner of the above-mentioned first aspect is implemented.

[0028] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the bird recognition method based on bird posture and image quality of some embodiments of the present disclosure, the bird recognition accuracy is improved. Specifically, the reason for the low bird recognition accuracy is that it is difficult to accurately recognize birds due to the interference of bird posture and image quality degradation. In practice, birds have various posture types in the natural environment, and different posture types will cause significant fluctuations in recognition accuracy. In addition, affected by weather (e.g., rain, snow, haze) and / or obstructions (e.g., dirt), the image quality will be degraded, and the use of low-quality images for bird recognition will greatly affect the accuracy and reliability of recognition. Based on this, the bird recognition method based on bird posture and image quality of some embodiments of the present disclosure, first, in response to the detection of bird activity in the feeding area, the real-time monitoring video corresponding to the above-mentioned feeding area is collected. By detecting the presence of bird activity in the feeding area and then starting the collection of real-time monitoring video, it can effectively avoid unnecessary power consumption caused by long-term video collection and save video storage space. Secondly, the above-mentioned real-time monitoring video is sampled to obtain a sequence of images to be detected. In practice, bird activities are continuous, which is reflected in the fact that the video screens of continuous frames in the real-time monitoring video contain similar bird behaviors. Video downsampling can effectively reduce the amount of data processing. Then, for each image to be detected in the above-mentioned image sequence to be detected, the following processing steps are performed: the first step is to perform bird posture recognition on the above-mentioned image to be detected to generate a bird posture recognition result, thereby determining the bird posture in the image to be detected. The second step is to determine the image quality result corresponding to the above-mentioned image to be detected, thereby quantifying the image quality of the image to be detected. The third step is to perform bird recognition on the above-mentioned image to be detected to determine the bird recognition result. The fourth step is to optimize the recognition result of the above-mentioned bird recognition result based on the above-mentioned bird posture recognition result and the above-mentioned image quality result to obtain the optimized bird recognition result, thereby weighted fusion of the bird recognition result from the perspective of bird posture and image quality to optimize the recognition result. Finally, based on the obtained optimized bird recognition result sequence, a global bird recognition result for the above-mentioned real-time monitoring video is generated. Finally, based on the optimized bird recognition result corresponding to each image to be detected, the bird recognition result corresponding to the real-time monitoring video is determined. In this way, the accuracy of bird recognition is greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0030] Figure 1is a flow chart of some embodiments of a bird identification method based on bird posture and image quality according to the present disclosure;

[0031] Figure 2 is a flow chart of other embodiments of a bird recognition method based on bird posture and image quality according to the present disclosure;

[0032] Figure 3 It is a schematic diagram of the relationship between the positions of infrared sensors, feeding areas and birds;

[0033] Figure 4 It is a schematic diagram of the positional relationship between the pressure sensor, the feeding area and the birds;

[0034] Figure 5 It is a schematic diagram of the positional relationship between the infrared sensor, the pressure sensor, the feeding area and the birds;

[0035] Figure 6 is a schematic structural diagram of some embodiments of a bird identification device based on bird posture and image quality according to the present disclosure;

[0036] Figure 7 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0037] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0038] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0039] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0040] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0041] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0042] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0043] refer to Figure 1 , shows a process 100 of some embodiments of a bird recognition method based on bird posture and image quality according to the present disclosure. The bird recognition method based on bird posture and image quality comprises the following steps:

[0044] Step 101, in response to detecting bird activity in a feeding area, collecting real-time monitoring video corresponding to the feeding area.

[0045] In some embodiments, the execution subject (e.g., a computing device) of the bird recognition method based on bird posture and image quality can collect real-time monitoring videos corresponding to the feeding area in response to detecting the presence of bird activities in the feeding area. Among them, the feeding area can be an area for feeding birds. In practice, by setting up a feeding area, the probability of attracting birds can be increased, and the probability of including birds in the collected real-time monitoring video can be increased. In addition, wild animals often have the characteristic of avoiding people. Therefore, it is often necessary to set up a feeding area in an area with low human density for collecting real-time monitoring videos. In this case, the camera used to shoot real-time monitoring videos can be powered by a lithium battery or a solar cell. In order to ensure power efficiency, the detection of bird activities in the feeding area is used as a condition to trigger the collection of real-time monitoring videos.

[0046] As an example, a bird presence sensor may be set toward the feeding area. Specifically, the bird presence sensor may be implemented by a millimeter wave radar, and by combining the echo of the radar wave emitted by the millimeter wave radar, it is determined whether there is bird activity in the feeding area.

[0047] It should be noted that the above-mentioned computing device can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the computing device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, for example, or it can be implemented as a single software or software module. No specific limitation is made here.

