Intelligent bird repelling method and system

Abstract

A bird repelling method, which is applicable to a bird repelling system. A moving mechanism (201) is combined with an intelligent image determination model to effectively identify and repel birds. The method comprises: first, on the basis of a position preset by a user, controlling a moving mechanism (201) to move, and capturing an image at the designated position; then, using an intelligent image model to determine bird positions, accordingly planning a movement path and a starting occasion for a repelling apparatus (203), and formulating a repelling strategy and implementing same; after repelling, capturing an image again, and comparing the images before and after the repelling to evaluate repelling effectiveness; and finally, storing an evaluation result as historical data, and performing strategy review and optimization.

Description

Intelligent bird repelling method and system TECHNICAL FIELD

[0001] The present invention relates to a bird repelling technology, and more particularly, to an intelligent image capturing and recognition based method and system for automatically repelling birds. BACKGROUND

[0002] In the field of preventing avian influenza and reducing feed loss, the prior art generally relies on artificially preset repelling methods, such as using sound devices, fake hawk models, flash devices or other visual and auditory stimulation means to repel birds. However, these methods have many obvious shortcomings. Maintaining and operating these devices consumes manpower and cost, increasing the management burden of the farm, which is particularly disadvantageous for large-scale farms. In summary, the traditional artificially preset bird repelling method is difficult to adapt to the changing behavior of birds and environmental conditions, and has defects such as decreasing effect, resource waste, and epidemic prevention risk.

[0003] Figure 1 shows a schematic diagram of a bird repelling device of the prior art. Referring to Figure 1, the bird repelling device includes a moving mechanism 100, a camera device 101, and a laser emitting device 102. The bird repelling device can be programmed to control the moving mechanism 100 and the laser emitting device 102 in several different modes. In addition, each mode can be assigned to a specific time period, so that various bird repelling modes are implemented at different times of the day. The results are observed through the camera device 101.

[0004] However, since the repelling method is mostly preset parameters, it lacks the ability to respond immediately, and it is difficult to adjust according to different bird species or current conditions, resulting in limited repelling effect. Birds themselves have a high learning and adaptation ability, and are prone to habitual reactions to fixed and repetitive stimuli, thereby ignoring these repelling means and making them ineffective. In addition, the current bird repelling device is mostly passive in operation, and cannot respond immediately according to the actual activity period or invasion path of the bird flock, reducing the efficiency of epidemic prevention. Therefore, the bird repelling technology needs to develop towards intelligence and dynamic adjustment.

[0005] The currently widely used patrol mode, such as the camera device 101 scanning at fixed angles and intervals, is not only time-consuming, but also inefficient. Such a mechanism often spends a lot of resources in areas where no birds appear, missing immediate repelling opportunities. At the same time, it is also unable to concentrate resources on the hot areas where birds frequently gather for intensive monitoring and repelling, greatly reducing the overall effectiveness. Furthermore, the traditional bird repelling technology lacks an immediate safety recognition mechanism. When non-target objects such as personnel, vehicles or livestock appear within the repelling range, the system cannot recognize and stop operating, causing the sound and light repelling devices to possibly interfere with or even harm humans and animals. SUMMARY

[0006] One object of the preferred embodiments of the present application is to provide an intelligent bird expelling method and system that uses image technology to determine bird location, number, and other information, and formulates strategies for bird expelling based on this information and the condition of the expelling device.

[0007] Therefore, a preferred embodiment of the present application provides a bird expelling method, which includes: setting at least one expelling location by a control device; directing a lens of an image capturing device toward the at least one expelling location by a moving mechanism; performing a first image capturing on the at least one expelling location by the lens of the image capturing device to obtain a first image; determining the locations of a plurality of birds in the first image by a smart image determination model; determining the moving trajectory of the moving mechanism and the activation of an expelling device based on the locations of the plurality of birds; determining an expelling strategy based on the expelling device; executing the expelling strategy by the expelling device and the moving mechanism; performing a second image capturing to obtain a second image after the execution of the expelling strategy; determining the second image and the first image by the smart image determination model to obtain an expelling effectiveness evaluation data; storing the expelling effectiveness evaluation data; and performing a strategy review calculation based on the expelling effectiveness evaluation data to change the expelling strategy.

