CN116630897B - Airport bird repellent intelligent auxiliary control system based on image recognition - Google Patents
Airport bird repellent intelligent auxiliary control system based on image recognition Download PDFInfo
- Publication number
- CN116630897B CN116630897B CN202310889401.6A CN202310889401A CN116630897B CN 116630897 B CN116630897 B CN 116630897B CN 202310889401 A CN202310889401 A CN 202310889401A CN 116630897 B CN116630897 B CN 116630897B
- Authority
- CN
- China
- Prior art keywords
- birds
- bird
- suspected
- frequency
- fanning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000005871 repellent Substances 0.000 title claims abstract description 34
- 230000002940 repellent Effects 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 claims description 24
- 210000003128 head Anatomy 0.000 claims description 22
- 230000009471 action Effects 0.000 claims description 17
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 238000001514 detection method Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000013450 outlier detection Methods 0.000 claims description 4
- 230000001846 repelling effect Effects 0.000 description 13
- 230000002159 abnormal effect Effects 0.000 description 8
- 230000000694 effects Effects 0.000 description 7
- 230000006399 behavior Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 241001314440 Triphora trianthophoros Species 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01M—CATCHING, TRAPPING OR SCARING OF ANIMALS; APPARATUS FOR THE DESTRUCTION OF NOXIOUS ANIMALS OR NOXIOUS PLANTS
- A01M29/00—Scaring or repelling devices, e.g. bird-scaring apparatus
- A01M29/06—Scaring or repelling devices, e.g. bird-scaring apparatus using visual means, e.g. scarecrows, moving elements, specific shapes, patterns or the like
- A01M29/10—Scaring or repelling devices, e.g. bird-scaring apparatus using visual means, e.g. scarecrows, moving elements, specific shapes, patterns or the like using light sources, e.g. lasers or flashing lights
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01M—CATCHING, TRAPPING OR SCARING OF ANIMALS; APPARATUS FOR THE DESTRUCTION OF NOXIOUS ANIMALS OR NOXIOUS PLANTS
- A01M29/00—Scaring or repelling devices, e.g. bird-scaring apparatus
- A01M29/16—Scaring or repelling devices, e.g. bird-scaring apparatus using sound waves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention relates to the technical field of image data processing, in particular to an airport bird repellent intelligent auxiliary control system based on image recognition, which comprises the following components: dividing birds in a flight period into suspected birds and normal birds; obtaining the fanning frequency and the warning frequency of suspected birds; obtaining the credibility of the suspected birds according to the difference between the fanning frequency of the suspected birds and the fanning frequency of the normal birds and the difference between the warning frequency of the suspected birds and the warning frequency of the normal birds, and obtaining the unreliable birds in the suspected birds according to the credibility of the suspected birds; the bird repellent control is performed based on the repellent priority of each bird obtained based on the number of times that an untrusted bird appears in each bird and the total number of occurrences of each bird. The invention avoids the situation that birds stay for a long time to attract other birds to stay and the bird trunk with expelling resistance does not walk, ensures higher expelling efficiency and avoids potential safety hazard.
Description
Technical Field
The invention relates to the technical field of image data, in particular to an airport bird repellent intelligent auxiliary control system based on image recognition.
Background
Birds moving near airports pose a threat to aviation security by potentially flying into the engines causing space accidents. The bird repellent system can effectively reduce the activity of birds near an airport to ensure aviation safety, and generally uses different devices to monitor birds such as radar, infrared rays, cameras and the like, wherein the cameras have the functions of providing high-resolution images and monitoring in real time and can be used for monitoring the activity condition of the birds.
Using a camera device may be difficult to monitor birds in a more complex environment, as birds may be out of the monitoring range of the camera system, resulting in an inability to determine the bird's repellent effect; some birds are set up in the airport for a long time, so that the expelling effect cannot be confirmed, and even birds which are difficult to expel or have expelling resistance can attract other birds to stay, so that birds which are prevented from expelling are omitted, and potential safety hazards are caused when the birds are set up in the airport.