[0048] Step 102: sampling the real-time monitoring video to obtain a sequence of images to be detected.

[0049] In some embodiments, the execution subject may perform video sampling on the real-time monitoring video to obtain a sequence of images to be detected. The sequence of images to be detected is an ordered sequence of images obtained by downsampling the real-time monitoring video. In practice, the real-time monitoring video may be sampled by downsampling the frame rate to obtain the sequence of images to be detected.

[0050] As an example, the video frame rate of the real-time monitoring video may be 30 FPS (Frame Per Second). Therefore, the video frame rate of the monitoring video may be reduced, for example, to 15 FPS, so as to obtain the image sequence to be detected.

[0051] Step 103: for each image to be detected in the sequence of images to be detected, perform the following processing steps:

[0052] Step 1031 , performing bird posture recognition on the image to be detected to generate a bird posture recognition result.

[0053] In some embodiments, the above-mentioned execution subject can perform bird posture recognition on the image to be detected to generate a bird posture recognition result. Among them, the bird posture recognition result represents the confidence of the bird in different postures contained in the image to be detected. In practice, there are forward posture, lateral posture and back posture. Among them, the forward posture represents that the bird is facing the camera, so that the collected image to be detected contains the front of the bird. The lateral posture represents that the bird is facing the camera sideways, so that the collected image to be detected contains the side of the bird. The back posture represents that the bird is facing the camera back, so that the collected image to be detected contains the north side of the bird. Specifically, considering the limited computing power of the edge end (camera), a lightweight recognition model can be used, for example, a MobileNet model, a ResNet model, to perform bird posture recognition on the image to be detected to generate a bird posture recognition result. For the recognition model used for bird posture recognition, an image data set with annotated bird postures can be collected and divided into a training sample set and a verification sample set. In addition, in order to enrich the number of samples, data enhancement methods such as flipping and cropping can be used. Thereby improving the robustness of the recognition model obtained by subsequent training for different posture recognition. In the model training phase, the cross entropy loss function is used to optimize the model. Stochastic gradient descent or Adam optimizer is used to control the weight update of the model during the iteration process. In addition, in the model training or testing phase, the accuracy of different postures is statistically counted to map the corresponding weights. For example, if the accuracy of identifying the bird's posture as a forward posture is 80%, the corresponding weight is 0.8. For another example, if the accuracy of identifying the bird's posture as a sideways posture is 95%, the corresponding weight is 0.95. For another example, if the accuracy of identifying the bird's posture as a back posture is 60%, the corresponding weight is 0.6. In practice, the recognition model can be continuously iterated based on the continuously enriched training samples, so the (confidence) weight corresponding to the bird's posture is continuously optimized.

[0054] Step 1032: Determine the image quality result corresponding to the image to be detected.

[0055] In some embodiments, the above-mentioned execution entity can determine the image quality result corresponding to the image to be detected. Among them, the image quality result represents whether the image to be detected is clear or whether there is occlusion. In practice, the image quality result can be represented by a score value between 0 and 1, that is, the lower the score value, the less clear the image to be detected is or the existence of occlusion. The higher the score, the clearer the image to be detected is or the absence of occlusion. Specifically, a lightweight convolutional neural network, such as a MobileNet or ResNet model, can be used to determine the image quality result corresponding to the image to be detected. Images annotated with corresponding image quality results can be collected as training samples and verification samples. The loss function can use a mean square error loss function. In the training stage, an Adam optimizer can be used to control the weight update of the model during the iteration process.

[0056] Step 1033: Perform bird recognition on the image to be detected to determine a bird recognition result.

[0057] In some embodiments, the above-mentioned execution entity can perform bird recognition on the image to be detected to determine the bird recognition result. The bird classification result represents the classification results and corresponding confidence levels for different bird types. In practice, a lightweight model such as the ResNet model, the EfficientNet model, and the ViT (Vision Transformer) model can be used to perform bird recognition on the image to be detected to determine the bird recognition result. A multi-category cross entropy loss function can be used to ensure the ability of the trained model to distinguish fine-grained categories. The output format is a 1×N vector, which represents the recognition confidence level for N bird categories.

[0058] Step 1034, optimizing the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result.

[0059] In some embodiments, the execution subject may optimize the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result. In practice, the execution subject may use the maximum confidence contained in the bird recognition result and the image quality result as weights to update the confidence for N bird categories contained in the bird recognition result to obtain a bird recognition result containing the updated confidence corresponding to the bird category. Specifically, the update method is simplified to the following formula:

[0060] P adjusted (c) = w pose × quality ×P model (c)

[0061] Among them, P adjusted(c) represents the updated confidence for bird category c. pose Indicates the maximum confidence contained in the bird posture recognition results. quality Characterize the image quality results. model (c) Characterizes the confidence level of the bird category C contained in the bird posture recognition information.

