CN116189018A - Method, equipment, system and storage medium for counting wild aerial cattle groups - Google Patents

Method, equipment, system and storage medium for counting wild aerial cattle groups Download PDF

Info

Publication number
CN116189018A
CN116189018A CN202310021725.8A CN202310021725A CN116189018A CN 116189018 A CN116189018 A CN 116189018A CN 202310021725 A CN202310021725 A CN 202310021725A CN 116189018 A CN116189018 A CN 116189018A
Authority
CN
China
Prior art keywords
image
frame
cattle
aerial
ith
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.)
Pending
Application number
CN202310021725.8A
Other languages
Chinese (zh)
Inventor
张军国
王远
李柏灿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Forestry University
Original Assignee
Beijing Forestry University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Forestry University filed Critical Beijing Forestry University
Priority to CN202310021725.8A priority Critical patent/CN116189018A/en
Publication of CN116189018A publication Critical patent/CN116189018A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/17—Terrestrial scenes taken from planes or by drones
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • 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
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • 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
    • 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
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
    • Y02A40/70—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in livestock or poultry

Abstract

The embodiment of the invention discloses a method, equipment, a system and a storage medium for counting wild aerial photography flocks, wherein the method comprises the following steps: preprocessing field aerial videos; performing feature detection and extraction on the preprocessed field aerial video by using a YOLOv4 target detection model to obtain the number of cattle group identification in each frame of image and individual features of each cattle, wherein the individual features comprise position features, attribute features and score features; inputting the position feature and the attribute feature of each cow in the front and back frame images into a SuperGlue feature matching model, and acquiring the feature corresponding relation of the cow groups in the front and back frame images and the superposition quantity of the cow groups in the back frame images and the cow groups in the front frame images; and calculating to obtain the number of the flocks in the ith frame of image according to the number of the flocks in the ith frame of image, the number of the flocks in the ith-1 frame of image and the corresponding superposition number, wherein i is a positive integer, and i is more than or equal to 2. The method realizes accurate matching and flock counting of flock targets between frames.

