CN114115359A - Unmanned aerial vehicle mountain sheep searching system and working method thereof - Google Patents
Unmanned aerial vehicle mountain sheep searching system and working method thereof Download PDFInfo
- Publication number
- CN114115359A CN114115359A CN202111251088.0A CN202111251088A CN114115359A CN 114115359 A CN114115359 A CN 114115359A CN 202111251088 A CN202111251088 A CN 202111251088A CN 114115359 A CN114115359 A CN 114115359A
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- searching
- sheep
- target
- 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.)
- Granted
Links
- 241001416149 Ovis ammon Species 0.000 title claims abstract description 15
- 238000000034 method Methods 0.000 title claims description 31
- 238000001514 detection method Methods 0.000 claims abstract description 53
- 241000283690 Bos taurus Species 0.000 claims abstract description 31
- 230000005540 biological transmission Effects 0.000 claims abstract description 28
- RZVHIXYEVGDQDX-UHFFFAOYSA-N 9,10-anthraquinone Chemical compound C1=CC=C2C(=O)C3=CC=CC=C3C(=O)C2=C1 RZVHIXYEVGDQDX-UHFFFAOYSA-N 0.000 claims abstract description 14
- 238000004891 communication Methods 0.000 claims abstract description 4
- 241000283898 Ovis Species 0.000 claims description 38
- 238000012549 training Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 230000000638 stimulation Effects 0.000 claims description 10
- 238000002372 labelling Methods 0.000 claims description 5
- 238000013459 approach Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 claims description 3
- 238000009432 framing Methods 0.000 claims description 3
- 238000004806 packaging method and process Methods 0.000 claims description 3
- 238000010422 painting Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 2
- 241001494479 Pecora Species 0.000 abstract 3
- 238000007689 inspection Methods 0.000 description 3
- 241000283707 Capra Species 0.000 description 2
- 244000144992 flock Species 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 241000282461 Canis lupus Species 0.000 description 1
- 244000025254 Cannabis sativa Species 0.000 description 1
- 241001465754 Metazoa Species 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 244000144980 herd Species 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000009304 pastoral farming Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/12—Target-seeking control
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention provides an unmanned aerial vehicle mountain sheep searching system which comprises an unmanned aerial vehicle subsystem and a ground terminal subsystem; the ground terminal system comprises a touch display screen, a processor and a first image transmission device, wherein a searching module is installed in the processor; the searching module is used for generating a searching instruction, and the searching instruction at least comprises a searching range and a searching target type; the unmanned aerial vehicle subsystem comprises an unmanned aerial vehicle body, and a battery pack, a flight control device, an onboard computer, a video recording device and a second image transmission device which are arranged on the unmanned aerial vehicle body; a route planning module and a target detection model are installed in the airborne computer. According to the invention, the unmanned aerial vehicle is adopted to replace a manual mode to search for lost cattle and sheep, and the lost cattle and sheep can be automatically driven to return to the destination, so that the efficiency and safety of searching sheep in mountainous areas are effectively improved; carry on picture simultaneously and pass device and 4G communication module on unmanned aerial vehicle, ensure that data can effectively transmit to the ground end.
Description
Technical Field
The invention relates to the technical field of grazing of unmanned aerial vehicles, in particular to a mountain sheep searching system of an unmanned aerial vehicle and a working method thereof.
Background
The western mountainous areas of China have complex terrains and sparse population, and people in non-urban areas are mainly in the nomadic industry. The mountainous area has complex terrain, the mountains are continuous, and the great elevation difference exists from the feet to the top of the mountains, so that the method is a great challenge to the production and the life of farmers. In order to adapt to the terrain, most herdsmen in mountain areas choose goats/yaks which can stand stably on a hillside to breed, flocks of cattle and sheep go out of the shed in the daytime and automatically go to nearby hillsides to eat grass, and the flocks of cattle and sheep return to the shed after sunset. The herdsman only counts the number at night. The condition that cattle and sheep are lost is often met, and herdsmen need to go to the hillside to search by themselves. Every time one cow is lost, the loss of ten thousand yuan can be caused to herdsmen.