[0062] Step 104: Generate a global bird recognition result for the real-time monitoring video according to the obtained optimized bird recognition result sequence.

[0063] In some embodiments, the execution subject may generate a global bird recognition result for the real-time surveillance video based on the obtained optimized bird recognition result sequence. In practice, since the optimized bird recognition result includes updated confidences for N bird categories, a statistical method may be used to determine the bird category with the most occurrences and the maximum confidence in the bird recognition result sequence as the global bird recognition result.

[0064] As an example, the bird category may include: c1, c2, c3. The bird recognition result sequence may include [0.1, 0.3, 0.7], [0.1, 0.2, 0.9], [0.1, 0.2, 0.9], where the confidence of the first dimension in the bird recognition result corresponds to c1, the confidence of the second dimension in the bird recognition result corresponds to c2, and the confidence of the third dimension in the bird recognition result corresponds to c3. Therefore, c3 is the global bird recognition result.

[0065] The above-mentioned embodiments of the present disclosure have the following beneficial effects: through the bird recognition method based on bird posture and image quality of some embodiments of the present disclosure, the bird recognition accuracy is improved. Specifically, the reason for the low bird recognition accuracy is that it is difficult to accurately recognize birds due to the interference of bird posture and image quality degradation. In practice, birds have various posture types in the natural environment, and different posture types will cause significant fluctuations in recognition accuracy. In addition, affected by weather (e.g., rain, snow, haze) and / or obstructions (e.g., dirt), the image quality will be degraded, and the use of low-quality images for bird recognition will greatly affect the accuracy and reliability of recognition. Based on this, the bird recognition method based on bird posture and image quality of some embodiments of the present disclosure, first, in response to the detection of bird activity in the feeding area, the real-time monitoring video corresponding to the above-mentioned feeding area is collected. By detecting the presence of bird activity in the feeding area and then starting the collection of real-time monitoring video, it can effectively avoid unnecessary power consumption caused by long-term video collection and save video storage space. Secondly, the above-mentioned real-time monitoring video is sampled to obtain a sequence of images to be detected. In practice, bird activities are continuous, which is reflected in the fact that the video screens of continuous frames in the real-time monitoring video contain similar bird behaviors. Video downsampling can effectively reduce the amount of data processing. Then, for each image to be detected in the above-mentioned image sequence to be detected, the following processing steps are performed: the first step is to perform bird posture recognition on the above-mentioned image to be detected to generate a bird posture recognition result, thereby determining the bird posture in the image to be detected. The second step is to determine the image quality result corresponding to the above-mentioned image to be detected, thereby quantifying the image quality of the image to be detected. The third step is to perform bird recognition on the above-mentioned image to be detected to determine the bird recognition result. The fourth step is to optimize the recognition result of the above-mentioned bird recognition result based on the above-mentioned bird posture recognition result and the above-mentioned image quality result to obtain the optimized bird recognition result, thereby weighted fusion of the bird recognition result from the perspective of bird posture and image quality to optimize the recognition result. Finally, based on the obtained optimized bird recognition result sequence, a global bird recognition result for the above-mentioned real-time monitoring video is generated. Finally, based on the optimized bird recognition result corresponding to each image to be detected, the bird recognition result corresponding to the real-time monitoring video is determined. In this way, the accuracy of bird recognition is greatly improved.

[0066] Further references Figure 2 , which shows a process 200 of another embodiment of a bird recognition method based on bird posture and image quality. The process 200 of the bird recognition method based on bird posture and image quality includes the following steps:

[0067] Step 201 , detecting heat source changes and / or pressure changes in the feeding area by using an infrared sensor and / or a pressure sensor to determine whether there is bird activity in the feeding area.

[0068] In some embodiments, the execution entity (e.g., a computing device) of the bird recognition method based on bird posture and image quality can detect heat source changes and / or pressure changes in the feeding area through infrared sensors and / or pressure sensors to determine whether there is bird activity in the feeding area.

[0069] As an example, see Figure 2 Schematic diagram of the positional relationship between the infrared sensor, the feeding area and the bird. The infrared sensor 2 can be directed toward the feeding area 1 to detect changes in the heat source in the feeding area 1. When there is a bird 3 in the feeding area 1, the infrared sensor 1 will detect the presence of a heat source in the feeding area 1, thereby indicating the presence of bird activity in the feeding area 1.