Description

Method, equipment, system and storage medium for counting wild aerial cattle groups
Technical Field
The invention relates to the technical field of wild animal image processing, in particular to a method, equipment and system for counting wild aerial cattle groups and a storage medium.
Background
The animal husbandry of the grassland pasture has important positions in the middle of the resources of China, relates to domestic and folk life, and needs an accurate, quick, economical and practical animal protection network system, thereby improving the development strength and progress of the grassland animal husbandry. In the aspect of development of grassland animal husbandry, for protecting grassland livestock and taking cattle groups as an example, basic work which needs to be done is to quickly obtain important information such as the number and activity track of wild cattle groups and survival conditions, and meanwhile, the grassland animal information is essential for many applications, including protection of wild endangered species living together, tracking of invasive species and monitoring of forest and wild animals and plants. Traditional grassland pasture livestock population monitoring comprises remote sensing information shooting, label mark monitoring and the like. Such population surveys can be cost prohibitive, logistically challenging and time consuming, especially in large and inconvenient areas, and are subject to uncertainty and risk; the fixed monitoring and the real-time monitoring in the animal farm consume a great deal of manual energy, and the working efficiency is very low. The unmanned aerial vehicle is an unmanned aerial vehicle capable of completing control flight through setting autonomous flight or signal receiving, and has the characteristics of low cost, flexibility, convenience, simple structure, simple operation, small volume, free take-off and landing and the like. The machine body is provided with a visual sensor, a temperature and humidity sensor and a laser radar sensor, so that the machine body is applied to monitoring of a moving target. The method can effectively overcome the defects of high cost, poor real-time performance, invalid repeated data acquisition and the like caused by the traditional monitoring means.
The research finds that the deep neural network is an effective image detection method at present, and simultaneously in a multi-target detection algorithm for the wild cattle group, the deep neural network is used as a basic research of a monitoring target feature extraction part, and the characteristics of each monitoring target in the population are provided for the graph neural network adopted by frame-by-frame target matching. However, in the unmanned aerial vehicle aerial photography field animal population process, aerial photography movement of the unmanned aerial vehicle is quicker when relative to ground target movement, so that displacement of the flock target between frames can be regarded as having smaller displacement. Because of the non-directional displacement of the flight in the unmanned aerial vehicle aerial photographing process, the conventional algorithm has the serious problems of obvious target loss and shielding false detection based on the existing real conditions.
Disclosure of Invention
Aiming at the technical defects in the prior art, the embodiment of the invention aims to provide a method, equipment, a system and a storage medium for counting the flock in the wild aerial photography, which realize the accurate matching of the flock targets and the flock counting between frames and solve the problems of target loss, shielding false detection and the like caused by the movement of the flock and the speed change of an unmanned aerial vehicle during the flight.
In order to achieve the above object, in a first aspect, an embodiment of the present invention provides a method for tracking and counting wild aerial cattle, including:
preprocessing field aerial videos;
performing feature detection and extraction on the preprocessed field aerial video by using a YOLOv4 target detection model to obtain the number of cattle group identification in each frame of image and individual features of each cattle, wherein the individual features comprise position features, attribute features and score features;
inputting the position feature and the attribute feature of each cow in the front and back frame images into a SuperGlue feature matching model, and acquiring the feature corresponding relation of the cow groups in the front and back frame images and the superposition quantity of the cow groups in the back frame images and the cow groups in the front frame images;
and calculating to obtain the number of the flocks in the ith frame of image according to the number of the flocks in the ith frame of image, the number of the flocks in the ith-1 frame of image and the corresponding superposition number, wherein i is a positive integer, and i is more than or equal to 2.
Further, the preprocessing of the outdoor aerial video includes:
removing aerial video clips and interference video clips lost by targets, and reserving effective video clips;
carrying out slow, quick, rotation, frame skip and static treatment on the effective video clips;
and storing the processed effective video clips in MP4 format.
Further, after acquiring the individual features of each cow in each frame of image, the method further comprises:
and converting the format and the information of the individual features to form corresponding feature set points.
Further, the specific calculation mode of the flock statistical number in the ith frame image is as follows:
if the number of cattle groups in the ith-1 frame image is M i-1 The number of cattle group identification in the ith frame image is N i The corresponding superposition quantity is K i The number S of flocks counted in the ith frame image i The method comprises the following steps:
S i ＝M i-1 +N i -K i ；
wherein, the flock statistical quantity S in the first frame of image 1 ＝M 1 。
In a second aspect, an embodiment of the present invention further provides a tracking and counting device for field aerial beef flocks, including:
the preprocessing module is used for preprocessing the field aerial video;
the first acquisition module is used for carrying out feature detection and extraction on the preprocessed field aerial video by using the YOLOv4 target detection model to acquire the number of cattle group identification in each frame of image and individual features of each cattle, wherein the individual features comprise position features, attribute features and score features;
the second acquisition module is used for inputting the position characteristic and the attribute characteristic of each cow in the front and back frame images into a SuperGlue characteristic matching model to acquire the characteristic corresponding relation of the cow groups in the front and back frame images and the superposition quantity of the cow groups in the back frame images and the cow groups in the front frame images;
the statistical quantity calculation module is used for calculating the statistical quantity of the flocks in the ith frame of image according to the identification quantity of the flocks in the ith frame of image, the identification quantity of the flocks in the ith-1 frame of image and the corresponding superposition quantity, wherein i is a positive integer, and i is more than or equal to 2.
Further, the preprocessing of the outdoor aerial video includes:
removing aerial video clips and interference video clips lost by targets, and reserving effective video clips;
carrying out slow, quick, rotation, frame skip and static treatment on the effective video clips;
and storing the processed effective video clips in MP4 format.
Further, the apparatus further comprises:
and the conversion module is used for carrying out format and information conversion on the individual characteristics to form corresponding characteristic gathering points.