At present, the main problems encountered by minority herdsmen engaged in breeding cattle and sheep in mountain areas are as follows: 1. the mountain roads are rugged, numerous in mountains, the altitude difference from the feet of the mountain to the top of the mountain is more than two kilometers, no stable mountain roads can run, and partial mountainous roads need to be climbed, so that the mountain roads are very dangerous. 2. The single cattle and sheep are too scattered, the target size is small, the single cattle and sheep are easy to be confused with shrubs, stones and snow mountains, cannot be distinguished easily, and cannot be seen clearly only by naked eyes. 3. The herdsman only goes out to find lost cattle and sheep, and needs to climb over a plurality of mountains within a few days each time, 4. the mountain road is steep, wild animals such as wolf herds also exist in mountain forests, and the herdsman alone goes out to find sheep and is easy to encounter danger.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides an unmanned aerial vehicle goat searching system in mountainous areas and a working method thereof. The unmanned aerial vehicle has flexible flight and is easy to take off and land vertically; the carried laser radar module can be used for modeling mountainous terrain, so that the flight safety of the unmanned aerial vehicle is guaranteed; the carried target detection module is easy to identify small targets, and the identification effect by adopting an algorithm is better than that by naked eyes. On the same route, the herdsman needs a day from the foot of the mountain to the top of the mountain, and the adoption of the unmanned aerial vehicle can only need a few minutes, which is enough for the unmanned aerial vehicle to charge and patrol at regular time; the efficiency and the security of seeking sheep in the mountainous area are effectively improved.
In order to achieve the purpose, the invention adopts the following technical scheme:
the embodiment of the invention provides an unmanned aerial vehicle mountain sheep searching system, which comprises an unmanned aerial vehicle subsystem and a ground terminal subsystem;
the ground terminal system comprises a touch display screen, a processor and a first image transmission device, wherein a searching module is installed in the processor; the searching module is used for generating a searching instruction, and the searching instruction at least comprises a searching range and a searching target type; the first image transmission device monitors the working state of the unmanned aerial vehicle subsystem in a remote desktop mode through a point-to-point image transmission technology;
the unmanned aerial vehicle subsystem comprises an unmanned aerial vehicle body, and a battery pack, a flight control device, an onboard computer, a video recording device and a second image transmission device which are arranged on the unmanned aerial vehicle body; a route planning module and a target detection model are installed in the airborne computer;
the battery pack is respectively connected with the flight control device, the onboard computer, the video recording device and the image transmission device so as to provide electric energy required by normal work of the battery pack;
the route planning module is used for receiving a searching instruction sent by the ground terminal system and automatically planning a flight route according to a searching range; the flight control device is used for controlling the unmanned aerial vehicle to fly along the flight route sent by the route planning module; the video recording device is used for shooting ground painting images in the flight process of the unmanned aerial vehicle; the target detection model processes the map image shot by the video recording device and detects whether a search target exists on the image; the second image transmission device is used for receiving the control signal transmitted by the ground terminal system and sending an image signal and a detection result to the ground terminal system;
when the target detection model finds the search target, the airborne computer sends the image containing the search target and the current unmanned aerial vehicle position information to the ground terminal system together.
Further, the search module automatically acquires a to-be-detected area selected by a user on the touch display screen, converts an area boundary of the to-be-detected area into a GPS coordinate, and transmits the GPS coordinate to an airborne computer of the unmanned aerial vehicle.
Further, the route planning module is used for automatically generating an S-shaped flight path according to the area to be detected.
Further, the flight control device comprises a distance sensor, and the distance sensor is used for detecting the distance value between the unmanned aerial vehicle body and the mountain surface in real time in the flight process of the unmanned aerial vehicle;
the flight control device adjusts the flight height of the unmanned aerial vehicle body in real time according to the distance value fed back by the distance sensor, so that the distance value between the unmanned aerial vehicle body and the mountain surface is maintained within a preset distance value range.
Further, the types of the search targets include cattle and sheep.
Further, the unmanned aerial vehicle subsystem further comprises an active stimulation module, wherein the active stimulation module is used for driving the airborne sound box to emit driving sound so as to drive the target in the image to return to the destination.