In order to solve the problem of bird driving carry-over, birds with low credibility are screened out through the camera system shooting, and the audio frequency and the light irradiation frequency of the birds are replaced at regular time according to the bird varieties with low credibility in each month, so that the effect of driving the birds and protecting airport safety is achieved.
Disclosure of Invention
The invention provides an airport bird-repellent intelligent auxiliary control system based on image recognition, which aims to solve the existing problems.
The airport bird-repellent intelligent auxiliary control system based on image recognition adopts the following technical scheme:
the embodiment of the invention provides an airport bird repellent intelligent auxiliary control system based on image recognition, which comprises the following modules:
the data acquisition module is used for acquiring video data of the birds to be repelled;
the bird state detection module is used for obtaining an alert period and a flight period of birds according to images in the video data and dividing the birds in the flight period into suspected birds and normal birds; obtaining the fanning frequency and the warning frequency of suspected birds;
the bird credibility calculation module is used for obtaining the credibility of the suspected birds according to the difference between the fanning frequency of the suspected birds and the fanning frequency of the normal birds and the difference between the warning frequency of the suspected birds and the warning frequency of the normal birds, and obtaining the unreliable birds in the suspected birds according to the credibility of the suspected birds;
the bird repelling control module is used for obtaining the repelling priority of each bird according to the occurrence times of the unreliable birds in each bird and the occurrence total number of each bird, and performing bird repelling control according to the repelling priority of each bird.
Preferably, the method for obtaining the warning period and the flight period of the birds according to the images in the video data comprises the following specific steps:
obtaining the variety of birds on each frame of video image in the video data and the bounding box of each bird by using a target detection algorithm, wherein the Euclidean distance between the center points of the bounding boxes of the same bird in two adjacent frames of images is recorded as the flight distance of the bird in the two adjacent frames of images; after the driving behavior is generated, the flying distance obtained by every two adjacent frames forms a flying distance sequence for the same bird; presetting a distance threshold value th1, and recording a time period corresponding to a flight distance continuously larger than th1 as a flight time period in a flight distance sequence; and (3) recording a period corresponding to the flight distance which is continuously less than or equal to th1 as an alert period in the flight speed sequence.
Preferably, the method for classifying birds in a flight period into suspected birds and normal birds comprises the following specific steps:
acquiring center points of bounding boxes corresponding to all birds in a flight period, inputting the center points of all bounding boxes into an outlier detection algorithm to obtain outliers, marking the birds corresponding to the outliers as suspected birds, and marking the birds except the suspected birds as normal birds in all birds in the flight period.
Preferably, the step of obtaining the fanning frequency includes:
acquiring key points of birds, wherein the key points comprise key points at the head and neck junction of the birds, key points of eyes of the birds, key points of wing tips and key points at the tail parts of the birds; the middle points of all key points of birds are marked as points A, the key points of the tips of wings are respectively marked as points B, the key points at the tail parts are marked as points C, and the included angles between line segments where A and B are positioned and line segments where A and C are positioned are obtained and marked as angles a; taking the time corresponding to each frame of image as an abscissa, taking the angle a of the same bird in each frame of image as an ordinate, obtaining a curve, recording the absolute value of the difference value of the abscissas corresponding to two adjacent maximum points on the curve as a fanning period length, recording the average value of fanning period lengths corresponding to all the two adjacent maximum points on the curve as the average fanning period of the bird, and recording the reciprocal of the average fanning period of the bird as the fanning frequency of the bird.
Preferably, the step of obtaining the alert frequency includes:
according to the key points of birds, head raising actions and head twisting actions of the birds are obtained, the occurrence times of the head raising actions and the head twisting actions are counted, and the warning frequency is recorded.