Further, the specific calculation mode of the flock statistical number in the ith frame image is as follows:
if the number of cattle groups in the ith-1 frame image is M i-1 The number of cattle group identification in the ith frame image is N i The corresponding superposition quantity is K i The number S of flocks counted in the ith frame image i The method comprises the following steps:
S i ＝M i-1 +N i -K i ；
wherein, the flock statistical quantity S in the first frame of image 1 ＝M 1 。
In a third aspect, an embodiment of the present invention further provides a wild aerial cattle group tracking counting system, including a processor, an input device, an output device, and a memory, where the processor, the input device, the output device, and the memory are connected to each other, where the memory is configured to store a computer program, the computer program includes program instructions, and the processor is configured to invoke the program instructions to perform the method according to the first aspect.
In a fourth aspect, embodiments of the present invention also provide a computer readable storage medium storing a computer program comprising program instructions which, when executed by a processor, cause the processor to perform the method according to the first aspect.
According to the invention, a field flock counting method based on unmanned aerial vehicle aerial photography is built by adopting a YOLOv4 target detection model and a SuperGlue feature matching model. The method for counting the cattle group in the video through the YOLOv4 detects and extracts the characteristics of the cattle group targets in the video, and is combined with a characteristic matching network SuperGlue to form an end-to-end YOLOv4-SuperGlue field cattle group counting method, so that the problems of label change caused by individual shielding and label mutation caused by aerial unmanned aerial vehicle speed change are solved, accuracy and precision are improved, accurate matching of the cattle group targets between frames and cattle group counting are realized, and the problems of target loss and shielding false detection caused by speed change during cattle group movement and unmanned aerial vehicle flight are solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below.
Fig. 1 is a flowchart of a tracking and counting method for field aerial cattle groups provided by an embodiment of the invention;
FIG. 2 is a reference schematic diagram of an output result of a Yolov4 object detection model provided by an embodiment of the present invention after feature detection and extraction of an image;
FIG. 3 is a reference schematic diagram of the result output by the SuperGlue feature matching model provided by the embodiment of the invention;
fig. 4 is a schematic structural diagram of a tracking and counting device for field aerial beef flocks according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a tracking and counting system for wild aerial cattle groups according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be understood that the terms "comprises" and "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
As used in this specification and the appended claims, the term "if" may be interpreted as "when..once" or "in response to a determination" or "in response to detection" depending on the context. Similarly, the phrase "if a determination" or "if a [ described condition or event ] is detected" may be interpreted in the context of meaning "upon determination" or "in response to determination" or "upon detection of a [ described condition or event ]" or "in response to detection of a [ described condition or event ]".
It is noted that unless otherwise indicated, technical or scientific terms used herein should be given the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.
In this embodiment, adopt six rotor unmanned aerial vehicle that the wheelbase is 80cm to carry on the 4K camera of 35mm focus and take the slope and take photo by plane the shadow mode, take photo by plane the tracking of taking photo by plane to grassland flock according to unmanned aerial vehicle planning route, take photo lens installs in unmanned aerial vehicle below, remove flock by plane with its being applied to monitoring field, take photo by plane the video, count flock by plane the identification quantity in the video of taking photo by plane from the field.
As shown in fig. 1, the embodiment of the invention provides a flowchart of a tracking and counting method for field aerial beef flocks. The method may comprise the steps of:
s100: preprocessing field aerial videos.
Specifically, the preprocessing of the field aerial video includes:
s101: and removing the aerial video clips and the interference video clips lost by the target, and reserving the effective video clips.
When aerial photography is carried out in the wild, the shooting direction of the camera lens can be changed at any time, so that a cattle group target in a shot video fragment is lost, the camera lens can be polluted, the shot video is interfered, the aerial photography video fragment and the interference video fragment with the lost target are required to be removed for accelerating the processing speed, and only the effective video fragment is reserved.
S102: and carrying out slow, quick, rotation, frame skip and static processing on the effective video clips.
S103: and storing the processed effective video clips in MP4 format.
Stored in MP4 format, a frame width of 3840 pixels, a frame height of 2160 pixels, a data rate of 18688kbps, a total bit rate of 18771kbps, and a video frame rate of 29.97f/s.
S200: and performing feature detection and extraction on the preprocessed field aerial video by using a YOLOv4 target detection model to obtain the number of cow group identification in each frame of image and individual features of each cow, wherein the individual features comprise position features, attribute features and score features.
The Yolov4 target detection model mainly comprises a Backbone network backbox, namely CSPDarknet53, neck, SPP, PAN and Head, namely Yolov3, wherein the Backbone network backbox comprises Bag of Freebie (sBoF) and Bag of specialty (BoS) for backbox which are respectively CutMix, mosaic data enhancement, dropBlock regularization, label-like smoothing and Mish activation, cross-phase part Connection (CSP) and multi-input weighted residual connection (MiWRC); bag of Freebies (BoF) and Bag of specialty (BoS) for backbones for detectors are CIoU-loss, cmBN, dropBlock regularization, mosoic data enhancement, self-countermeasure training, grid sensitivity cancellation, handling of single ground truth using multiple anchors, cosine annealing scheduler, optimal superparameters, random training shapes and rish activations, SPP blocks, SAM blocks, PAN path aggregation blocks, DIoU-NMS, respectively. The output result of the image after feature detection and extraction by using the YOLOv4 object detection model is shown in fig. 2.
Because the data format of the individual characteristics of each cow acquired by using the YOLOv4 target detection model is not consistent with the data format required by the SuperGlue characteristic matching model, the data format cannot be directly input into the SuperGlue characteristic matching model, and therefore the individual characteristics are required to be subjected to format and information conversion to form corresponding characteristic gathering points, so that the data format of the converted individual characteristics accords with the data format required by the SuperGlue characteristic matching model.
S300: and inputting the position characteristics and the attribute characteristics of each cow in the front and rear frame images into a SuperGlue characteristic matching model, and acquiring the characteristic corresponding relation of the cow groups in the front and rear frame images and the superposition quantity of the cow groups in the rear frame images and the cow groups in the front frame images.