Further, the target detection model processes the shot image based on a uoloV4 algorithm; if the detected confidence of the object in the shot image is larger than a first preset confidence threshold value, automatically shooting and storing position information; if the confidence coefficient of the detected object is larger than a second preset confidence coefficient threshold value, automatically triggering an alarm device to remind ground station workers, and simultaneously triggering an active stimulation module to carry out automatic driving and automatic tracking; the second preset confidence threshold is greater than the first preset confidence threshold.
Further, the training process of the target detection model comprises the following steps:
gather a plurality of sections videos of the different backgrounds of different angles that ordinary aerial photography unmanned aerial vehicle shot, take out frame processing to the video, extract a certain amount of sample picture, send into uolo original algorithm and detect in batches, carry out data processing to the testing result, select the target that satisfies following condition simultaneously: (1) detecting targets with the target frame area larger than 100 unit pixels, (2) detecting targets with the type of cattle and sheep, carrying out batch labeling, carrying out manual screening, adding targets which cannot be detected by a labeling script, deleting targets with detection errors, correcting the target frame, and generating a training data set;
putting the marked training data set into an image enhancement module, performing rotation, distortion, stretching and inversion processing on the sample pictures, expanding the number of the shot sample pictures, and correspondingly transforming the mark frame to obtain a complete target detection data set;
sending the target detection data set into a YOLO algorithm for training, and setting relevant parameters of a detection network by combining shooting parameters of an unmanned aerial vehicle, processing speed of an airborne computer and the class characteristics of a detection target; and continuously adjusting the learning rate in the training process to obtain a target detection model.
Further, the unmanned aerial vehicle subsystem further comprises a 4G communication module; when the unmanned aerial vehicle picture-transmitted signal is not good, the shot image, the detection result and the current position coordinate of the unmanned aerial vehicle are packaged into a ros message, and the message is stored in a public message pool by adopting a publishing-subscribing mechanism, so that the ground terminal system can obtain the position of the cattle and sheep by subscribing the message.
Based on the unmanned aerial vehicle mountain area sheep searching system, the invention also provides a working method of the unmanned aerial vehicle mountain area sheep searching system, and the working method comprises the following steps:
s1, configuring global project parameters including expected confidence, ground clearance, flight speed and active stimulation parameters;
s2, the unmanned aerial vehicle receives the remote control signal, checks the GPS signal, the running state of the onboard computer and the automatic takeoff after the image transmission channel is stable, and hovers;
s3, automatically connecting the ground terminal system with an onboard computer of the unmanned aerial vehicle, manually framing a search area on a terminal satellite map of a display screen by a user, generating a search instruction, and transmitting the search instruction to a path planning module of the onboard computer, so that the path planning module automatically plans an S-shaped flight route according to a search range;
s4, controlling the unmanned aerial vehicle to patrol along the hillside according to the planned flight route, and simultaneously carrying out target detection on the shot ground picture image:
when the pictures are shot and the suspected lost cattle and sheep pictures exist, the unmanned aerial vehicle automatically changes the circuit to approach the detection; if the confidence coefficient of the detected cattle and sheep is more than or equal to 50 percent after approaching, automatically photographing and recording, recording the position of the unmanned aerial vehicle at the moment, packaging the data into ros information, transmitting the ros information back to the ground terminal system through a second image transmission device, and requesting ground further processing;
and S5, driving according to the driving instruction sent by the herdsman.
The invention has the beneficial effects that:
according to the invention, the unmanned aerial vehicle is adopted to replace a manual mode to search for the lost cattle and sheep, and the lost cattle and sheep can be automatically driven to return to the destination, so that the efficiency and the safety of searching sheep in mountainous areas are effectively improved. In consideration of the situation that the mountain environment is complex and the signal is not good in some cases, the unmanned aerial vehicle is provided with the image transmission device and the 4G communication module at the same time, and data can be effectively transmitted to the ground end.
Drawings
Fig. 1 is a schematic structural view of a mountain sheep searching system of an unmanned aerial vehicle according to an embodiment of the invention.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings.
It should be noted that the terms "upper", "lower", "left", "right", "front", "back", etc. used in the present invention are for clarity of description only, and are not intended to limit the scope of the present invention, and the relative relationship between the terms and the terms is not limited by the technical contents of the essential changes.