Preferably, the method for obtaining the credibility of the suspected bird according to the difference between the fanning frequency of the suspected bird and the fanning frequency of the normal bird and the difference between the warning frequency of the suspected bird and the warning frequency of the normal bird comprises the following specific steps:
firstly, the warning difference and the fanning difference of each suspected bird are obtained, and the calculation formula is as follows:
wherein ,mean value of warning frequency of all normal birds, < ->Mean value of the fanning frequency of all normal birds, < >>Indicating the frequency of vigilance of each suspected bird, +.>Representing the fanning frequency of each suspected bird; />Indicating the vigilance difference of each suspected bird, +.>Representing the fanning difference for each suspected bird;
the calculation formula of the credibility of suspected birds is as follows:
wherein
Is the credibility of each suspected bird, < +.>Is a judgment value for the fanning frequency and the warning frequency.
Preferably, the untrusted birds are suspected birds having a confidence level less than a preset threshold.
Preferably, the method for obtaining the driving priority of each bird according to the number of occurrence times of the untrusted birds in each bird and the total occurrence number of each bird comprises the following specific steps:
counting the occurrence times of the unreliable birds in any bird, and numbering all birds from small to large according to the sequence of the occurrence times of the unreliable birds to obtain a first number corresponding to each bird;
according to the sequence from big to small of the total number of each bird, numbering all birds from small to big again to obtain a second number corresponding to each bird;
the method for calculating the driving priority of each bird comprises the following steps:
wherein ,、/>a first number and a second number corresponding to each bird are represented, and e represents a natural constant; DP represents the driving priority for each bird.
Preferably, the bird repellent control is performed according to the repellent priority of each bird, and the specific steps are as follows:
and obtaining a preset number of birds with the largest driving priority, recording the birds as target birds, obtaining driving signals of the target birds, and performing driving control according to the driving signals of the target birds.
The technical scheme of the invention has the beneficial effects that: because birds with high expelling priority have expelling resistance, the invention analyzes the credibility of the birds by analyzing the flight and warning behavior characteristics of the birds, thereby obtaining the expelling priority of the birds, ensuring that the birds which are difficult to expel are expelled in a targeted way every day, avoiding the conditions that the birds stay for a long time and attract other birds to stay and the bodies of the birds with expelling resistance do not walk, ensuring higher expelling efficiency and avoiding potential safety hazards.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a frame structure diagram of an airport bird repellent intelligent auxiliary control system based on image recognition of the present invention;
Detailed Description
In order to further describe the technical means and effects adopted by the airport bird repellent intelligent auxiliary control system for achieving the preset aim of the invention, the following detailed description is given of specific implementation, structure, characteristics and effects thereof of the airport bird repellent intelligent auxiliary control system based on image recognition according to the invention by combining the attached drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The specific scheme of the airport bird repellent intelligent auxiliary control system based on image recognition provided by the invention is specifically described below with reference to the accompanying drawings.
Referring to fig. 1, a frame structure diagram of an airport bird repellent intelligent auxiliary control system based on image recognition according to an embodiment of the present invention is shown, where the system includes the following modules:
and a data acquisition module: video data of the bird repelled shot by the camera are collected.
Monitoring cameras are installed in high-risk areas of airports, and are mainly installed in high-risk areas where birds gather, such as lawns and woodlands of airports, control towers and tarmac, water sources near airports, farms and the like. The reaction of the birds after being driven is shot through the camera device on the ground.
All birds repel bird types and corresponding numbers in days of day 1 month of extraction, and the data is stored as a bird population. And (3) storing the shot image data locally, and carrying out enhanced denoising pretreatment on the acquired video image to obtain a reaction image of birds. A group of video data is stored every half month, and the bird characteristics are analyzed through the video data, so that a basis is provided for adjusting bird repelling strategies in the future.
Bird status detection module: and obtaining the fanning frequency and the warning frequency of birds according to each frame of image in the video data.