Specifically, the number of the cattle groups in the rear frame image and the number of the cattle groups in the front frame image are the number of the cattle individuals repeatedly appearing in the rear frame image and the front frame image. The result of the SuperGlue feature matching model output is shown in FIG. 3.
S400: and calculating to obtain the number of the flocks in the ith frame of image according to the number of the flocks in the ith frame of image, the number of the flocks in the ith-1 frame of image and the corresponding superposition number, wherein i is a positive integer, and i is more than or equal to 2.
Specifically, the specific calculation mode of the flock statistical number in the ith frame of image is as follows:
if the number of cattle groups in the ith-1 frame image is M i-1 The number of cattle group identification in the ith frame image is N i The corresponding superposition quantity is K i The number S of flocks counted in the ith frame image i The method comprises the following steps:
S i ＝M i-1 +N i -K i ；
wherein, the flock statistical quantity S in the first frame of image 1 ＝M 1 。
For example, the number of cattle groups in the 6 th frame image is M 6 =10, i.e. the herd comprises 10 cattle; cattle group identification number M in 7 th frame image 7 =14, i.e. the herd comprises 14 calves; the number of the overlapping of the cattle groups in the 7 th frame image and the 6 th frame image is K 7 =8, i.e. cattle in 8 th frame 6 images also appear in 7 th frame images. Thus, in the 7 th frame image, the number of 8 cows was counted in the 6 th frame image, and 6 cows were newly photographed. Therefore, the flock statistical quantity S in the 7 th frame image is calculated 7 The method comprises the following steps: s is S 7 =10+14-8=16, i.e. the number of flocks in image 7 was 16.
In particular, since the first frame image is not preceded by the previous frame image, the flock count number S in the first frame image 1 ＝M 1 。
In a second aspect, the embodiment of the invention also provides a tracking and counting device for the wild aerial beef flock. As shown in fig. 4, the apparatus may include:
the preprocessing module 201 is used for preprocessing field aerial videos;
a first obtaining module 202, configured to perform feature detection and extraction on the preprocessed field aerial video by using a YOLOv4 target detection model, to obtain the number of bovine group identifications in each frame of image and individual features of each bovine, where the individual features include a position feature, an attribute feature and a score feature;
the second obtaining module 203 is configured to input the position feature and the attribute feature of each cow in the front and rear frame images into a SuperGlue feature matching model, and obtain a feature correspondence of the cow in the front and rear frame images and a number of overlapping cow in the rear frame image and the cow in the front frame image;
the statistic number calculating module 204 is configured to calculate, according to the number of cattle groups identified in the ith frame of image, the number of cattle groups identified in the ith-1 frame of image, and the corresponding overlapping number, the statistic number of cattle groups in the ith frame of image, where i is a positive integer, and i is greater than or equal to 2.
Specifically, the preprocessing of the field aerial video includes:
removing aerial video clips and interference video clips lost by targets, and reserving effective video clips;
carrying out slow, quick, rotation, frame skip and static treatment on the effective video clips;
and storing the processed effective video clips in MP4 format.
Further, the apparatus further comprises:
the conversion module 205 is configured to perform format and information conversion on the individual features to form corresponding feature collection points.
Further, the specific calculation mode of the flock statistical number in the ith frame image is as follows:
if the number of cattle groups in the ith-1 frame image is M i-1 The number of cattle group identification in the ith frame image is N i The corresponding superposition quantity is K i The number S of flocks counted in the ith frame image i The method comprises the following steps:
S i ＝M i-1 +N i -K i ；
wherein, the flock statistical quantity S in the first frame of image 1 ＝M 1 。
Based on the same inventive concept, the embodiment of the invention also provides an off-the-spot aerial photography flock tracking and counting system. As shown in fig. 5, the system may include: one or more processors 101, one or more input devices 102, one or more output devices 103, and a memory 104, the processors 101, input devices 102, output devices 103, and memory 104 being interconnected by a bus 105. The memory 104 is used for storing a computer program comprising program instructions, the processor 101 being configured to invoke the program instructions for performing the method of the above-described embodiment of the wild aerial cattle group tracking counting method.
It should be appreciated that in embodiments of the present invention, the processor 101 may be a central processing unit (Central Processing Unit, CPU), which may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSPs), application specific integrated circuits (Application Specific Integrated Circuit, ASICs), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The input device 102 may include a keyboard or the like, and the output device 103 may include a display (LCD or the like), a speaker or the like.
The memory 104 may include read only memory and random access memory and provides instructions and data to the processor 101. A portion of the memory 104 may also include non-volatile random access memory. For example, the memory 104 may also store information of device type.
In a specific implementation, the processor 101, the input device 102, and the output device 103 described in the embodiments of the present invention may execute the implementation described in the embodiments of the method for tracking and counting an wild aerial photo flock provided in the embodiments of the present invention, which is not described herein again.
It should be noted that, for the specific workflow of the system for tracking and counting the wild aerial photo flock, reference may be made to the foregoing method embodiment, and details are not repeated herein.
Further, an embodiment of the present invention also provides a readable storage medium storing a computer program, the computer program including program instructions that when executed by a processor implement: the wild aerial photography flock tracking and counting method.
The computer readable storage medium may be an internal storage unit of the background server according to the foregoing embodiment, for example, a hard disk or a memory of the system. The computer readable storage medium may also be an external storage device of the system, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the system. Further, the computer readable storage medium may also include both internal storage units and external storage devices of the system. The computer readable storage medium is used to store the computer program and other programs and data required by the system. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.
Those of ordinary skill in the art will appreciate that the elements and algorithm steps described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of the two, and that the elements and steps of the examples have been generally described in terms of function in the foregoing description to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention is essentially or a part contributing to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
While the invention has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions of equivalents may be made and equivalents will be apparent to those skilled in the art without departing from the scope of the invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.