Fig. 1 is a schematic structural view of a mountain sheep searching system of an unmanned aerial vehicle according to an embodiment of the invention. Referring to fig. 1, the sheep searching system comprises an unmanned aerial vehicle subsystem and a ground terminal subsystem.
Ground terminal system
The ground terminal system comprises a touch display screen, a processor and a first image transmission device, wherein a searching module is installed in the processor; the searching module is used for generating a searching instruction, and the searching instruction at least comprises a searching range and a searching target type; the first image transmission device monitors the working state of the unmanned aerial vehicle subsystem in a remote desktop mode through a point-to-point image transmission technology.
Illustratively, the ground terminal system can adopt a mobile phone, a one-piece machine or a PC terminal and the like. Taking a mobile phone as an example, a herdsman can select a search area on a mobile phone screen, and can also check the working state of the unmanned aerial vehicle or a target image and a target position returned by the unmanned aerial vehicle in real time through the mobile phone screen, and further issue a related control instruction.
(II) unmanned aerial vehicle subsystem
The unmanned aerial vehicle subsystem comprises an unmanned aerial vehicle body, and a battery pack, a flight control device, an onboard computer, a video recording device and a second image transmission device which are arranged on the unmanned aerial vehicle body; a route planning module and a target detection model are installed in the airborne computer.
The battery pack is respectively connected with the flight control device, the onboard computer, the video recording device and the image transmission device so as to provide electric energy required by normal operation of the battery pack.
The route planning module is used for receiving a searching instruction sent by the ground terminal system and automatically planning a flight route according to a searching range; the flight control device is used for controlling the unmanned aerial vehicle to fly along the flight route sent by the route planning module; the video recording device is used for shooting ground painting images in the flight process of the unmanned aerial vehicle; the target detection model processes the map image shot by the video recording device and detects whether a search target exists on the image; the second image transmission device is used for receiving the control signal transmitted by the ground terminal system and sending the image signal and the detection result to the ground terminal system.
When the target detection model finds the search target, the airborne computer sends the image containing the search target and the current unmanned aerial vehicle position information to the ground terminal system together.
The search module automatically acquires the area to be detected selected by the user on the touch display screen, converts the area boundary of the area to be detected into GPS coordinates and transmits the GPS coordinates to the airborne computer of the unmanned aerial vehicle. And the route planning module automatically generates an S-shaped flight path according to the area to be detected. The unmanned aerial vehicle keeps being 20 meters at the minimum from the ground through distance sensor at the in-process of patrolling and examining, and effective visual angle scope is 150m at the minimum. The inspection speed is 15km/h, the battery can last for 30min on average, and two mountain bodies can be inspected.
The unmanned aerial vehicle subsystem further comprises an active stimulation module for driving the airborne sound box to emit driving sound so as to drive the target in the image to return to the destination.
For example, the object detection model processes the captured image based on yoloV4 algorithm; if the detected confidence of the object in the shot image is larger than a first preset confidence threshold (such as 50%), automatically shooting and storing position information; if the confidence of the detected object is larger than a second preset confidence threshold (such as 90%), and the second preset confidence threshold is larger than the first preset confidence threshold, an alarm device is automatically triggered to remind ground station workers, and an active stimulation module is simultaneously triggered to automatically drive and track.
Illustratively, the training process of the target detection model includes the steps of:
gather a plurality of sections videos of the different backgrounds of different angles that ordinary aerial photography unmanned aerial vehicle shot, take out frame processing to the video, extract a certain amount of sample picture, send into yolo original algorithm and detect in batches, carry out data processing to the testing result, select the target that satisfies following condition simultaneously: (1) the area of a target frame is larger than that of a detection target of 100 unit pixels, and (2) the targets of cattle and sheep in the detection category are labeled in batches and screened manually, the targets which cannot be detected by a labeling script are added, the targets with detection errors are deleted, the target frame is corrected, and a training data set is generated. And putting the marked training data set into an image enhancement module, rotating, distorting, stretching and reversing the sample pictures, expanding the number of the shot sample pictures, and correspondingly transforming the mark frame to obtain a complete target detection data set. Sending the target detection data set into a YOLO algorithm for training, and setting relevant parameters of a detection network by combining shooting parameters of an unmanned aerial vehicle, processing speed of an airborne computer and the class characteristics of a detection target; and continuously adjusting the learning rate in the training process to obtain a target detection model.