[0070] As yet another example, see Figure 4 The schematic diagram of the positional relationship among the pressure sensor, the feeding area and the birds is shown, wherein a pressure sensor 4 is arranged below the feeding area 1. When there are birds 3 in the feeding area 1, the pressure sensor 4 can detect the pressure change in the feeding area 1, thereby indicating the presence of bird activities in the feeding area 1.

[0071] As yet another example, see Figure 5 The schematic diagram of the positional relationship between the infrared sensor, the pressure sensor, the feeding area and the bird is shown. The infrared sensor 2 can face the feeding area 1 to detect the change of the heat source in the feeding area 1. The pressure sensor 4 is arranged below the feeding area 1 to detect the pressure change in the feeding area 1. When there is a bird 3 in the feeding area 1, the infrared sensor 2 will detect the change of the heat source, and the pressure sensor 4 will detect the pressure change, so as to indicate the presence of bird activities in the feeding area 1.

[0072] Step 202, in response to detecting the presence of bird activity in the feeding area, collecting real-time monitoring video corresponding to the feeding area.

[0073] In some embodiments, the specific implementation of step 202 and the technical effects thereof can be referred to in Figure 1 The corresponding step 102 in the embodiment will not be described in detail here.

[0074] Step 203: Downsample the real-time monitoring video at preset time intervals to obtain a sequence of images to be detected.

[0075] In some embodiments, the above-mentioned execution subject may downsample the real-time monitoring video at a preset time interval to obtain a sequence of images to be detected. In practice, the preset time interval is less than the acquisition time of the real-time monitoring video. For example, the acquisition time of the real-time monitoring video may be 120 seconds. The preset time interval may be 5 seconds, that is, the real-time monitoring video is framed once every 5 seconds to obtain the sequence of images to be detected.

[0076] In some optional implementations of some embodiments, the execution subject performs video sampling on the real-time monitoring video to obtain a sequence of images to be detected, including:

[0077] In the first step, the real-time monitoring video is downsampled at preset time intervals to obtain a candidate image sequence.

[0078] In practice, since the movement of birds is random, that is, video downsampling is performed at regular time intervals, the candidate images in the obtained candidate image sequence may not contain birds.

[0079] In the second step, for each candidate image in the above candidate image sequence, the following screening steps are performed:

[0080] The first sub-step is to perform image binarization processing on the candidate image to obtain a binarized image.

[0081] In practice, binarization processing can reduce the amount of subsequent data processing and alleviate the data processing pressure at the edge (camera).

[0082] The second sub-step is to dilate the binarized image to obtain a dilated image.

[0083] In practice, the Canndy operator can be used to dilate the above binarized image to obtain a dilated image. Specifically, during the binarization process, a threshold needs to be set for image binarization. If the threshold is not selected properly, the binarization may not be accurate enough, especially for small areas with unclear boundaries. After binarization, the boundary of the area will shrink. The dilation process can restore the boundary of the area to a certain extent.

[0084] The third sub-step is to determine the image difference between the expanded image and the background image to obtain at least one difference set region.

[0085] The background image is an image when there is no occluder in the feeding area. Specifically, the background image can be a binary image when there is no occluder in the feeding area. At least one difference set region can represent at least one existing occluder.

[0086] The fourth sub-step is to determine the candidate image as the image to be detected in response to the presence of a difference set region whose corresponding area is larger than a preset area in the at least one difference set region.

[0087] In practice, the difference region can represent bird occlusion or dirt occlusion. However, birds often have a certain size, so the corresponding difference region also has a certain area. Therefore, by setting the preset area as the threshold, the candidate images containing only the difference region of a small area or the candidate images not containing the difference region are filtered, that is, only the candidate images that may contain birds are retained as the images to be detected. In this way, further image screening is achieved based on image sampling at fixed time intervals.

[0088] Step 204: for each image to be detected in the sequence of images to be detected, perform the following processing steps:

[0089] Step 2041 , performing bird posture recognition on the image to be detected to generate a bird posture recognition result.

[0090] In some embodiments, the execution subject performs bird posture recognition on the image to be detected to generate a bird posture recognition result, which may include the following steps:

[0091] The first step is to determine the probability distribution of bird postures corresponding to the above-mentioned images to be detected through a pre-trained bird posture recognition model.