Birds that develop tolerance to common bird repellents may react relatively poorly to the repellents or be completely disregarded; they may not escape the airport as quickly as normal intolerant birds, but stay around the airport for a longer period of time. Normal intolerant birds are more easily frightened and repelled by common repelling means. It appears that a class of birds that are less aggressive to escape are present on the image, they are more prone to avoid the active area of the bird repellent device than to leave the airport, and the speed and behaviour of flight are distinguishable from normal birds. Therefore, birds with slow flying speeds are identified through motion detection, and then according to two periods of reaction of the birds, an alert action period generated after being driven and a flying period after taking off are started respectively. In the warning stage, the bird motion amplitude is smaller and quieter; birds in the flight stage have higher flight speed and relatively obvious action and motion amplitude.
The specific acquisition method of the warning period and the flight period is as follows:
and detecting the variety of birds in each frame of video image by using a target detection algorithm based on deep learning. The target detection algorithm used in this embodiment refers to obtaining a bounding box of birds and a variety of birds in an image by using a YOLOV4 network, where the YOLOV4 network is a commonly known target detection technology, and specific network structures and training methods are not described in detail in this embodiment.
The Euclidean distance between the center points of bounding boxes in two adjacent frames of images of the same bird is recorded as the flight distance of the bird in the two adjacent frames of images. After the driving action is generated, the obtained flight distance of every two adjacent frames forms a flight distance sequence for the same bird. Setting a distance threshold th1, and if 2 or more continuous flight distances are greater than th1 in the flight distance sequence, recording the corresponding time periods of the continuous flight distances as flight time periods; if the flight distances of 2 or more continuous flight speed sequences are smaller than or equal to th1, the time periods corresponding to the flight distances which are continuously distributed are recorded as warning time periods;
in this embodiment, th1=2.0 is taken as an example, and other values may be set in other implementations, and this embodiment is not particularly limited. Further, the repelling action in this embodiment includes emitting light signals and sound signals having bird repelling function.
Acquiring all birds in a flight period, acquiring the center points of surrounding frames corresponding to each bird, inputting the center points of the surrounding frames of the birds into an outlier detection algorithm to acquire the center points of all outliers, namely, the outliers for short, wherein the birds corresponding to the outliers are marked as suspected birds, and the birds except the suspected birds in the flight period are marked as normal birds. Wherein the outlier detection algorithm used in the present embodiment is the CLOF algorithm.
For single-variety birds to be detected, two birds, namely abnormal birds and normal birds, exist at the same time, and when the abnormal birds fly, the flapping frequency of wings is obviously abnormal to that of the normal birds. Therefore, the difference value of wing flapping frequency can be compared with that of normal homologous birds to describe the abnormality degree of the birds in the flight period. Thus, there is a need for confidence detection for each individual suspected bird in an image.
Before credibility detection, the position and the gesture of birds need to be tracked by a tracking algorithm so as to conveniently acquire the motion gesture of the birds. When tracking the position and the gesture of birds, a plurality of key points of the birds need to be extracted for tracking, and a skeleton model can be constructed by the key points, so that the movement characteristics of the birds can be conveniently extracted. The tips of the wings of the birds are selected as key points of the two wings, the center of the junction of the head and neck of the birds and eyes of the birds are selected as two key points of the head, and finally the tail of the birds is selected as the last key point.
It should be noted that, in this embodiment, two methods for obtaining key points are provided, where the first method is to obtain all the corner points in the bounding box of the bird in the collected image by using SIFT corner detection algorithm, and then obtain the corner points belonging to the wing tip, the head and neck junction, the eyes and the tail in these corner points by using random forest algorithm, and use the corner points as key points. The second approach is to use a keypoint detection network, such as the HRNet network, to obtain the above-described keypoints on birds. The mean value of the two methods is known in the art, and the second method can avoid the shielding problem, so the second method is used in this embodiment, and the specific process is not repeated.