Claims (10)

1. A tracking and counting method for field aerial cattle groups is characterized by comprising the following steps:
preprocessing field aerial videos;
performing feature detection and extraction on the preprocessed field aerial video by using a YOLOv4 target detection model to obtain the number of cattle group identification in each frame of image and individual features of each cattle, wherein the individual features comprise position features, attribute features and score features;
inputting the position feature and the attribute feature of each cow in the front and back frame images into a SuperGlue feature matching model, and acquiring the feature corresponding relation of the cow groups in the front and back frame images and the superposition quantity of the cow groups in the back frame images and the cow groups in the front frame images;
and calculating to obtain the number of the flocks in the ith frame of image according to the number of the flocks in the ith frame of image, the number of the flocks in the ith-1 frame of image and the corresponding superposition number, wherein i is a positive integer, and i is more than or equal to 2.
2. The method for tracking and counting wild aerial cattle groups according to claim 1, wherein the preprocessing of the wild aerial videos comprises the following steps:
removing aerial video clips and interference video clips lost by targets, and reserving effective video clips;
carrying out slow, quick, rotation, frame skip and static treatment on the effective video clips;
and storing the processed effective video clips in MP4 format.
3. A method of tracking and counting field aerial cattle groups according to claim 1, wherein after the individual characteristics of each cattle in each frame of image are obtained, the method further comprises:
and converting the format and the information of the individual features to form corresponding feature set points.
4. The method for tracking and counting the wild aerial cattle groups according to claim 1, wherein the specific calculation mode of the cattle group statistical number in the ith frame of image is as follows:
if the number of cattle groups in the ith-1 frame image is M i-1 The number of cattle group identification in the ith frame image is N i The corresponding superposition quantity is K i The number S of flocks counted in the ith frame image i The method comprises the following steps:
S i ＝M i-1 +N i -K i ；
wherein, the flock statistical quantity S in the first frame of image 1 ＝M 1 。
5. Wild flock of cattle of taking photo by plane tracks counting equipment, a serial communication port, include:
the preprocessing module is used for preprocessing the field aerial video;
the first acquisition module is used for carrying out feature detection and extraction on the preprocessed field aerial video by using the YOLOv4 target detection model to acquire the number of cattle group identification in each frame of image and individual features of each cattle, wherein the individual features comprise position features, attribute features and score features;
the second acquisition module is used for inputting the position characteristic and the attribute characteristic of each cow in the front and back frame images into a SuperGlue characteristic matching model to acquire the characteristic corresponding relation of the cow groups in the front and back frame images and the superposition quantity of the cow groups in the back frame images and the cow groups in the front frame images;
the statistical quantity calculation module is used for calculating the statistical quantity of the flocks in the ith frame of image according to the identification quantity of the flocks in the ith frame of image, the identification quantity of the flocks in the ith-1 frame of image and the corresponding superposition quantity, wherein i is a positive integer, and i is more than or equal to 2.
6. A field aerial cattle herd tracking and counting device according to claim 5, wherein the preprocessing of the field aerial video comprises:
removing aerial video clips and interference video clips lost by targets, and reserving effective video clips;
carrying out slow, quick, rotation, frame skip and static treatment on the effective video clips;
and storing the processed effective video clips in MP4 format.
7. A field aerial cattle herd tracking and counting device according to claim 5, wherein the device further comprises:
and the conversion module is used for carrying out format and information conversion on the individual characteristics to form corresponding characteristic gathering points.
8. The field aerial cattle group tracking and counting device according to claim 5, wherein the specific calculation mode of the cattle group statistical number in the ith frame of image is as follows:
if the number of cattle groups in the ith-1 frame image is M i-1 The number of cattle group identification in the ith frame image is N i The corresponding superposition quantity is K i The number S of flocks counted in the ith frame image i The method comprises the following steps:
S i ＝M i-1 +N i -K i ；
wherein, the flock statistical quantity S in the first frame of image 1 ＝M 1 。
9. A wild-type aerial herd tracking and counting system comprising a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory being interconnected, wherein the memory is adapted to store a computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the method of any of claims 1-4.
10. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program comprising program instructions which, when executed by a processor, cause the processor to perform the method of any of claims 1-4.
CN202310021725.8A 2023-01-06 2023-01-06 Method, equipment, system and storage medium for counting wild aerial cattle groups Pending CN116189018A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310021725.8A CN116189018A (en) 2023-01-06 2023-01-06 Method, equipment, system and storage medium for counting wild aerial cattle groups