[0092] Among them, the above-mentioned bird posture probability distribution represents the confidence of different bird postures corresponding to the bird contained in the image to be detected. In practice, considering the limited computing power of the edge end (camera), and considering that birds will occupy a certain screen when appearing in the image to be detected, the bird posture recognition model can be composed of five layers of serially connected convolution layers. Specifically, the bird posture recognition model may include: convolution layer A, convolution layer B, convolution layer C, convolution layer D, convolution layer E and fully connected layer F. Among them, convolution layer A, convolution layer B and convolution layer C each contain 3 parallel connected convolution layers and one fusion layer. Taking convolution layer A as an example, convolution layer A includes: convolution layer A1, convolution layer A2, convolution layer A3 and fusion layer A4. Among them, the convolution kernel of convolution layer A1 is 1×1. The convolution kernel of convolution layer A2 is 3×3. The convolution kernel of convolution layer A3 is 5×5. The fusion layer A4 is used to superimpose the three features output by the convolutional layers A1, A2, and A3. By setting the convolutional layers A1, A2, and A3, feature extraction under different receptive fields is achieved. The vector dimension of the feature vector output by the fully connected layer is 1×3, which includes the confidence of the forward posture, side posture, and back posture.

[0093] The second step is to determine the bird posture recognition result according to the bird posture distribution.

[0094] In practice, the above-mentioned execution entity may use the bird posture corresponding to the maximum confidence in the bird posture distribution as the bird posture recognition result.

[0095] Step 2042: Perform image quality scoring on the image to be detected using a pre-trained image quality scoring model to obtain an image quality result.

[0096] In some embodiments, the execution subject can perform image quality scoring on the image to be detected through a pre-trained image quality scoring model to obtain an image quality result. In practice, considering that both bird posture recognition and image quality scoring are based on the image to be detected, that is, there are reused image features in the process of bird posture recognition and image quality scoring, in order to further alleviate the data processing pressure of the edge end (camera), the image quality scoring model may include: convolution layer A, convolution layer B, convolution layer C, convolution layer G, convolution layer H, convolution layer I, fully connected layer J and fully connected layer H. Among them, the image quality scoring model and the bird posture recognition model share convolution layer A, convolution layer B and convolution layer C. Convolution layer D and convolution layer E are deep feature extractions of the output features of convolution layer C, and the corresponding parameters are mainly for bird posture recognition. Therefore, convolution layer G, convolution layer H, and convolution layer I are selected in the image quality scoring model. In addition, considering that the image quality result is a 1×1 vector, that is, it corresponds to a value of 0-1. Therefore, the fully connected layer J and the fully connected layer H are set to transform the feature dimension of the output of the convolutional layer I to 1×1.

[0097] Step 2043: Perform bird recognition on the image to be detected by using the pre-trained fine-grained bird recognition model to determine a bird recognition result.

[0098] In some embodiments, the above-mentioned execution subject can perform bird recognition on the image to be detected through a pre-trained fine-grained bird recognition model to determine the bird recognition result. In practice, further considering the computing power pressure of the edge end (camera), a lightweight fine-grained bird recognition model and a non-lightweight fine-grained bird recognition model can be pre-set. The execution subject can be more complex and adaptively select a lightweight fine-grained bird recognition model or a non-lightweight fine-grained bird recognition model to perform bird recognition on the image to be detected to determine the bird recognition result. Specifically, the lightweight fine-grained bird recognition model can use the EfficientNet-B1 model as the backbone network. The non-lightweight fine-grained bird recognition model can use the EfficientNet-B7 model backbone network. In order to further improve the feature processing efficiency and reduce the edge computing pressure. The input of the fine-grained bird recognition model is a feature map after the feature processing of convolution layer A, convolution layer B, and convolution layer C. In this way, the use and processing efficiency of features can be greatly improved. In addition, during the training stage, since the bird posture recognition model, image quality scoring model and fine-grained bird recognition model have common model parts (convolutional layer A, convolutional layer B and convolutional layer C), in order to improve the model training efficiency, the bird posture recognition model, image quality scoring model and fine-grained bird recognition model are trained in parallel, and the voting method is used to update the weights when updating convolutional layer A, convolutional layer B and convolutional layer C.

[0099] Step 2044, using the bird posture recognition result and the image quality result as optimization weights, the bird recognition result is confidence-weighted fused to obtain an optimized bird recognition result.

[0100] In some embodiments, the above-mentioned execution subject can use the bird posture recognition result and the image quality result as optimization weights, perform confidence-weighted fusion on the bird recognition result, and obtain the optimized bird recognition result. Specifically, please refer to the formula corresponding to the update method shown in step 1034, but considering that the maximum confidence value range of the bird posture recognition result is 0-1, and the area range of the image quality result is 0-1, that is, the maximum confidence value and the image quality result included in the bird posture recognition result are used as weights, which may cause the updated confidence contained in the optimized bird recognition result to be reduced in order of magnitude, which is manifested as an extremely small confidence. Therefore, the formula corresponding to the update method of step 1034 can expand the image quality result tenfold. This can offset the existing order of magnitude reduction problem to a certain extent.