The flapping frequency of the wing skeleton movement can be calculated by detecting the flapping period of the wing skeleton in the flying stage after the driving action. The calculation method comprises the following steps:
in each frame of video image in the flight period, the middle points of all key points on the bird skeleton are marked as points A, the key points of the wing tips are respectively marked as points B, the key points at the tail are marked as points C, and the included angles between the line segments where A and B are located and the line segments where A and C are located are marked as angles a. Namely, each bird corresponds to an angle a in each frame of image; and taking the time corresponding to each frame of image as an abscissa, taking the angle a of the same bird in each frame of image as an ordinate, obtaining a curve for representing the change rule of the flying gesture of the bird, and recording the absolute value of the difference value of the abscissas corresponding to two adjacent maximum points on the curve as a fanning period length, wherein the average value of the fanning period lengths corresponding to all the two adjacent maximum points on the curve is recorded as the average fanning period of the bird, and the reciprocal of the average fanning period of the bird is recorded as the fanning frequency of the bird.
It should be noted that, there are two key points B of the wing tip, and two fanning frequencies can be calculated at this time, and the average of the two fanning frequencies is the final result of the fanning frequency of the bird in this embodiment.
The frequency of fanning of each suspected bird during the flight phase is thus obtained.
Through the method, the fanning frequency of birds in flying can be obtained, but the birds are classified and analyzed only by the characteristic, so that misjudgment can occur, and the situation that the birds get ill or the old fall behind the bird group can occur; the analysis should also continue during the alert period.
For suspected birds with single warning period, two birds, namely abnormal birds and normal birds, exist at the same time, the abnormal birds should have the condition that the warning frequency is obviously abnormal to the normal birds, and the warning frequency of the abnormal birds is generally lower than that of the normal birds. The birds with fanning frequency are extracted and analyzed through motion tracking, and the images of the warning when the birds are on the ground are analyzed, so that the two parameters are ensured to be stored on the same bird, and the birds can be conveniently classified later.
Birds with abnormal birds resistant to common bird repelling means are calm, the number of times of warning behaviors of the birds is less than that of normal birds, the head of the birds is lifted in a warning period, and the number of times of observing the head in four weeks by rotating the head left and right is the warning frequency. The head is lifted once, the head is twisted once in one direction, and all times are accumulated to be warning frequency in a warning period.
In this embodiment, the method for acquiring the bird head raising motion and the head twisting motion includes: all key points on the wing skeleton of the bird are input into a full-link network, and the network outputs action types corresponding to the wing skeleton, wherein the action types comprise head lifting actions, head twisting actions and other actions. The full link network technology is a common known technology, and specific network results, training and usage methods thereof are not described in detail in this embodiment. Other methods may be used in other embodiments to obtain the head-up motion and the head-twisting motion, for example, using a machine learning algorithm, which is not described in detail in this embodiment.
The fanning frequency and the warning frequency of the suspected birds are obtained so far and are used for representing gesture movement characteristics of the suspected birds in the flight period and the warning period. And similarly, obtaining the fanning frequency and the warning frequency of the normal birds.
Bird credibility calculation module: obtaining the credibility of birds according to the fanning frequency and the warning frequency of the birds, and obtaining the unreliable birds according to the credibility.
Firstly, analyzing the difference of gesture motion characteristics of suspected birds and normal birds in a flight period and a warning period, and enabling the suspected birds to be:
wherein Mean value of warning frequency of all normal birds, < ->Mean value of the fanning frequency of all normal birds, < >>Indicating the frequency of vigilance of each suspected bird, +.>Representing the frequency of fanning of each suspected bird. />Representing police of each suspected birdAbstinence of different, ->Indicating the fanning difference for each suspected bird.
The method for the credibility of suspected birds is as follows:
wherein :
is the credibility of each suspected bird, < +.>Is a judgment value for the fanning frequency and the warning frequency.
So far, the credibility of each suspected bird is obtained.