For example, 30 video segments with different backgrounds and different angles shot by a common aerial photography unmanned aerial vehicle are adopted, the time duration of each video segment is 10-15 seconds, frame extraction is carried out, one image is extracted from each 3 frames, and about 3700 images are extracted in total; and expanding the number of the shot pictures of the plurality of cattle and sheep into 55000 pictures through rotating, twisting, pulling, reversing and other modes. And sending the detection data set into a YOLO algorithm for training, and setting the detection network as 416 x 416 and the confidence threshold as 20% by combining the characteristics of small target image shot by the unmanned aerial vehicle, airborne computer processing speed limitation and the like. And setting the size to be 2 and the filters to be 21 by combining the class characteristics of the detection target, and continuously adjusting the learning rate in the training process. And obtaining a target detection model after training.
And if the cattle and sheep targets are detected in the unmanned aerial vehicle inspection process, requesting indication from the ground terminal. And receiving a signal and operating according to the request instruction. This step may occur in three situations:
the first condition, when unmanned aerial vehicle picture signal is good, then instruct to ground end request, if ground end staff audits to confirm finds the cattle and sheep that lose, then herdsman can go to by oneself. Also can take initiative amazing, the automatic broadcast sound of machine-carried audio amplifier module drives, and the cattle and sheep of unmanned aerial vehicle automatic tracking frame selection drives simultaneously. If the cattle and sheep advancing direction is the direction of workers far away from the ground end, the position of the unmanned aerial vehicle can be automatically changed, so that the unmanned aerial vehicle returns to track towards the correct direction.
In the second situation, when the unmanned aerial vehicle is too remote and cannot receive the image transmission information, the 4G module is automatically started to be connected with the Internet, the camera is used for drawing, the detection result and the current position coordinate of the unmanned aerial vehicle are packaged into ros information, the ros information is stored in a public information pool by adopting a publishing-subscribing mechanism, and the ground end is used for subscribing the ros information, so that the position of the cattle and sheep is obtained.
And in the third situation, when the pattern signal of the unmanned aerial vehicle is lost and the position is not covered by the 4G network, the data is stored in the SD card carried by the airborne computer, the data is actively returned, and the command is sent again in the area with good signal to be returned.
Based on the unmanned aerial vehicle mountain area sheep searching system, the embodiment also provides a working method of the unmanned aerial vehicle mountain area sheep searching system, and the working method comprises the following steps:
s1, configuring global project parameters including expected confidence, ground clearance, flying speed and active stimulation parameters.
And S2, the unmanned aerial vehicle receives the remote control signal, checks the GPS signal, the running state of the airborne computer and takes off automatically after the image transmission channel is stable, and hovers.
And S3, automatically connecting the ground terminal system with an onboard computer of the unmanned aerial vehicle, manually framing a search area on a terminal satellite map of the display screen by a user, generating a search instruction, and transmitting the search instruction to a path planning module of the onboard computer, so that the path planning module automatically plans an S-shaped flight route according to a search range.
S4, controlling the unmanned aerial vehicle to patrol along the hillside according to the planned flight route, and simultaneously carrying out target detection on the shot ground picture image:
when the pictures are shot and the suspected lost cattle and sheep pictures exist, the unmanned aerial vehicle automatically changes the circuit to approach the detection; and if the confidence coefficient of the cattle and sheep detected after approaching is not less than 50%, automatically shooting and recording, recording the position of the unmanned aerial vehicle at the moment, packaging the data into ros information, transmitting the ros information back to the ground terminal system through the second image transmission device, and requesting ground further processing. Preferably, the inspection process can be carried with an infrared camera for auxiliary investigation.
And S5, driving according to the driving instruction sent by the herdsman.
The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and the technical solutions that are based on the idea of the present invention are all within the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may be made by those skilled in the art without departing from the principle of the invention.
Claims (10)
1. An unmanned aerial vehicle mountain sheep searching system is characterized by comprising an unmanned aerial vehicle subsystem and a ground terminal subsystem;
the ground terminal system comprises a touch display screen, a processor and a first image transmission device, wherein a searching module is installed in the processor; the searching module is used for generating a searching instruction, and the searching instruction at least comprises a searching range and a searching target type; the first image transmission device monitors the working state of the unmanned aerial vehicle subsystem in a remote desktop mode through a point-to-point image transmission technology;
the unmanned aerial vehicle subsystem comprises an unmanned aerial vehicle body, and a battery pack, a flight control device, an onboard computer, a video recording device and a second image transmission device which are arranged on the unmanned aerial vehicle body; a route planning module and a target detection model are installed in the airborne computer;
the battery pack is respectively connected with the flight control device, the onboard computer, the video recording device and the image transmission device so as to provide electric energy required by normal work of the battery pack;
the route planning module is used for receiving a searching instruction sent by the ground terminal system and automatically planning a flight route according to a searching range; the flight control device is used for controlling the unmanned aerial vehicle to fly along the flight route sent by the route planning module; the video recording device is used for shooting ground painting images in the flight process of the unmanned aerial vehicle; the target detection model processes the map image shot by the video recording device and detects whether a search target exists on the image; the second image transmission device is used for receiving the control signal transmitted by the ground terminal system and sending an image signal and a detection result to the ground terminal system;
when the target detection model finds the search target, the airborne computer sends the image containing the search target and the current unmanned aerial vehicle position information to the ground terminal system together.
2. The unmanned aerial vehicle mountain sheep searching system of claim 1, wherein the search module automatically obtains a to-be-detected area selected by a user on a touch display screen, converts an area boundary of the to-be-detected area into a GPS coordinate, and transmits the GPS coordinate to an onboard computer of the unmanned aerial vehicle.
3. The unmanned aerial vehicle mountain sheep searching system of claim 2, wherein the route planning module is configured to automatically generate an S-shaped flight path according to the area to be detected.
4. The mountain area sheep searching system for unmanned aerial vehicles according to claim 1, wherein the flight control device comprises a distance sensor for detecting the distance value between the unmanned aerial vehicle body and the surface of the mountain body in real time during the flight of the unmanned aerial vehicle;
the flight control device adjusts the flight height of the unmanned aerial vehicle body in real time according to the distance value fed back by the distance sensor, so that the distance value between the unmanned aerial vehicle body and the mountain surface is maintained within a preset distance value range.
5. The unmanned mountain sheep searching system of claim 1, wherein the target types include cattle and sheep.
6. The unmanned aerial vehicle mountain sheep searching system of claim 1, wherein the unmanned aerial vehicle subsystem further comprises an active stimulation module for driving the airborne speaker to emit driving sound to drive the target in the image back to the destination.
7. The unmanned mountain sheep searching system according to claim 6, wherein the target detection model processes the captured image based on yoloV4 algorithm; if the detected confidence of the object in the shot image is larger than a first preset confidence threshold value, automatically shooting and storing position information; if the confidence coefficient of the detected object is larger than a second preset confidence coefficient threshold value, automatically triggering an alarm device to remind ground station workers, and simultaneously triggering an active stimulation module to carry out automatic driving and automatic tracking; the second preset confidence threshold is greater than the first preset confidence threshold.
8. The unmanned mountain sheep searching system according to claim 7, wherein the training process of the target detection model comprises the following steps:
gather a plurality of sections videos of the different backgrounds of different angles that ordinary aerial photography unmanned aerial vehicle shot, take out frame processing to the video, extract a certain amount of sample picture, send into yolo original algorithm and detect in batches, carry out data processing to the testing result, select the target that satisfies following condition simultaneously: (1) detecting targets with the target frame area larger than 100 unit pixels, (2) detecting targets with the type of cattle and sheep, carrying out batch labeling, carrying out manual screening, adding targets which cannot be detected by a labeling script, deleting targets with detection errors, correcting the target frame, and generating a training data set;
putting the marked training data set into an image enhancement module, performing rotation, distortion, stretching and inversion processing on the sample pictures, expanding the number of the shot sample pictures, and correspondingly transforming the mark frame to obtain a complete target detection data set;
sending the target detection data set into a YOLO algorithm for training, and setting relevant parameters of a detection network by combining shooting parameters of an unmanned aerial vehicle, processing speed of an airborne computer and the class characteristics of a detection target; and continuously adjusting the learning rate in the training process to obtain a target detection model.
9. The unmanned aerial vehicle mountainous area sheep searching system of claim 1, wherein the unmanned aerial vehicle subsystem further comprises a 4G communication module; when the unmanned aerial vehicle picture-transmitted signal is not good, the shot image, the detection result and the current position coordinate of the unmanned aerial vehicle are packaged into a ros message, and the message is stored in a public message pool by adopting a publishing-subscribing mechanism, so that the ground terminal system can obtain the position of the cattle and sheep by subscribing the message.
10. An operation method of the unmanned aerial vehicle mountain sheep searching system based on any one of claims 1-9, the operation method comprising the following steps:
s1, configuring global project parameters including expected confidence, ground clearance, flight speed and active stimulation parameters;
s2, the unmanned aerial vehicle receives the remote control signal, checks the GPS signal, the running state of the onboard computer and the automatic takeoff after the image transmission channel is stable, and hovers;
s3, automatically connecting the ground terminal system with an onboard computer of the unmanned aerial vehicle, manually framing a search area on a terminal satellite map of a display screen by a user, generating a search instruction, and transmitting the search instruction to a path planning module of the onboard computer, so that the path planning module automatically plans an S-shaped flight route according to a search range;
s4, controlling the unmanned aerial vehicle to patrol along the hillside according to the planned flight route, and simultaneously carrying out target detection on the shot ground picture image:
when the pictures are shot and the suspected lost cattle and sheep pictures exist, the unmanned aerial vehicle automatically changes the circuit to approach the detection; if the confidence coefficient of the detected cattle and sheep is more than or equal to 50 percent after approaching, automatically photographing and recording, recording the position of the unmanned aerial vehicle at the moment, packaging the data into ros information, transmitting the ros information back to the ground terminal system through a second image transmission device, and requesting ground further processing;
and S5, driving according to the driving instruction sent by the herdsman.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111251088.0A CN114115359B (en) | 2021-10-26 | 2021-10-26 | Unmanned aerial vehicle mountain area sheep searching system and working method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111251088.0A CN114115359B (en) | 2021-10-26 | 2021-10-26 | Unmanned aerial vehicle mountain area sheep searching system and working method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114115359A true CN114115359A (en) | 2022-03-01 |
| CN114115359B CN114115359B (en) | 2024-06-14 |
Family
ID=80377135
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111251088.0A Active CN114115359B (en) | 2021-10-26 | 2021-10-26 | Unmanned aerial vehicle mountain area sheep searching system and working method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114115359B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116389693A (en) * | 2023-06-02 | 2023-07-04 | 中国科学院空天信息创新研究院 | Automatic grassland livestock number monitoring device and method based on unmanned aerial vehicle aerial photography |
| CN117295009A (en) * | 2023-10-07 | 2023-12-26 | 广州精天信息科技股份有限公司 | Communication equipment deployment method and device, storage medium and intelligent terminal |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105929850A (en) * | 2016-05-18 | 2016-09-07 | 中国计量大学 | Unmanned plane system and method with capabilities of continuous locking and target tracking |
| CN109960275A (en) * | 2018-08-30 | 2019-07-02 | 中宇航通（北京）科技有限公司 | A kind of unmanned plane search and rescue system and method |
| CN110658852A (en) * | 2019-09-16 | 2020-01-07 | 苏州米龙信息科技有限公司 | Intelligent target searching method and system for unmanned aerial vehicle |
| CN110929589A (en) * | 2019-10-31 | 2020-03-27 | 浙江大华技术股份有限公司 | Method, device, computer device and storage medium for vehicle feature recognition |
| CN111178148A (en) * | 2019-12-06 | 2020-05-19 | 天津大学 | Ground target geographic coordinate positioning method based on unmanned aerial vehicle vision system |
| CN112764433A (en) * | 2020-12-25 | 2021-05-07 | 湖北航天飞行器研究所 | Device and method for searching unmanned aerial vehicle-mounted rocket debris based on deep learning |
| CN113538391A (en) * | 2021-07-25 | 2021-10-22 | 吉林大学 | Photovoltaic defect detection method based on Yolov4 and thermal infrared image |
-
2021
- 2021-10-26 CN CN202111251088.0A patent/CN114115359B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105929850A (en) * | 2016-05-18 | 2016-09-07 | 中国计量大学 | Unmanned plane system and method with capabilities of continuous locking and target tracking |
| CN109960275A (en) * | 2018-08-30 | 2019-07-02 | 中宇航通（北京）科技有限公司 | A kind of unmanned plane search and rescue system and method |
| CN110658852A (en) * | 2019-09-16 | 2020-01-07 | 苏州米龙信息科技有限公司 | Intelligent target searching method and system for unmanned aerial vehicle |
| CN110929589A (en) * | 2019-10-31 | 2020-03-27 | 浙江大华技术股份有限公司 | Method, device, computer device and storage medium for vehicle feature recognition |
| CN111178148A (en) * | 2019-12-06 | 2020-05-19 | 天津大学 | Ground target geographic coordinate positioning method based on unmanned aerial vehicle vision system |
| CN112764433A (en) * | 2020-12-25 | 2021-05-07 | 湖北航天飞行器研究所 | Device and method for searching unmanned aerial vehicle-mounted rocket debris based on deep learning |
| CN113538391A (en) * | 2021-07-25 | 2021-10-22 | 吉林大学 | Photovoltaic defect detection method based on Yolov4 and thermal infrared image |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116389693A (en) * | 2023-06-02 | 2023-07-04 | 中国科学院空天信息创新研究院 | Automatic grassland livestock number monitoring device and method based on unmanned aerial vehicle aerial photography |
| CN116389693B (en) * | 2023-06-02 | 2023-08-08 | 中国科学院空天信息创新研究院 | Automatic grassland livestock number monitoring device and method based on unmanned aerial vehicle aerial photography |
| CN117295009A (en) * | 2023-10-07 | 2023-12-26 | 广州精天信息科技股份有限公司 | Communication equipment deployment method and device, storage medium and intelligent terminal |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114115359B (en) | 2024-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210217148A1 (en) | Methods for agronomic and agricultural monitoring using unmanned aerial systems | |
| US10928829B2 (en) | Detection of traffic dynamics and road changes in autonomous driving | |
| US11365014B2 (en) | System and method for automated tracking and navigation | |
| CN107358796B (en) | Vehicle detection method based on unmanned aerial vehicle | |
| KR101536095B1 (en) | Grassland management system using drone | |
| KR101793509B1 (en) | Remote observation method and system by calculating automatic route of unmanned aerial vehicle for monitoring crops | |
| CN114115359B (en) | Unmanned aerial vehicle mountain area sheep searching system and working method thereof | |
| KR20170138225A (en) | The real time national imagery intelligence system or method based on the drone which hovering or surveillance flight around the drone docking station | |
| CN105046909A (en) | Agricultural loss assessment assisting method based on small-sized unmanned aerial vehicle | |
| KR20190000439A (en) | Unmanned air vehicle for birds control and operating method by using the same | |
| CN108983809A (en) | The method and unmanned plane of accurate identification positioning surrounding based on unmanned plane | |
| CN115493594A (en) | Patrol control system and method based on BIM | |
| Chakraborty et al. | Development of UAV based glacial lake outburst monitoring system | |
| TWI771231B (en) | Sensing system, sensing data acquisition method and control device | |
| CN115984719A (en) | Forest region hidden danger identification method and equipment based on unmanned aerial vehicle | |
| TWI813077B (en) | Driving control system, control method, and control device | |
| US10959058B1 (en) | Object tracking systems and methods | |
| CN109815880A (en) | Water hyacinth identifying system and method based on low altitude remote sensing image and deep learning | |
| CN107390707A (en) | Unmanned aerial vehicle control system | |
| CN207281593U (en) | Unmanned aerial vehicle control system | |
| US20230043836A1 (en) | Information processing apparatus, information processing method, and program | |
| JP7307867B1 (en) | Drones and delivery systems | |
| US20240071237A1 (en) | Information processing apparatus, method, and non-transitory computer readable medium | |
| KR102512529B1 (en) | Method and apparatus of operating and managing unmanned golf course | |
| CN114265429A (en) | Unmanned aerial vehicle control method and device, unmanned aerial vehicle and 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 |