[0101] In some optional implementations of some embodiments, the execution subject uses the bird posture recognition result and the image quality result as optimization weights, performs confidence-weighted fusion on the bird recognition result, and obtains the optimized bird recognition result, including:

[0102] In the first step, the bird posture recognition results and the image quality results are used as optimization weights to perform confidence weighted fusion on the bird recognition results to obtain the weighted fused bird recognition results.

[0103] In the second step, the above bird recognition results and the above weighted fusion bird recognition results are distributed and balanced through the balancing factor to obtain the above optimized bird recognition results.

[0104] In practice, the above execution entity can use the maximum confidence level, image quality results, and balance factors contained in the bird recognition results as weights to update the confidence levels for N bird categories contained in the bird recognition results, and obtain bird recognition results containing the updated confidence levels corresponding to the bird categories. For details, please refer to the following formula:

[0105] P adjusted (c) = (1-α) × (w pose × quality ×P model (c))+α×P model (c)

[0106] Among them, P adjusted (c) represents the updated confidence for bird category c. pose Indicates the maximum confidence contained in the bird posture recognition results. quality Characterize the image quality results. model (c) represents the confidence of the bird posture recognition information for the bird category C. represents the balance factor.

[0107] In some optional implementations of some embodiments, the execution subject uses the bird posture recognition result and the image quality result as optimization weights, performs confidence-weighted fusion on the bird recognition result, and obtains the optimized bird recognition result, including:

[0108] In the first step, the bird posture recognition results and the image quality results are used as optimization weights to perform confidence weighted fusion on the bird recognition results to obtain the weighted fused bird recognition results.

[0109] In practice, the weighted fusion may be performed using the formula corresponding to the updating method shown in step 1034 , or the weighted fusion may be performed using the formula corresponding to the updating method in step 2044 .

[0110] In the second step, the weighted fusion bird recognition results are normalized through the activation layer to obtain the optimized bird recognition results.

[0111] In practice, the above weighted fusion bird recognition results are normalized by activating the Softmax function of the layer to obtain the above optimized bird recognition results, that is, the confidence value with a reduced order of magnitude is remapped to the range of 0-1.

[0112] Step 205: Generate a global bird recognition result for the real-time monitoring video according to the obtained optimized bird recognition result sequence.

[0113] In some embodiments, the specific implementation of step 205 and the technical effects thereof can be referred to in Figure 1 The corresponding step 105 in the embodiment will not be described in detail here.

[0114] from Figure 2 It can be seen that Figure 1 Compared with the description of some corresponding embodiments, the present disclosure further optimizes the model structure of the bird posture recognition model, the image quality scoring model and the fine-grained bird recognition model on the basis of reducing the computing pressure of the edge end (camera) as much as possible, and avoids repeated processing of features by sharing the model structure, thereby alleviating the computing pressure of the edge end. In addition, by setting a balancing factor or an activation layer, the problem of a decrease in the order of magnitude of confidence after optimizing the weight is overcome. The accuracy of bird recognition is further guaranteed.

[0115] Further references Figure 6 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a bird identification device based on bird posture and image quality. These device embodiments are similar to Figure 1 Corresponding to the method embodiments shown, the bird identification device based on bird posture and image quality can be specifically applied to various electronic devices.

[0116] like Figure 6As shown, in some embodiments, a bird recognition device 600 based on bird posture and image quality includes: a collection unit 601, a video sampling unit 602, an execution unit 603 and a generation unit 604. The collection unit 601 is configured to collect the real-time monitoring video corresponding to the feeding area in response to detecting the presence of bird activities in the feeding area; the video sampling unit 602 is configured to perform video sampling on the real-time monitoring video to obtain a sequence of images to be detected; the execution unit 603 is configured to perform the following processing steps for each image to be detected in the sequence of images to be detected: perform bird posture recognition on the image to be detected to generate a bird posture recognition result; determine the image quality result corresponding to the image to be detected; perform bird recognition on the image to be detected to determine a bird recognition result; optimize the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result; the generation unit 604 is configured to generate a global bird recognition result for the real-time monitoring video according to the obtained optimized bird recognition result sequence.

[0117] In some optional implementations of some embodiments, the bird identification device based on bird posture and image quality also includes: a detection unit (not shown in the figure), which is configured to, before the above-mentioned response to detecting the presence of bird activity in the feeding area and collecting the real-time monitoring video corresponding to the above-mentioned feeding area, the above-mentioned method also includes: detecting heat source changes and / or pressure changes in the feeding area through infrared sensors and / or pressure sensors to determine whether there is bird activity in the feeding area.