This embodiment will be described by taking th2=0.05 and th3=0.07 as examples.
When the credibility is 1, the warning difference of suspected birds is indicatedAnd fanning difference->Satisfies the condition and />At this time, the warning difference and the fanning difference of the suspected birds are compared with those of other normal birdsIf the difference is not large, then the suspected birds with the credibility of 1 are recorded as credible birds.
When suspected birds are alert to differencesAnd fanning difference->Does not satisfy the condition->Andwhen the difference in vigilance or fanning of the suspected birds is larger than the difference in vigilance or fanning of other normal birds, the suspected birds with the credibility less than 1 are marked as the unreliable birds.
Bird repelling control module: and obtaining bird repelling priority according to the occurrence times of the untrusted birds, and carrying out bird repelling according to the bird repelling priority.
For any bird, the number of occurrence times of the untrusted birds in the bird is obtained, and all birds are numbered from small to large according to the sequence of the occurrence times of the untrusted birds, so that the first number corresponding to each bird is obtained. According to the sequence from big to small of the total number of each bird, numbering all birds from small to big again to obtain a second number corresponding to each bird;
the method for calculating the driving priority of each bird comprises the following steps:
、/>a first number and a second number corresponding to each bird are represented, and e represents a natural constant; DP represents each birdWhen the first number and the second number corresponding to each bird are larger, the total number corresponding to birds and the number of occurrence times of the untrusted birds are relatively more, so that the birds need to be driven preferentially.
The above analysis based on the video data of one month thus obtained the driving priority of each bird in the previous month. Then on each day of the next month, three birds with the highest driving priority are obtained, driving audio sounds, flash lamp frequencies and brightness corresponding to the three birds are obtained, and then the unmanned aerial vehicle is used for emitting corresponding sounds and light signals to drive.
Because the birds with high driving priority have driving resistance, the birds with high driving priority are guaranteed to be driven in a targeted manner every day, the birds are prevented from staying for a long time and other birds are prevented from staying, the higher driving efficiency is guaranteed, and potential safety hazards are avoided.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (6)
1. Airport bird repellent intelligent auxiliary control system based on image recognition, which is characterized by comprising the following modules:
the data acquisition module is used for acquiring video data of the birds to be repelled;
the bird state detection module is used for obtaining an alert period and a flight period of birds according to images in the video data and dividing the birds in the flight period into suspected birds and normal birds; obtaining the fanning frequency and the warning frequency of suspected birds;
the bird credibility calculation module is used for obtaining the credibility of the suspected birds according to the difference between the fanning frequency of the suspected birds and the fanning frequency of the normal birds and the difference between the warning frequency of the suspected birds and the warning frequency of the normal birds, and obtaining the unreliable birds in the suspected birds according to the credibility of the suspected birds;
the bird repellent control module is used for obtaining the repellent priority of each bird according to the occurrence times of the unreliable birds in each bird and the occurrence total number of each bird, and performing bird repellent control according to the repellent priority of each bird;
the method for obtaining the warning period and the flight period of birds according to the images in the video data comprises the following specific steps:
obtaining the variety of birds on each frame of video image in the video data and the bounding box of each bird by using a target detection algorithm, wherein the Euclidean distance between the center points of the bounding boxes of the same bird in two adjacent frames of images is recorded as the flight distance of the bird in the two adjacent frames of images; after the driving behavior is generated, the flying distance obtained by every two adjacent frames forms a flying distance sequence for the same bird; presetting a distance threshold value th1, and recording a time period corresponding to a flight distance continuously larger than th1 as a flight time period in a flight distance sequence; in the flying speed sequence, a period corresponding to the flying distance continuously smaller than or equal to th1 is recorded as an alert period;
the method for classifying birds in a flight period into suspected birds and normal birds comprises the following specific steps:
acquiring center points of surrounding frames corresponding to all birds in a flight period, inputting the center points of all the surrounding frames into an outlier detection algorithm to obtain outliers, marking the birds corresponding to the outliers as suspected birds, and marking the birds except the suspected birds as normal birds in all the birds in the flight period;
the method for obtaining the credibility of the suspected birds according to the difference between the fanning frequency of the suspected birds and the fanning frequency of the normal birds and the difference between the warning frequency of the suspected birds and the warning frequency of the normal birds comprises the following specific steps:
firstly, the warning difference and the fanning difference of each suspected bird are obtained, and the calculation formula is as follows:
wherein ,mean value of warning frequency of all normal birds, < ->Mean value of the fanning frequency of all normal birds, < >>Indicating the frequency of vigilance of each suspected bird, +.>Representing the fanning frequency of each suspected bird; />Indicating the vigilance difference of each suspected bird, +.>Representing the fanning difference for each suspected bird;
the calculation formula of the credibility of suspected birds is as follows:
wherein
Is the credibility of each suspected bird, < +.>Is a judgment value for the fanning frequency and the warning frequency.