[0008] Another preferred embodiment of the present application provides a bird expelling system, which includes a moving mechanism, an image capturing device, an expelling device, and a control device. The image capturing device is arranged on the moving mechanism. The expelling device is arranged on the moving mechanism, and the image capturing device and the expelling device are substantially aligned in the same direction. The control device is coupled to the moving mechanism, the image capturing device, and the expelling device, and includes a smart image determination model. The control device stores at least one expelling location set by a user.

[0009] In operation, the control device controls the moving mechanism to direct a lens of the image capturing device toward the at least one expelling location, and then controls the lens of the image capturing device to perform a first image capturing on the at least one expelling location to obtain a first image. The smart image determination model determines the locations of a plurality of birds in the first image. The control device determines the moving trajectory of the moving mechanism and the activation of the expelling device based on the locations of the plurality of birds. The control device then determines an expelling strategy based on the expelling device. The control device controls the expelling device and the moving mechanism to execute the expelling strategy. After the execution of the expelling strategy, the control device controls the lens of the image capturing device to perform a second image capturing to obtain a second image. The smart image determination model determines the second image and the first image to obtain an expelling effectiveness evaluation data. The control device stores the expelling effectiveness evaluation data. The control device performs a strategy review calculation based on the expelling effectiveness evaluation data to fine-tune the expelling strategy.

[0010] So that the foregoing and other objects, features and advantages of the present application can be better understood, a more detailed description of the present application will be rendered by reference to the following detailed description, DETAILED DESCRIPTION, which is illustrated in the accompanying drawings. Wherever possible and practical, identical reference numerals will be used throughout the drawings to refer to identical or similar components. BRIEF DESCRIPTION OF DRAWINGS

[0011] The accompanying drawings are included to provide a further understanding of the application, and are incorporated in and constitute a part of this specification. The drawings illustrate exemplary embodiments of the application and, together with the description, serve to explain principles of the application.

[0012] FIG. 1 shows a schematic diagram of a prior art bird repelling device.

[0013] FIG. 2 shows a schematic diagram of a bird repelling system according to a preferred embodiment of the present application.

[0014] FIG. 3 shows a schematic diagram of a sequential scanning of a bird repelling system according to a preferred embodiment of the present application.

[0015] FIG. 4 shows a schematic diagram of a focal point coordinate scanning of a bird repelling system according to a preferred embodiment of the present application.

[0016] FIG. 5 shows a flow chart of a bird repelling method using the bird repelling system according to a preferred embodiment of the present application.

[0017] FIG. 6 shows a schematic diagram of a bird repelling system according to a preferred embodiment of the present application, with a desired repelling area set.

[0018] FIG. 7 shows a schematic diagram of a bird repelling system according to a preferred embodiment of the present application, with a targeting path of a moving mechanism 201 driving a repelling device 203.

[0019] FIGS. 8A to 8D show a schematic diagram of a repelling strategy of a repelling device 203 of a bird repelling system according to a preferred embodiment of the present application.

[0020] FIG. 9 shows a flow chart of a bird repelling system according to a preferred embodiment of the present application.

[0021] FIG. 10 shows a schematic diagram of a bird repelling system according to a preferred embodiment of the present application, with a targeting path of a moving mechanism 201 driving a repelling device 203. DETAILED DESCRIPTION

[0022] Reference will now be made in detail to the exemplary embodiments of the present application, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers will be used throughout the drawings to refer to the same or like parts. Additionally, the exemplary embodiments are merely intended to make the design concept of the present application complete, and the following exemplary embodiments are not intended to limit the present application.

[0023] Figure 2 shows a schematic diagram of a bird repelling system according to a preferred embodiment of the present application. Referring to Figure 2, the bird repelling system comprises a moving mechanism 201, an image capturing device 202, a repelling device 203, and a control device 204. The image capturing device 202 can be a camera, a digital camera, an infrared camera, or the like, without being limited to the present application. The image capturing device 202 is configured on the moving mechanism 201, which can be freely rotated, such as a horizontal rotation, a vertical rotation, or a Pan / Tilt / Zoom (PTZ) control mechanism. The repelling device 203 is configured on the moving mechanism 201, such as being configured together with the image capturing device 202, in the present embodiment. In this way, the position of the lens of the image capturing device 202 is essentially the position of the repelling device 203. In the present embodiment, the repelling device 203 is mainly a long-range aiming device, such as a laser module, a water gun, a directional sound wave device, an air gun, or a sprayer.

[0024] The water gun is used for water spraying. The sound wave device is used for sound wave emission. The laser module is used for laser emission. The air gun is used for physical repelling. The sprayer can release a chemical agent or a water-based repellent. In the present embodiment, the repelling device 203 can be one or more, which is mainly determined according to different situations or the needs of customers or farmers. In other preferred embodiments, the bird repelling system can further comprise a non-directional repelling device, such as a remote-controlled moving device, for physical repelling. This part is also a design option, without being limited to the present application.

[0025] The image capturing device 202 and the repelling device 203 are essentially aimed in the same direction in design. The control device 204 is coupled to the moving mechanism 201, the image capturing device 202, and the repelling device 203. In the present embodiment, the control device 204 is similar to a small computer, which internally comprises a pre-trained edge computing intelligent image judgment model. In the present embodiment, the equipment manufacturer will pre-train the intelligent image judgment model according to different situations, such as using bird photos of a roof scene to train the intelligent image judgment model if the place to repel birds is a roof, or using bird photos of a cowshed scene to train the intelligent image judgment model if the place to repel birds is a cowshed. The edge computing intelligent image judgment model is mainly used to allow the user to directly apply the intelligent image judgment model on the edge device without connecting to a server, which can reduce the operation delay and does not need to consider the security and efficiency of communication, and the application can be very flexible.

[0026] The device generally adopts sequential scanning when the user does not set it. Fig. 3 is a schematic diagram of sequential scanning of the bird expelling system according to a preferred embodiment of the present application. Referring to Fig. 3, in this embodiment, the moving mechanism 201 sequentially aims the lens of the image capturing device 202 at the above-mentioned preset expelling positions one by one according to a fixed sequence, performs the expelling, and then enters the next expelling position. With this preset setting, the device periodically moves to scan the entire active area, but the disadvantage is that a lot of time is wasted in scanning the area without birds.

[0027] Fig. 4 is a schematic diagram of focal point coordinate scanning of the bird expelling system according to a preferred embodiment of the present application. Referring to Fig. 4, in this embodiment, the user can manually specify areas in the active area and set, for example, three bird occurrence hot zones. By setting the regular coordinate points to specify the areas, multiple coordinate points can be selected as regular deterrent points. These coordinate points can be automatically started according to the preset time of the system, and the system automatically selects the regular coordinate point closest to the deterrent target as part of the path. In this way, the effectiveness of each deterrent action can be improved.

[0028] Fig. 5 is a flowchart of the bird expelling method using the bird expelling system according to a preferred embodiment of the present application. Referring to Fig. 5, the flow of the bird expelling method includes:

[0029] Step S501: Start.

[0030] Step S502: Initialize strategy parameters. First, when the control device 204 starts to operate, it reads the user's settings first, for example, before use, the bird expelling system first takes panoramic screenshots of all the blocks. Then, the user can set the coordinates of several expelling positions in the first area according to the needs, and select the areas that need to be expelled (hot zones) and the areas that do not need to be expelled for each expelling position. These settings will be read by the control device 204 first in this initialization strategy parameter step.

[0031] Step S503: Direct the lens of the image capturing device toward the expelling position through the moving mechanism. Referring to Fig. 4, in this step S503, the control device 204 controls the moving mechanism 201 to aim the lens of the image capturing device 202 at the first expelling position.

[0032] Step S504: Perform first image capturing on the expelling position through the lens of the image capturing device to obtain a first image. The control device 204 controls the image capturing device 202 to perform image capturing on the expelling position to obtain the initial image of the position, that is, the image before bird expelling.

[0033] Step S505: judging the positions of the birds in the first image by the intelligent image judging model. The control device 204 judges the positions of the birds in the first image by the intelligent image judging model. Fig. 6 shows a schematic diagram of the set-up of the bird expelling system of a preferred embodiment of the present application. Referring to Fig. 6, in this embodiment, the user sets the need-expelling area 601 for each expelling position, and the intelligent image judging model only judges the birds in the need-expelling area 601. The birds outside the need-expelling area 601 are ignored by the intelligent image judging model.

[0034] Step S506: determining the moving trajectory of the moving mechanism and the expelling device according to the positions of the birds. Fig. 7 shows a schematic diagram of the aiming path of the moving mechanism 201 driving the expelling device 203 of the bird expelling system of a preferred embodiment of the present application. The control device 204 calculates the moving mode of the moving mechanism 201 according to the positions of the birds, i.e. calculates the moving mode of the moving mechanism 201 according to the aiming path 701 of the expelling device 203 required by the positions of the birds.

[0035] Step S507: determining an expelling strategy according to the expelling device. In this embodiment, one of the expelling devices 203, e.g. the laser module, is selected as the expelling device 203 according to the expelling devices 203 possessed, and the expelling strategy is determined according to the selected laser module and historical data. Figs. 8A-8D show schematic diagrams of the expelling strategies of the expelling device 203 of the bird expelling system of a preferred embodiment of the present application. Referring to Figs. 8A-8C, in this embodiment, the laser module is taken as an example, when the laser module moves to the position of the first bird according to the path of Fig. 7, it does not simply emit laser light, because if the bird does not see the laser light, the bird will not be driven away, therefore, in this embodiment, three laser light emitting strategies are provided in Figs. 8A-8C. For example, the strategy of Fig. 8A takes the cross-shaped path as an example; the strategy of Fig. 8B takes the zigzag-shaped path as an example; and the strategy of Fig. 8C takes the diamond-shaped path as an example.

[0036] Further, referring to Fig. 8D, for example, there is a small group of birds in the threshold area 801. The bird group in the small area 801 needs more frequent monitoring to improve the bird expelling effect. Therefore, in this embodiment, a threshold value of the area is set. If the activity range of the bird group is within the threshold value, it is defined as a "small-range bird group". The threshold value is a certain percentage of the monitoring area. In this embodiment, the expelling strategy is that the laser beam randomly moves in the scanning range and reflects when encountering the boundary, thereby improving the probability of the deterrent effect.

[0037] Step S508: execute the eviction strategy via the eviction device and the moving mechanism. After the eviction strategy is confirmed, the control device 204 starts to control the moving mechanism 201 and the eviction device 203 to execute the eviction strategy.

[0038] Step S509: after the eviction strategy is executed, perform second image capturing to obtain a second image. After the eviction strategy is performed for each bird gathering place, the control device 204 controls the image capturing device 202 to perform image capturing again at the eviction location to obtain an eviction result image of the location, i.e. an image when the bird eviction is performed.

[0039] Step S510: determine the second image and the first image via the intelligent image determination model to obtain an eviction effectiveness evaluation data. In this step S509, the control device 204 starts the local intelligent image determination model again to determine the first image before eviction and the second image after eviction, and obtains the eviction effectiveness evaluation data.

[0040] Step S511: store the eviction effectiveness evaluation data. The eviction effectiveness evaluation data is stored by the control device 204 as historical data.

[0041] Step S512: determine whether all eviction locations are performed. If not, the control device 204 returns to step S503 to perform again until all eviction locations are performed, and step S513 is performed.

[0042] Step S513: perform a strategy review algorithm according to the eviction effectiveness evaluation data to change the eviction strategy. In this embodiment, the strategy review algorithm can be, for example, comparing the historical data and the eviction effectiveness evaluation data to perform: calculating a current eviction rate of each of the plurality of locations; comparing the current eviction rate and the historical eviction rate of each of the plurality of locations to obtain an eviction trend of each of the plurality of locations; and changing the eviction strategy of each of the plurality of locations according to the historical eviction quantity of each of the plurality of locations, the current eviction quantity of each of the plurality of locations, and the eviction trend of each of the plurality of locations.

[0043] By the above strategy review algorithm, the expulsion strategy can be changed, such as changing the expulsion duration of the expulsion device; changing the limit detection time of the expulsion device; changing the expulsion device, such as replacing the laser, using a water gun, or ultrasound, etc.; changing the threat frequency of the expulsion device; and changing the threat specific coordinate position of the expulsion device. The above strategy review algorithm can be selectively performed by a general algorithm or by an artificial intelligence model. Generally speaking, the bird expulsion system adopts edge computing, and in the case of insufficient computing power and limited expulsion method and expulsion strategy, it is not necessary to use artificial intelligence, and a general algorithm is also a better option, which can reduce the power consumption of the bird expulsion system. The present application is not limited thereto. After step S513 is performed, return to step S503 to continue the next round of expulsion.

[0044] The above embodiment takes the roof of FIG. 7 as an example, and the environment of the roof is relatively simple and will not have obstacles or risks, so the operation of the bird expulsion system can be relatively simple. However, the bird expulsion method will change after the environment changes. FIG. 9 shows a flowchart of a bird expulsion method using the bird expulsion system according to a preferred embodiment of the present application. Referring to FIG. 9 and FIG. 5, the operation flow of the bird expulsion method further includes:

[0045] Step S901: determining obstacles by the first image through an intelligent image judgment model.

[0046] Step S902: determining non-bird organisms for risk avoidance by the first image through an intelligent image judgment model. For example, the environment set by the bird expulsion system is a cowshed, and when expelling birds, it is likely to injure the cows, such as the laser hitting the eyes of the cows eating feed causing the cows to be startled. Or the laser may also interfere with the workers. These are all need to be risk-avoided. Through the intelligent image judgment model, the above obstacles, non-bird organisms, etc. can be labeled out.

[0047] Step S903: determining the movement trajectory of the movement mechanism and the expulsion device according to the determined obstacles and risk objects when determining the movement trajectory. This step is the modified step S506.

[0048] Step S904: According to the expelling device, the judged obstacles and the risk objects, the expelling strategy is determined to avoid the judged obstacles and the risk objects. This step is the modified step S507. FIG. 10 shows a schematic diagram of the aiming path of the moving mechanism 201 driving the expelling device 203 of the bird expelling system according to a preferred embodiment of the present application. Referring to FIG. 10, 1001 represents the judged obstacles and the risk objects, and 1002 is the judged aiming path for avoiding the above-mentioned obstacles and risk objects 1001. In this way, these obstacles can be avoided. In addition, when the expelling strategy is executed, the laser can also be prevented from covering the judged obstacles and risk objects 1001.

[0049] Although the above-mentioned embodiments have considered the complex situations of risks and obstacles, in fact, some potential bird locations may also be included in the actual situation. The so-called potential bird location is a location where birds need to be expelled, but the image capturing device 202 cannot analyze the location where the birds exist. Generally, such locations are relatively far away from the lens of the image capturing device 202. In another preferred embodiment, step S506 can also determine the movement trajectory of the moving mechanism and the expelling device according to the potential bird location. In the above-mentioned embodiments, the potential bird location can also be a location where some areas are particularly prone to bird gathering without the need for the intelligent image judgment model to judge. Therefore, the present application is not limited thereto.

[0050] In summary, the preferred embodiment of the present application provides a novel bird expelling system which combines automation, intelligent image recognition and dynamic strategy adjustment to achieve efficient and targeted bird expelling. First of all, the system captures the bird activities in the specified area through the image capturing device, and accurately locates the bird location using the built-in edge operation intelligent image judgment model. Then, the system dynamically plans the aiming path and expelling strategy of the expelling device (such as laser, water gun, sound wave device, etc.) according to the location, number and hot zone range of the birds. This strategy not only can be executed in a targeted manner, but also can be self-optimized through strategy review algorithm according to the performance evaluation data (such as the expelling rate and expelling trend) after each expelling, to automatically adjust the expelling behavior next time, such as changing the expelling duration, detection frequency, or even replacing the expelling device. This design not only effectively avoids the energy and time waste caused by the traditional sequential scanning mode, but also provides a more flexible and adaptive solution according to different scenes and bird habits. Therefore, the bird expelling system of the present application can greatly improve the expelling efficiency while reducing the disturbance to the environment, and provides a smart solution for the bird damage problem in the fields of agriculture, public facilities, etc.

[0051] The specific embodiments set forth above are presented for the purpose of facilitating the understanding of the present application, and are not to be construed as limitations of the present application, as the spirit and scope of the present application are to be determined solely by the appended claims. Thus, the present application is to cover all modifications and variations of this application that come within the scope of the following claims and their equivalents.

[0052] [Symbol explanation] 100: moving mechanism 101: photographing device 102: laser emitting device 201: moving mechanism 202: image capturing device 203: expelling device 204: control device S501-S513, S901-S904: flow of bird expelling method of a preferred embodiment of the present application 601: expelling area 701: aiming path 801: threshold area 1001: judged obstacle and risk object 1002: judged aiming path

Claims

1. A bird expelling method, applicable to a bird expelling system, the bird expelling system comprising a control device, a moving mechanism, an image capturing device, an intelligent image judgment model, and an expelling device, the bird expelling method comprising: setting at least one expelling position via the control device; directing a lens of the image capturing device toward the at least one expelling position via the moving mechanism; performing a first image capturing for the at least one expelling position via the lens of the image capturing device to obtain a first image; judging a plurality of bird positions in the first image via the intelligent image judgment model; deciding a moving track of the moving mechanism and the expelling device according to the plurality of bird positions; deciding an expelling strategy according to the expelling device; executing the expelling strategy via the expelling device and the moving mechanism; performing a second image capturing to obtain a second image after executing the expelling strategy; judging the second image and the first image via the intelligent image judgment model to obtain an expelling effectiveness evaluation data; storing the expelling effectiveness evaluation data; and performing a strategy review algorithm according to the expelling effectiveness evaluation data to change the expelling strategy.

2. The bird repelling method as claimed in claim 1, wherein performing the first image capturing for the at least one expelling position via the lens of the image capturing device to obtain the first image, further comprising: judging an obstacle via the intelligent image judgment model by the first image.

3. The method of claim 2, wherein the bird repellent is applied to the surface of the structure in an amount of from about 0.1 to about 1.0 grams per square foot of surface area. deciding the moving track of the moving mechanism and the expelling device according to the plurality of bird positions, further comprising: deciding the moving track of the moving mechanism and the expelling device according to the judged obstacle.

4. The method of claim 2, wherein the bird repelling agent is applied to the surface of the structure in an amount of from about 0.1 to about 10 grams per square meter. deciding the moving track of the moving mechanism and the expelling device according to the plurality of bird positions, further comprising: deciding the expelling strategy according to the expelling device and the judged obstacle to avoid the judged obstacle.

5. The bird repelling method as claimed in claim 1, wherein The intelligent image judgment model is an edge computing intelligent image judgment model to reduce the operation delay.

6. The bird repelling method as claimed in claim 1, wherein The expelling device comprises at least one of: a water gun for water spraying expelling; a sound wave device for sound wave emitting; a laser module for emitting laser to expel; an air gun for physical expelling; and a sprayer for releasing chemical or aqueous expelling agent. The moving mechanism is a pan / tilt / zoom (PTZ) control mechanism.

7. The bird repelling method as claimed in claim 1, wherein deciding the moving track of the moving mechanism and the expelling device according to a potential bird position.

8. The bird repelling method as claimed in claim 1, wherein performing the first image capturing for the at least one expelling position via the lens of the image capturing device to obtain the first image, further comprising: judging a non-bird creature via the intelligent image judgment model by the first image to avoid the risk.

9. The bird repelling method as claimed in claim 1, wherein performing the strategy review algorithm according to the expelling effectiveness evaluation data to change the expelling strategy, comprising: ​ 10. The bird repelling method as claimed in claim 1, wherein ​ According to a historical data and the eviction effectiveness evaluation data, the strategy review algorithm is performed to change at least one of the following: change the eviction duration of the eviction device; change the limit detection time of the eviction device; change the eviction device; change the intimidation frequency of the eviction device; and change the intimidation specific coordinate position of the eviction device.

11. The method of repelling birds as recited in claim 1, wherein, According to the eviction effectiveness evaluation data, the strategy review algorithm is performed to change the eviction strategy, further comprising: comparing a historical data and the eviction effectiveness evaluation data to perform: calculating the current eviction rate of each of the multiple locations; comparing the current eviction rate and the historical eviction rate of each of the multiple locations to obtain the eviction trend of each of the multiple locations; and changing the eviction strategy of each of the multiple locations according to the historical eviction quantity of each of the multiple locations, the current eviction quantity of each of the multiple locations, and the eviction trend of each of the multiple locations.

12. The method of repelling birds as claimed in claim 1, wherein, when the at least one eviction location includes multiple eviction locations, after the step of storing the eviction effectiveness evaluation data is performed, the bird eviction method is executed again by the mobile mechanism to direct the lens of the image capturing device to the next eviction location until the multiple eviction locations are all executed, and then the strategy review algorithm is executed to change the eviction strategy.

13. A bird eviction system, comprising: a mobile mechanism; an image capturing device configured on the mobile mechanism; an eviction device configured on the mobile mechanism, wherein the image capturing device and the eviction device are substantially aligned in the same direction; and a control device coupled to the mobile mechanism, the image capturing device, and the eviction device, comprising an intelligent image judgment model, wherein the control device stores at least one eviction location set by a user; wherein the control device controls the mobile mechanism to direct a lens of the image capturing device to the at least one eviction location; wherein the control device controls the lens of the image capturing device to perform a first image capturing for the at least one eviction location to obtain a first image; wherein the intelligent image judgment model is used to determine multiple bird locations in the first image; wherein the control device determines to activate the eviction device and the movement trajectory of the mobile mechanism according to the multiple bird locations; wherein the control device determines an eviction strategy according to the eviction device; wherein the control device controls the eviction device and the mobile mechanism to execute the eviction strategy; wherein after the eviction strategy is executed, the control device controls the lens of the image capturing device to perform a second image capturing to obtain a second image; wherein the intelligent image judgment model is used to determine the second image and the first image to obtain eviction effectiveness evaluation data; wherein the control device stores the eviction effectiveness evaluation data; wherein the control device performs a strategy review algorithm to change the eviction strategy according to the eviction effectiveness evaluation data.

14. The bird eviction system as claimed in claim 13, further comprising: a remote cloud storage device coupled to the control device to store the eviction performance evaluation data as a historical data.

15. A bird repelling system as claimed in claim 13, wherein, The control device is configured to determine an obstacle by the first image through the intelligent image determination model, wherein, according to the determined obstacle, the control device determines to activate the eviction device and a moving trajectory of the moving mechanism.

16. A bird dispersing system as claimed in claim 15, wherein, The control device is further configured to determine the eviction strategy according to the eviction device and the determined obstacle to avoid the determined obstacle.

17. A bird repelling system as claimed in claim 13, wherein, The intelligent image determination model is an edge computing intelligent image determination model to reduce the operation delay.

18. A bird repelling system as claimed in claim 13, wherein, The eviction device includes at least one of: a water gun to perform water spray eviction; a sound wave device to perform sound wave emission; a laser module to emit laser to perform eviction; an air gun to perform physical eviction; and a sprayer to release chemical or water-based eviction agent. The moving mechanism is a pan / tilt / zoom (PTZ) control mechanism.

19. A bird repelling system as claimed in claim 13, wherein, The control device is further configured to determine to activate the eviction device and a moving trajectory of the moving mechanism according to a potential bird location.

20. A bird repelling system as claimed in claim 13, wherein, The control device is further configured to determine a non-bird species by the first image through the intelligent image determination model to perform risk avoidance.

21. A bird repelling system as claimed in claim 13, wherein, The control device changes the eviction strategy, including:

22. A bird repelling system as claimed in claim 13, wherein, performing the strategy review calculation according to a historical data and the eviction performance evaluation data to change at least one of: change the eviction duration of the eviction device; change the limited detection time of the eviction device; change the eviction device; change the intimidation frequency of the eviction device; and change the intimidation specific coordinate position of the eviction device. The control device performs the strategy review calculation to change the eviction strategy, further including:

23. A bird repelling system as claimed in claim 13, wherein, comparing a historical data and the eviction performance evaluation data to perform: calculate the current eviction rate of each of the plurality of locations; compare the current eviction rate and the historical eviction rate of each of the plurality of locations to obtain the eviction trend of each of the plurality of locations; and change the eviction strategy of each of the plurality of locations according to the historical eviction quantity of each of the plurality of locations, the current eviction quantity of each of the plurality of locations, and the eviction trend of each of the plurality of locations. When the at least one eviction location includes a plurality of eviction locations, 24. A bird repelling system as claimed in claim 13, wherein, after the step of storing the eviction performance evaluation data is performed, the method continues to perform the bird eviction method by the moving mechanism to direct the lens of the image capturing device to the next eviction location until the plurality of eviction locations are all performed, and then perform the strategy review calculation to change the eviction strategy. ​

Images