[0118] In some optional implementations of some embodiments, the video sampling unit 602 is further configured to: perform video downsampling on the real-time monitoring video at preset time intervals to obtain a sequence of images to be detected.

[0119] In some optional implementations of some embodiments, the video sampling unit 602 is further configured to: perform video downsampling on the real-time monitoring video at preset time intervals to obtain a candidate image sequence; for each candidate image in the candidate image sequence, perform the following screening steps: perform image binarization on the candidate image to obtain a binarized image; perform dilation on the binarized image to obtain a dilated image; determine the image difference between the dilated image and the background image to obtain at least one difference set area, wherein the background image is an image when there is no obstruction in the feeding area; in response to the presence of a difference set area whose corresponding area area is greater than a preset area area in at least one of the difference set areas, determine the candidate image as an image to be detected.

[0120] In some optional implementations of some embodiments, the execution unit 603 is further configured to: determine the bird posture probability distribution corresponding to the image to be detected through a pre-trained bird posture recognition model, wherein the bird posture probability distribution represents the confidence level of different bird postures corresponding to the birds contained in the image to be detected; and determine the bird posture recognition result based on the bird posture distribution.

[0121] In some optional implementations of some embodiments, the execution unit 603 is further configured to: perform image quality scoring on the image to be detected by using a pre-trained image quality scoring model to obtain the image quality result.

[0122] In some optional implementations of some embodiments, the execution unit 603 is further configured to: perform bird recognition on the image to be detected by using a pre-trained fine-grained bird recognition model to determine the bird recognition result.

[0123] In some optional implementations of some embodiments, the execution unit 603 is further configured to: use the bird posture recognition result and the image quality result as optimization weights, perform confidence-weighted fusion on the bird recognition result, and obtain the optimized bird recognition result.

[0124] In some optional implementations of some embodiments, the execution unit 603 is further configured to: perform confidence-weighted fusion on the bird recognition results using the bird posture recognition results and the image quality results as optimization weights to obtain the weighted fused bird recognition results; and perform distribution balancing on the bird recognition results and the weighted fused bird recognition results using a balancing factor to obtain the optimized bird recognition results.

[0125] In some optional implementations of some embodiments, the execution unit 603 is further configured to: perform confidence-weighted fusion on the bird recognition results using the bird posture recognition results and the image quality results as optimization weights to obtain weighted fused bird recognition results; and perform normalization on the weighted fused bird recognition results through an activation layer to obtain the optimized bird recognition results.

[0126] It can be understood that the units described in the bird identification device 600 based on bird posture and image quality are similar to those described in the reference Figure 1 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the bird identification device 600 based on bird posture and image quality and the units included therein, and will not be described in detail here.

[0127] Reference below Figure 7, which shows a schematic structural diagram of an electronic device (eg, a computing device) 700 suitable for implementing some embodiments of the present disclosure. Figure 7 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0128] like Figure 7 As shown, the electronic device 700 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory 702 or a program loaded from a storage device 708 into a random access memory 703. Various programs and data required for the operation of the electronic device 700 are also stored in the random access memory 703. The processing device 701, the read-only memory 702, and the random access memory 703 are connected to each other via a bus 704. An input / output interface 705 is also connected to the bus 704.

[0129] Typically, the following devices may be connected to the input / output interface 705: an input device 706 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 707 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 708 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 709. The communication device 709 may allow the electronic device 700 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 7 The electronic device 700 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 7 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0130] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from a network through a communication device 709, or installed from a storage device 708, or installed from a read-only memory 702. When the computer program is executed by the processing device 701, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0131] It should be noted that the computer-readable medium recorded in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0132] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (Hyper Text Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0133] The computer-readable medium may be included in the electronic device; or it may exist independently without being assembled into the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: in response to detecting the presence of bird activities in the feeding area, collects the real-time monitoring video corresponding to the feeding area; performs video sampling on the real-time monitoring video to obtain a sequence of images to be detected; for each image to be detected in the sequence of images to be detected, performs the following processing steps: performs bird posture recognition on the image to be detected to generate a bird posture recognition result; determines the image quality result corresponding to the image to be detected; performs bird recognition on the image to be detected to determine the bird recognition result; optimizes the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result; generates a global bird recognition result for the real-time monitoring video according to the obtained optimized bird recognition result sequence.

[0134] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0135] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0136] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be provided in a processor, for example, may be described as: a processor including an acquisition unit, a video sampling unit, an execution unit, and a generation unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the acquisition unit may also be described as "a unit for collecting real-time monitoring video corresponding to the above-mentioned feeding area in response to detecting the presence of bird activity in the feeding area".

[0137] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0138] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) and the technical solutions formed.

Claims

1. A bird recognition method based on bird posture and image quality, comprising: In response to detecting bird activity in the feeding area, collecting real-time monitoring video corresponding to the feeding area; Performing video sampling on the real-time monitoring video to obtain an image sequence to be detected; For each image to be detected in the sequence of images to be detected, the following processing steps are performed: Performing bird posture recognition on the image to be detected to generate a bird posture recognition result; Determine an image quality result corresponding to the image to be detected; Performing bird recognition on the image to be detected to determine a bird recognition result; Optimizing the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result; According to the obtained optimized bird recognition result sequence, a global bird recognition result for the real-time monitoring video is generated.

2. The method according to claim 1, wherein: Before collecting the real-time monitoring video corresponding to the feeding area in response to detecting the presence of bird activities in the feeding area, the method further includes: By using infrared sensors and / or pressure sensors, changes in heat sources and / or pressure changes in the feeding area are detected to determine whether there is bird activity in the feeding area.

3. The method according to claim 1, wherein: The step of sampling the real-time monitoring video to obtain a sequence of images to be detected includes: The real-time monitoring video is downsampled at preset time intervals to obtain a sequence of images to be detected.

4. The method according to claim 1, wherein: The step of sampling the real-time monitoring video to obtain a sequence of images to be detected includes: Downsampling the real-time surveillance video at preset time intervals to obtain a candidate image sequence; For each candidate image in the candidate image sequence, the following screening steps are performed: Performing image binarization processing on the candidate image to obtain a binarized image; Performing dilation processing on the binary processed image to obtain a dilated image; Determine an image difference between the expanded image and a background image to obtain at least one difference set region, wherein the background image is an image when there is no occluder in the feeding area; In response to the presence of a corresponding region whose area is greater than a preset region in the at least one difference region, the candidate image is determined as the image to be detected.

5. The method according to claim 1, wherein: The performing bird posture recognition on the image to be detected to generate a bird posture recognition result includes: Determine the probability distribution of bird postures corresponding to the image to be detected by a pre-trained bird posture recognition model, wherein the probability distribution of bird postures represents the confidence level of different bird postures corresponding to the bird contained in the image to be detected; The bird posture recognition result is determined according to the bird posture distribution.

6. The method according to claim 1, wherein: The determining of the image quality result corresponding to the image to be detected includes: The image quality score is performed on the image to be detected by using a pre-trained image quality score model to obtain the image quality result.

7. The method according to claim 1, wherein: The performing bird recognition on the image to be detected to determine the bird recognition result includes: The bird recognition is performed on the image to be detected by using a pre-trained fine-grained bird recognition model to determine the bird recognition result.

8. The method according to claim 1, wherein: The step of optimizing the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result includes: The bird posture recognition result and the image quality result are used as optimization weights, and the bird recognition result is confidence-weighted fused to obtain the optimized bird recognition result.

9. The method according to claim 8, wherein: The step of taking the bird posture recognition result and the image quality result as optimization weights and performing confidence-weighted fusion on the bird recognition result to obtain the optimized bird recognition result includes: Taking the bird posture recognition result and the image quality result as optimization weights, performing confidence weighted fusion on the bird recognition result to obtain a weighted fused bird recognition result; The bird recognition result and the weighted fusion bird recognition result are distributed and balanced by a balance factor to obtain the optimized bird recognition result.

10. The method according to claim 8, wherein: The step of taking the bird posture recognition result and the image quality result as optimization weights and performing confidence-weighted fusion on the bird recognition result to obtain the optimized bird recognition result includes: Taking the bird posture recognition result and the image quality result as optimization weights, performing confidence weighted fusion on the bird recognition result to obtain a weighted fused bird recognition result; The weighted fusion bird recognition result is normalized through the activation layer to obtain the optimized bird recognition result.

11. A bird identification device based on bird posture and image quality, comprising: A collection unit is configured to collect real-time monitoring video corresponding to the feeding area in response to detecting the presence of bird activities in the feeding area; A video sampling unit is configured to perform video sampling on the real-time monitoring video to obtain a sequence of images to be detected; The execution unit is configured to perform the following processing steps for each image to be detected in the sequence of images to be detected: performing bird posture recognition on the image to be detected to generate a bird posture recognition result; determining an image quality result corresponding to the image to be detected; Performing bird recognition on the image to be detected to determine a bird recognition result; Optimizing the bird recognition result according to the bird posture recognition result and the image quality result to obtain an optimized bird recognition result; The generating unit is configured to generate a global bird recognition result for the real-time monitoring video according to the obtained optimized bird recognition result sequence.

12. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 10.

13. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 10 is implemented.

Images