2. The intelligent auxiliary control system for airport bird repellent based on image recognition of claim 1, wherein the step of obtaining the fanning frequency comprises the following steps:
acquiring key points of birds, wherein the key points comprise key points at the head and neck junction of the birds, key points of eyes of the birds, key points of wing tips and key points at the tail parts of the birds; the middle points of all key points of birds are marked as points A, the key points of the tips of wings are respectively marked as points B, the key points at the tail parts are marked as points C, and the included angles between line segments where A and B are positioned and line segments where A and C are positioned are obtained and marked as angles a; taking the time corresponding to each frame of image as an abscissa, taking the angle a of the same bird in each frame of image as an ordinate, obtaining a curve, recording the absolute value of the difference value of the abscissas corresponding to two adjacent maximum points on the curve as a fanning period length, recording the average value of fanning period lengths corresponding to all the two adjacent maximum points on the curve as the average fanning period of the bird, and recording the reciprocal of the average fanning period of the bird as the fanning frequency of the bird.
3. The intelligent auxiliary control system for airport bird repellent based on image recognition according to claim 2, wherein the step of obtaining the warning frequency is as follows:
according to the key points of birds, head raising actions and head twisting actions of the birds are obtained, the occurrence times of the head raising actions and the head twisting actions are counted, and the warning frequency is recorded.
4. The image recognition-based intelligent auxiliary control system for airport bird repellent according to claim 1, wherein the following is adopted
An untrusted bird is a suspected bird having a confidence level less than a preset threshold.
5. The intelligent auxiliary control system for bird repellent in an airport based on image recognition according to claim 1, wherein the method for obtaining the repellent priority of each bird according to the number of occurrence of untrusted birds in each bird and the total number of occurrence of each bird comprises the following specific steps:
counting the occurrence times of the unreliable birds in any bird, and numbering all birds from small to large according to the sequence of the occurrence times of the unreliable birds to obtain a first number corresponding to each bird;
according to the sequence from big to small of the total number of each bird, numbering all birds from small to big again to obtain a second number corresponding to each bird;
the method for calculating the driving priority of each bird comprises the following steps:
wherein ,、/>a first number and a second number corresponding to each bird are represented, and e represents a natural constant; DP represents the driving priority for each bird.