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310021725.8A CN116189018A (en) 2023-01-06 2023-01-06 Method, equipment, system and storage medium for counting wild aerial cattle groups

Publications (1)

Publication Number Publication Date
CN116189018A true CN116189018A (en) 2023-05-30

Family

ID=86439608

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310021725.8A Pending CN116189018A (en) 2023-01-06 2023-01-06 Method, equipment, system and storage medium for counting wild aerial cattle groups

Country Status (1)

Country Link
CN (1) CN116189018A (en)

Similar Documents

Publication Publication Date Title
US11627726B2 (en) System and method of estimating livestock weight
CN107667903B (en) Livestock breeding living body weight monitoring method based on Internet of things
CN108447091B (en) Target positioning method and device, electronic equipment and storage medium
Tsai et al. Assessment of dairy cow heat stress by monitoring drinking behaviour using an embedded imaging system
CN108229351B (en) Animal breeding method, animal breeding device and electronic equipment
Guzhva et al. Now you see me: Convolutional neural network based tracker for dairy cows
CN111325217B (en) Data processing method, device, system and medium
JP7108941B2 (en) ANIMAL INFORMATION MANAGEMENT SYSTEM AND ANIMAL INFORMATION MANAGEMENT METHOD
Kaixuan et al. Target detection method for moving cows based on background subtraction
Lu et al. An automatic splitting method for the adhesive piglets’ gray scale image based on the ellipse shape feature
CN112036206A (en) Monitoring method, device and system based on identification code type pig ear tag
CN114721403B (en) Automatic driving control method and device based on OpenCV and storage medium
CN115100732A (en) Fishing detection method and device, computer equipment and storage medium
WO2023041904A1 (en) Systems and methods for the automated monitoring of animal physiological conditions and for the prediction of animal phenotypes and health outcomes
Chae et al. Identifying the mating posture of cattle using deep learning-based object detection with networks of various settings
CN113724250A (en) Animal target counting method based on double-optical camera
CN116189018A (en) Method, equipment, system and storage medium for counting wild aerial cattle groups
CN115147782A (en) Dead animal identification method and device
CN109492541B (en) Method and device for determining type of target object, plant protection method and plant protection system
CN115760904A (en) Livestock and poultry statistical method, device, electronic equipment and medium
CN111150402A (en) Method, device, storage medium and electronic device for determining livestock form parameters
EP4210004A1 (en) An animal detecting device, system, method and computer program
KR101824203B1 (en) Apparatus and method for calculation of target object
KR102424901B1 (en) method for detecting estrus of cattle based on object detection algorithm
Shu et al. Non-contact respiration rate measurement of multiple cows in a free-stall barn using computer vision methods

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