6. The intelligent auxiliary control system for bird repellent in an airport based on image recognition according to claim 1, wherein the bird repellent control according to the repellent priority of each bird comprises the following specific steps:
and obtaining a preset number of birds with the largest driving priority, recording the birds as target birds, obtaining driving signals of the target birds, and performing driving control according to the driving signals of the target birds.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310889401.6A CN116630897B (en) | 2023-07-20 | 2023-07-20 | Airport bird repellent intelligent auxiliary control system based on image recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310889401.6A CN116630897B (en) | 2023-07-20 | 2023-07-20 | Airport bird repellent intelligent auxiliary control system based on image recognition |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116630897A CN116630897A (en) | 2023-08-22 |
CN116630897B true CN116630897B (en) | 2023-09-22 |
Family
ID=87621529
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310889401.6A Active CN116630897B (en) | 2023-07-20 | 2023-07-20 | Airport bird repellent intelligent auxiliary control system based on image recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116630897B (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106846371A (en) * | 2016-12-23 | 2017-06-13 | 深圳大学 | A kind of birds wing vibration frequency measuring method and device based on video image |
CN106954049A (en) * | 2017-05-13 | 2017-07-14 | 西安费斯达自动化工程有限公司 | The airport birds information acquisition method of panorama and precise image tracking system |
CN112056302A (en) * | 2020-10-10 | 2020-12-11 | 苏州电力设计研究院有限公司 | Bird repelling device, system and method for main transformer cross arm |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106462978A (en) * | 2014-05-07 | 2017-02-22 | 日本电气株式会社 | Object detection device, object detection method, and object detection system |
-
2023
- 2023-07-20 CN CN202310889401.6A patent/CN116630897B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106846371A (en) * | 2016-12-23 | 2017-06-13 | 深圳大学 | A kind of birds wing vibration frequency measuring method and device based on video image |
CN106954049A (en) * | 2017-05-13 | 2017-07-14 | 西安费斯达自动化工程有限公司 | The airport birds information acquisition method of panorama and precise image tracking system |
CN112056302A (en) * | 2020-10-10 | 2020-12-11 | 苏州电力设计研究院有限公司 | Bird repelling device, system and method for main transformer cross arm |
Non-Patent Citations (1)
Title |
---|
基于支持向量机的机场智能驱鸟决策;陈唯实;闫军;张洁;李敬;;北京航空航天大学学报(07);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN116630897A (en) | 2023-08-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11751560B2 (en) | Imaging array for bird or bat detection and identification | |
CN108710126B (en) | Automatic target detection and eviction method and system | |
US11544490B2 (en) | Avian detection systems and methods | |
US20130050400A1 (en) | Arrangement and Method to Prevent a Collision of a Flying Animal with a Wind Turbine | |
Sugumar et al. | An improved real time image detection system for elephant intrusion along the forest border areas | |
CN113034397B (en) | High-altitude parabolic detection method capable of realizing automatic tracing of self-adaptive track in multiple environments in real time | |
US20220361471A1 (en) | Intelligent insect trap and monitoring system | |
Hill et al. | Of birds, blades and barriers: Detecting and analysing mass migration events at alpha ventus | |
CN113569644B (en) | Airport bird target detection method based on machine vision | |
CN116630897B (en) | Airport bird repellent intelligent auxiliary control system based on image recognition | |
Bhusal et al. | Bird deterrence in a vineyard using an unmanned aerial system (uas) | |
Shah et al. | Use of Deep Learning Applications for Drone Technology | |
Elsayed et al. | Visual drone detection in static complex environment | |
Zubkov et al. | Investigation of the neural networks effectiveness in recognizing moving drones | |
Bhusal et al. | Bird detection, tracking and counting in wine grapes | |
Avramović et al. | Automatic monitoring of honeybees’ activity outside of the hive from UHD video | |
CN109669180A (en) | A kind of continuous wave radar unmanned plane detection method | |
Sakhare et al. | Intelligent Video Surveillance System for Indian Farms. | |
EP3779925A1 (en) | Camera-based angle tracking of swarms for collision avoidance | |
So et al. | Development and Efficacy of Intelligent Laser Bird Repeller with Ptz Camera and Caffe Framework | |
Sabhahit et al. | An animal intrusion detection and repellent system for farm safety using IoT | |
Mirzaei et al. | Regularized logistic regression based classification for infrared images | |
CN115965990A (en) | Unmanned aerial vehicle bird repelling method and system and readable storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |