CN117437385A - Agricultural pest identification system and method based on visual analysis - Google Patents
Agricultural pest identification system and method based on visual analysis Download PDFInfo
- Publication number
- CN117437385A CN117437385A CN202311765601.7A CN202311765601A CN117437385A CN 117437385 A CN117437385 A CN 117437385A CN 202311765601 A CN202311765601 A CN 202311765601A CN 117437385 A CN117437385 A CN 117437385A
- Authority
- CN
- China
- Prior art keywords
- visual
- pest
- acquisition module
- vision
- analysis
- 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
- 230000000007 visual effect Effects 0.000 title claims abstract description 228
- 241000607479 Yersinia pestis Species 0.000 title claims abstract description 217
- 238000000034 method Methods 0.000 title claims abstract description 16
- 230000035929 gnawing Effects 0.000 claims abstract description 18
- 230000004438 eyesight Effects 0.000 claims description 154
- 238000004891 communication Methods 0.000 claims description 17
- 241000238631 Hexapoda Species 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000005856 abnormality Effects 0.000 claims description 3
- NJPPVKZQTLUDBO-UHFFFAOYSA-N novaluron Chemical compound C1=C(Cl)C(OC(F)(F)C(OC(F)(F)F)F)=CC=C1NC(=O)NC(=O)C1=C(F)C=CC=C1F NJPPVKZQTLUDBO-UHFFFAOYSA-N 0.000 claims 1
- 238000011835 investigation Methods 0.000 description 3
- RZVHIXYEVGDQDX-UHFFFAOYSA-N 9,10-anthraquinone Chemical compound C1=CC=C2C(=O)C3=CC=CC=C3C(=O)C2=C1 RZVHIXYEVGDQDX-UHFFFAOYSA-N 0.000 description 2
- 241000237858 Gastropoda Species 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 241000894007 species Species 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004313 glare Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64U—UNMANNED AERIAL VEHICLES [UAV]; EQUIPMENT THEREFOR
- B64U20/00—Constructional aspects of UAVs
- B64U20/80—Arrangement of on-board electronics, e.g. avionics systems or wiring
- B64U20/87—Mounting of imaging devices, e.g. mounting of gimbals
-
- 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
- G06V10/757—Matching configurations of points or features
-
- 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
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- 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
- 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
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64U—UNMANNED AERIAL VEHICLES [UAV]; EQUIPMENT THEREFOR
- B64U2101/00—UAVs specially adapted for particular uses or applications
- B64U2101/30—UAVs specially adapted for particular uses or applications for imaging, photography or videography
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Remote Sensing (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Mechanical Engineering (AREA)
- Aviation & Aerospace Engineering (AREA)
- Catching Or Destruction (AREA)
Abstract
The invention discloses an agricultural pest identification system and method based on visual analysis, and relates to the technical field of crop pest visual analysis. The visual analysis module introduces the second visual recognition of the gnawing surface features on the basis of the visual analysis of the pest features, and is matched with the judgment of the common pest types of crops on the basis that the primary pest recognition result is consistent with the second judgment result, if one of the common pest types is the common pest type, the visual recognition result can be determined to be the common pest type, the accuracy of the visual recognition of the crops can be greatly improved, and more accurate and comprehensive pest data can be provided for pest control of the crops.
Description
Technical Field
The invention relates to the technical field of visual analysis of crop pests, in particular to an agricultural pest identification system and method based on visual analysis.
Background
Crop pests are the culprit of crop losses. Different control strategies are provided for different pests, and accurate identification of the pests becomes a key point of pest control and is also a great difficulty in agriculture.
The existing crop insect condition visual monitoring mainly adopts a lamplight trapping mode to carry out visual shooting on various trapped insect bodies, carries out insect condition analysis by adopting a visual algorithm according to the shot visual picture, and then formulates a control strategy according to an analysis result.
At present, crop pest identification is mainly based on feature identification and extraction of collected pest images, and then based on a pest feature database feature matching mode, species identification of pest is realized, and the identification success rate of the traditional visual identification algorithm is low, mainly for the following reasons:
1. illumination condition influence: under different illumination conditions, the image characteristics of pests can be greatly changed, so that the recognition effect of an algorithm is affected. For example, glare or shadows can cause a decrease in contrast of the pest image, making feature extraction and recognition more difficult.
2. Morphological variation of pests: during the growth and development of crop pests, the morphology of the pests may change greatly. Such variations may lead to a decrease in the accuracy of the algorithm in identifying different stages of pests. In addition, different species of pests may have different morphological characteristics, which also increases the complexity of the recognition algorithm.
3. Background complexity: crop pest images are often affected by background noise, such as extraneous information from soil, weeds, foliage, etc. These noises can lead to erroneous decisions by the algorithm in identifying pests. The higher the background complexity, the higher the computational resources and accuracy requirements required for the recognition algorithm.
Visual analysis results for crop pests are generally based only on a relatively high probability of analysis, resulting in a visual recognition rate of the pests that is not high.
Therefore, a visual recognition system and a matched visual recognition method for crop pests with more accurate recognition rate are required to be developed, and more accurate data is provided for agricultural pest control.
Disclosure of Invention
In order to solve the technical problems, the invention provides an agricultural pest identification system and method based on visual analysis. The following technical scheme is adopted:
the agricultural pest identification system based on visual analysis comprises a first visual acquisition module, a second visual acquisition module, a third visual acquisition module, a visual analysis module, a database module and an analysis result output module, wherein the first visual acquisition module is used for acquiring visual pictures with complete pests, the second visual acquisition module is used for acquiring visual pictures with complete leaves, the first visual acquisition module, the second visual acquisition module and the third visual acquisition module are respectively in communication connection with the visual analysis module, the database module stores pest feature databases, pest gnawing leaf feature databases and crop feature databases, the visual analysis module is respectively in communication connection with the database module and the analysis result output module, at least one complete pest visual picture acquired based on the first visual acquisition module, at least one complete leaf visual picture acquired by the second visual acquisition module and at least one complete crop visual picture acquired by the third visual acquisition module are respectively connected, the visual analysis module invokes data of the database module, and the visual analysis module adopts a visual feature similarity algorithm to perform pest feature similarity matching based on the complete pest visual pictures and the pest feature databases, and if the pest feature similarity matching is successful, and initial pest identification results are output; performing second similarity matching based on the complete foliar visual image and the foliar characteristic database of pest gnawing to obtain a second pest identification result; if the second pest identification result is consistent with the primary pest identification result, performing third similarity matching based on the complete crop visual image obtained by the third visual acquisition module and a crop feature database; when the similarity of the three features is judged to be matched, the visual analysis result of the successfully matched pests is output through the analysis result output module.
By adopting the technical scheme, the method is different from the conventional visual recognition method in that only the characteristic matching of the insect body is focused, the visual analysis module firstly performs pest characteristic matching based on the complete insect visual image and the pest characteristic database, if the matching is successful, a preliminary pest recognition result is obtained, however, a certain false recognition rate exists in the result, so that the second judgment is required further based on the complete leaf visual image and the pest gnawing leaf characteristic database, because different pests gnawing leaves or rod diameters have obvious characteristics, for example, snails can gnawing small holes with uneven edges on tender leaves, and the like, if the damage of the pests reaches a certain degree, corresponding eating characteristics are required for the complete leaf visual image, so that the pest recognition result can be obtained through the matching of the pest gnawing leaf characteristics, if the preliminary pest recognition result is consistent with the second judgment result, the pest recognition result is basically positive and accurate, if the primary pest recognition result is inconsistent, the false judgment is possible, and at the moment, workers are required to be reminded to intervene manually or to be recognized again;
the third judgment can be performed on the basis of the two judgments, the third judgment is performed on the basis of the complete crop visual image and the crop characteristic database which are obtained by the third visual acquisition module, the characteristic images of different crops and corresponding common pest type data are stored in the crop characteristic database, the types of target crops can be identified according to the complete crop visual image, further the common pest type data of the crops are obtained, the first and second judging consistent pest types are compared, if one of the pest types is the pest type, the visual identification result can be determined, if not, workers are required to be reminded to perform field investigation, whether the possibility of unknown damages of the unusual pests exists or not is provided, the accuracy of crop visual identification can be greatly improved, and more accurate and comprehensive pest data are provided for pest control of the crops.
Optionally, the device further comprises a flying body, and the first vision acquisition module, the second vision acquisition module and the third vision acquisition module are respectively installed at the bottom of the flying body.
Optionally, still include electronic universal cloud platform and shoot the controller, the base mounting of electronic universal cloud platform is at the bottom load mount portion of the body of flight, the movable part at electronic universal cloud platform is installed respectively to first vision acquisition module, second vision acquisition module and third vision acquisition module, shoot the execution action of controller control electronic universal cloud platform, first vision acquisition module, second vision acquisition module and third vision acquisition module.
Optionally, the shooting controller includes vision analysis chip and shooting control chip, vision analysis chip respectively with first vision acquisition module, second vision acquisition module and third vision acquisition module communication connection, vision analysis chip respectively analyzes the vision picture of first vision acquisition module, second vision acquisition module and third vision acquisition module, shooting control chip respectively with the flying body, electronic universal cloud platform, first vision acquisition module, second vision acquisition module and third vision acquisition module control connection, the flying body is located the flight hover position at crops top according to the vision analysis result control of vision analysis chip, the movable part action of electronic universal cloud platform is controlled, make first vision acquisition module gather at least one complete worm body vision picture, make second vision acquisition module gather at least one complete leaf surface vision picture, make third vision acquisition module gather at least one complete crops vision picture.
Optionally, the first vision acquisition module, the second vision acquisition module and the third vision acquisition module are vision cameras, and the flying body is a four-rotor unmanned aerial vehicle.
Through adopting above-mentioned technical scheme, adopt four rotor unmanned aerial vehicle to realize nimble shooting task, can be through the change of flight altitude and position, at last under the control of the electronic universal cloud platform movable part of cooperation under hovering state, make first vision acquire module, second vision acquire module and third vision acquire the module obtain the shooting result that satisfies the settlement condition, specific control logic can be through the preliminary analysis result of vision analysis chip to the shooting picture carry out shooting control, make the visual picture that follow-up shooting obtained accord with visual recognition demand more, help promoting the rate of accuracy of discernment.
Optionally, the visual analysis module comprises a data input module, a visual data memory and a visual analysis computer, wherein the data input module is respectively in communication connection with the data output ends of the first visual acquisition module, the second visual acquisition module and the third visual acquisition module, and stores the acquired visual data in the visual data memory, the visual analysis computer is respectively in communication connection with the visual data memory, the database module and the analysis result output module, invokes the visual data and the database data of the visual data memory to perform pest visual analysis, and outputs the analysis result through the analysis result output module.
Optionally, the analysis result output module is a display screen.
Through adopting above-mentioned technical scheme, the data input module can be wireless communication's mode with the connected mode of first vision acquisition module, second vision acquisition module and third vision acquisition module, can realize wireless interactive visual data of shooting through four rotor unmanned aerial vehicle's flight control system's wireless transmission module, finally enter visual data memory through data input module, visual analysis computer recall visual data memory's data and database module's data carry out pest visual analysis, visual analysis computer preassembles common visual processing software instruments such as visual feature recognition algorithm, the analysis result of final pest can be exported through the display screen.
The agricultural pest identification method based on visual analysis adopts an agricultural pest identification system based on visual analysis to carry out visual identification on crop pests, and comprises the following steps:
step 1, a flying body flies above a target crop, a shooting control chip controls a movable part of an electric universal cradle head to rotate, so that a third vision acquisition module shoots the whole view right above the target crop, and shooting is carried out for the first time to obtain a picture which is recorded as Pca; controlling the flying body to descend the flying height again, controlling the movable part of the electric universal cradle head to rotate, enabling the second vision acquisition module to be opposite to the leaf surface of the target crop, and shooting for the second time to obtain a picture which is recorded as Pba; finally, controlling the movable part of the electric universal holder to rotate, so that the shooting angles of the first vision acquisition module and the second vision acquisition module are the same, shooting is carried out for the third time, and the obtained picture is recorded as Paa;
step 2, inputting Pca, pba and Paa into a visual data memory through a data input module respectively, firstly carrying out insect characteristic frame selection and identification on Paa based on a visual characteristic identification algorithm by a visual analysis computer, carrying out characteristic similarity calculation on Paa based on a pest characteristic database, recording a pest characteristic matching result with the largest characteristic similarity as Sa, setting a pest characteristic similarity threshold Sax, judging that data based on the pest characteristic database output a preliminary pest judgment result if Sa is larger than Sax, continuously analyzing Pba, carrying out leaf characteristic frame selection and identification on Pba based on a visual characteristic identification algorithm by the visual analysis computer, carrying out similarity calculation on the leaf characteristic database based on the pest, recording a pest characteristic matching result with the largest characteristic similarity as Sb, setting a pest leaf feature similarity threshold Sbx, outputting a pest judgment result based on the pest leaf feature according to the pest leaf feature database, comparing the pest judgment result based on the leaf feature with the preliminary pest judgment result if Sb is larger than Sbx, and displaying the pest identification result through an analysis result output module if the pest judgment result is consistent.
Optionally, in step 1, after the first shooting is completed by the third vision acquisition module, the third vision acquisition module sends the Pca to the vision analysis chip, the vision analysis chip analyzes the Pca based on the vision feature recognition algorithm, if the target crop picture is judged to be complete, the analysis result is interacted with the shooting control chip, if the target crop picture is judged to be incomplete, the shooting control chip controls the flying body to change the flying height, shooting is performed again until the target crop picture is judged to be complete, the second vision acquisition module is controlled to perform the second shooting, and the shooting control logic of the second vision acquisition module and the shooting control logic of the first vision acquisition module are consistent with the third vision acquisition module.
Optionally, in step 2, if the pest judgment result based on the gnawing characteristics is consistent with the preliminary pest judgment result, continuously analyzing the Pca, performing crop characteristic frame selection and recognition on the Pca by the visual analysis computer based on the visual characteristic recognition algorithm, and performing matching based on the common pest types of the corresponding crops stored in the crop characteristic database, if the matching is successful, displaying the pest recognition result through the analysis result output module, and if the matching is unsuccessful, displaying the recognition abnormality.
In summary, the present invention includes at least one of the following beneficial technical effects:
the invention can provide an agricultural pest identification system and method based on visual analysis, wherein a visual analysis module firstly carries out pest characteristic matching based on an integral pest visual image and a pest characteristic database, if the matching is successful, a preliminary pest identification result is obtained, then a second judgment is carried out based on the integral pest visual image and the pest feeding leaf characteristic database, if the preliminary pest identification result is consistent with the second judgment result, a third judgment is carried out, the third judgment is carried out based on the integral crop visual image and a crop characteristic database obtained by a third visual acquisition module, the characteristic images of different crops and corresponding common pest type data are stored in the crop characteristic database, if one of the characteristic images and the corresponding common pest type data is used, the visual identification result is the pest, if not, workers are required to be reminded of carrying out field investigation, whether unknown damages of unusual pests exist or not is possible, the accuracy of the visual identification of crops can be greatly improved, and more accurate and comprehensive pest data are provided for pest control of crops.
Drawings
FIG. 1 is a schematic diagram of the electrical device connection principle of the visual analysis-based agricultural pest identification system of the present invention;
FIG. 2 is a schematic diagram of the structural principle of the agricultural pest identification system based on visual analysis of the present invention;
fig. 3 is a schematic structural view of three visual acquisition modules of the agricultural pest recognition system based on visual analysis of the invention, which are arranged on a flying body.
Reference numerals illustrate: 1. a first vision acquisition module; 2. a second vision acquisition module; 3. a third vision acquisition module; 4. a visual analysis module; 41. a data input module; 42. a visual data store; 43. a visual analysis computer; 5. a database module; 6. an analysis result output module; 7. a flying body; 8. electric universal cradle head; 9. a photographing controller; 91. a visual analysis chip; 92. and shooting the control chip.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
The embodiment of the invention discloses an agricultural pest identification system and method based on visual analysis.
Referring to fig. 1-3, embodiment 1, an agricultural pest identification system based on visual analysis includes a first visual acquisition module 1 for acquiring visual images of whole pests, a second visual acquisition module 2 for acquiring visual images of whole pests, a third visual acquisition module 3 for acquiring visual images of whole pests, a visual analysis module 4, a database module 5 and an analysis result output module 6, wherein the first visual acquisition module 1, the second visual acquisition module 2 and the third visual acquisition module 3 are respectively in communication connection with the visual analysis module 4, the database module 5 stores pest feature databases, pest gnawing blade feature databases and crop feature databases, the visual analysis module 4 is respectively in communication connection with the database module 5 and the analysis result output module 6, and based on at least one whole pest visual image acquired by the first visual acquisition module 1, at least one whole pest visual image acquired by the second visual acquisition module 2 and at least one whole pest visual image acquired by the third visual acquisition module 3, the visual analysis module 4 is invoked with data of the database module 5, and the visual analysis module 4 adopts a visual feature similarity algorithm to perform preliminary pest feature matching based on the whole pest feature databases if the pest feature databases are successfully matched; performing second similarity matching based on the complete foliar visual image and the foliar characteristic database of pest gnawing to obtain a second pest identification result; if the second pest identification result is consistent with the primary pest identification result, performing third similarity matching based on the complete crop visual image obtained by the third visual acquisition module and a crop feature database; when the similarity of the three features is judged to be matched, a visual analysis result of the matched pest is output through the analysis result output module 6.
Different from the conventional visual recognition method, the emphasis is only placed on the characteristic matching of the insect body, the visual analysis module 4 performs pest characteristic matching based on the complete insect visual image and the pest characteristic database, if the matching is successful, a preliminary pest recognition result is obtained, however, a certain false recognition rate exists in the result, so that the second judgment is required to be further performed based on the complete leaf visual image and the pest gnawing leaf characteristic database, because different pests gnawing leaf surfaces or rod diameters have obvious characteristics, for example snails can gnawing small holes with uneven edges on tender leaves, and for the complete leaf visual image, if the damage of the pests reaches a certain degree, corresponding gnawing characteristics are required, so that the pest recognition result can be obtained through the matching of the pest gnawing leaf surface characteristics, if the preliminary pest recognition result is consistent with the second judgment result, the pest recognition result is basically accurate, if the primary pest recognition result is inconsistent, the false judgment possibility exists at the moment, and the workers are required to be reminded of manual intervention, or recognition is performed again;
the third judgment can be performed on the basis of the two judgments, the third judgment is performed on the basis of the complete crop visual image and the crop characteristic database which are obtained by the third visual acquisition module 3, the characteristic images of different crops and corresponding common pest type data are stored in the crop characteristic database, the types of target crops can be identified according to the complete crop visual image, further the common pest type data of the crops are obtained, the first and second judging consistent pest types are compared, if one of the pest types is the pest type, the visual identification result can be determined to be the pest type, if not, workers are required to be reminded of carrying out field investigation, whether the possibility of unknown damages of unusual pests exists or not is provided, the accuracy of crop visual identification can be greatly improved, and more accurate and comprehensive pest data are provided for pest control of the crops.
Embodiment 2 further comprises a flying body 7, and the first vision acquisition module 1, the second vision acquisition module 2 and the third vision acquisition module 3 are respectively mounted at the bottom of the flying body 7.
Embodiment 3 still includes electronic universal cloud platform 8 and shooting controller 9, and electronic universal cloud platform 8's base mounting is at the bottom load mount portion of flight body 7, and first vision acquisition module 1, second vision acquisition module 2 and the movable part of installing at electronic universal cloud platform 8 respectively of third vision acquisition module 3, shooting controller 9 control electronic universal cloud platform 8, first vision acquisition module 1, second vision acquisition module 2 and the execution action of third vision acquisition module 3.
In embodiment 4, the shooting controller 9 includes a vision analysis chip 91 and a shooting control chip 92, the vision analysis chip 91 is respectively in communication connection with the first vision acquisition module 1, the second vision acquisition module 2 and the third vision acquisition module 3, the vision analysis chip 91 respectively analyzes the vision pictures of the first vision acquisition module 1, the second vision acquisition module 2 and the third vision acquisition module 3, the shooting control chip 92 is respectively in control connection with the flying body 7, the electric universal pan-tilt 8, the first vision acquisition module 1, the second vision acquisition module 2 and the third vision acquisition module 3, the flying body 7 is controlled to be located at the flying hovering position of the top of the crop according to the vision analysis result of the vision analysis chip 91, the action of the movable part of the electric universal pan-tilt 8 is controlled, so that the first vision acquisition module 1 acquires at least one complete insect vision picture, the second vision acquisition module 2 acquires at least one complete leaf surface vision picture, and the third vision acquisition module 3 acquires at least one complete crop vision picture.
In embodiment 5, the first vision acquisition module 1, the second vision acquisition module 2, and the third vision acquisition module 3 are all vision cameras, and the flying body 7 is a quadrotor unmanned plane.
The four-rotor unmanned aerial vehicle is adopted to realize flexible shooting tasks, the change of flying height and position can be realized, finally, the control of the movable part of the electric universal cradle head 8 is matched under the hovering state, so that the first vision acquisition module 1, the second vision acquisition module 2 and the third vision acquisition module 3 can obtain shooting results meeting set conditions, a specific control logic can carry out shooting control on preliminary analysis results of shooting pictures through the vision analysis chip 91, the vision pictures obtained through follow-up shooting are more in accordance with vision identification requirements, and the accuracy of identification is improved.
In embodiment 6, the visual analysis module 4 includes a data input module 41, a visual data memory 42 and a visual analysis computer 43, the data input module 41 is respectively connected with the data output ends of the first visual acquisition module 1, the second visual acquisition module 2 and the third visual acquisition module 3 in a communication manner, and the acquired visual data is stored in the visual data memory 42, the visual analysis computer 43 is respectively connected with the visual data memory 42, the database module 5 and the analysis result output module 6 in a communication manner, and the visual data and the database data of the visual data memory 42 are called for pest visual analysis, and the analysis result is output through the analysis result output module 6.
The analysis result output module 6 is a display screen.
The connection mode of the data input module 41 and the first visual acquisition module 1, the second visual acquisition module 2 and the third visual acquisition module 3 can be a wireless communication mode, visual data of wireless interaction shooting can be realized through a wireless transmission module of a flight control system of the quad-rotor unmanned aerial vehicle, finally the visual data are input into the visual data memory 42 through the data input module 41, the visual analysis computer 43 recalls the data of the visual data memory 42 and the data of the database module 5 to carry out pest visual analysis, the visual analysis computer 43 is preloaded with common visual processing software tools such as a visual feature recognition algorithm, and finally the analysis result of pests can be output through a display screen.
Embodiment 7, an agricultural pest identification method based on visual analysis, which performs visual identification of crop pests using an agricultural pest identification system based on visual analysis, comprising the steps of:
step 1, the flying body 7 flies above the target crops, the shooting control chip 92 controls the movable part of the electric universal cradle head 8 to rotate, so that the third vision acquisition module 3 shoots the whole view right above the target crops, and the first shooting is carried out, and the obtained picture is recorded as Pca; controlling the flying body 7 to descend the flying height again, controlling the movable part of the electric universal cradle head 8 to rotate, enabling the second vision acquisition module 2 to be opposite to the leaf surface of the target crop, and shooting for the second time to obtain a picture which is recorded as Pba; finally, controlling the movable part of the electric universal holder 8 to rotate, so that the shooting angles of the first vision acquisition module 1 and the second vision acquisition module 2 are the same, and shooting for the third time to obtain a picture which is recorded as Paa;
step 2, pca, pba and Paa are respectively input into a visual data memory 42 through a data input module 41, a visual analysis computer 43 firstly performs insect characteristic frame selection and identification on Paa based on a visual characteristic identification algorithm, performs characteristic similarity calculation on the basis of a pest characteristic database, marks the pest characteristic matching result with the largest characteristic similarity as Sa, sets a pest characteristic similarity threshold Sax, judges that data based on the pest characteristic database outputs a preliminary pest judgment result if Sa > Sax, continuously analyzes Pba, the visual analysis computer 43 performs leaf characteristic frame selection and identification on Pba based on a visual characteristic identification algorithm, performs similarity calculation on the basis of the pest feeding leaf characteristic database, marks the pest characteristic matching result with the largest characteristic similarity as Sb, sets a pest feeding leaf characteristic similarity threshold Sbx, and if Sb > Sbx, outputs a pest judgment result based on the feeding characteristic according to the pest feeding leaf characteristic database, compares the pest judgment result based on the feeding characteristic with the preliminary pest judgment result, and if the pest judgment result is consistent, and displays the pest identification result through an analysis result output module 6.
In embodiment 8, in step 1, after the third vision acquisition module 3 completes the first shooting, the Pca is sent to the vision analysis chip 91, the vision analysis chip 91 analyzes the Pca based on the vision feature recognition algorithm, if it is judged that the target crop picture is complete, the analysis result is interacted with the shooting control chip 92, if it is judged that the target crop picture is incomplete, the shooting control chip 92 controls the flying body 7 to change the flying height, shooting is performed again, until it is judged that the target crop picture is complete, the second vision acquisition module 2 is controlled to perform the second shooting, and the shooting control logic of the second vision acquisition module 2 and the first vision acquisition module 1 is consistent with the third vision acquisition module 3.
In embodiment 9, in step 2, if the pest judgment result based on the feeding feature is consistent with the preliminary pest judgment result, the analysis of Pca is continued, the visual analysis computer 43 performs crop feature frame selection and recognition on Pca based on the visual feature recognition algorithm, and performs matching based on the common pest types of the corresponding crops stored in the crop feature database, if the matching is successful, the pest recognition result is displayed through the analysis result output module 6, and if the matching is unsuccessful, the recognition abnormality is displayed.
The above embodiments are not intended to limit the scope of the present invention, and therefore: all equivalent changes in structure, shape and principle of the invention should be covered in the scope of protection of the invention.
Claims (10)
1. Agricultural pest identification system based on visual analysis, its characterized in that: the pest recognition system comprises a first visual acquisition module (1) for acquiring visual pictures of complete pests, a second visual acquisition module (2) for acquiring visual pictures of complete crops, a third visual acquisition module (3) for acquiring visual pictures of complete leaves, a visual analysis module (4), a database module (5) and an analysis result output module (6), wherein the first visual acquisition module (1), the second visual acquisition module (2) and the third visual acquisition module (3) are respectively in communication connection with the visual analysis module (4), the database module (5) stores a pest characteristic database, a pest gnawing leaf characteristic database and a crop characteristic database, the visual analysis module (4) is respectively in communication connection with the database module (5) and the analysis result output module (6), and if the pest characteristic databases are successfully matched with each other based on the similarity of the pest characteristic database, the pest characteristic database is firstly matched with the pest characteristic database, and the pest characteristic database is initially output; performing second similarity matching based on the complete foliar visual image and the foliar characteristic database of pest gnawing to obtain a second pest identification result; if the second pest identification result is consistent with the primary pest identification result, performing third similarity matching based on the complete crop visual image obtained by the third visual acquisition module and a crop feature database; when the similarity of the three features is judged to be matched, a visual analysis result of the matched pests is output through an analysis result output module (6).
2. The visual analysis-based agricultural pest identification system according to claim 1, wherein: the device further comprises a flying body (7), wherein the first vision acquisition module (1), the second vision acquisition module (2) and the third vision acquisition module (3) are respectively arranged at the bottom of the flying body (7).
3. The visual analysis-based agricultural pest identification system according to claim 2, wherein: still include electronic universal cloud platform (8) and shoot controller (9), the pedestal mounting of electronic universal cloud platform (8) is at the bottom load mount portion of the body of flight (7), install the movable part at electronic universal cloud platform (8) respectively in first vision acquisition module (1), second vision acquisition module (2) and third vision acquisition module (3), shoot controller (9) control electronic universal cloud platform (8), first vision acquisition module (1), second vision acquisition module (2) and the executive action of third vision acquisition module (3).
4. The visual analysis-based agricultural pest identification system according to claim 3, wherein: the shooting controller (9) comprises a vision analysis chip (91) and a shooting control chip (92), the vision analysis chip (91) is respectively in communication connection with the first vision acquisition module (1), the second vision acquisition module (2) and the third vision acquisition module (3), the vision analysis chip (91) is respectively used for analyzing vision pictures of the first vision acquisition module (1), the second vision acquisition module (2) and the third vision acquisition module (3), the shooting control chip (92) is respectively in control connection with the flying body (7), the electric universal cradle head (8), the first vision acquisition module (1), the second vision acquisition module (2) and the third vision acquisition module (3), the flying body (7) is controlled to be located at a hovering position of the top of a crop according to a vision analysis result of the vision analysis chip (91), and a movable part of the electric universal cradle head (8) is controlled to act, so that the first vision acquisition module (1) acquires at least one complete insect vision picture, the second vision acquisition module (2) acquires at least one complete leaf surface vision picture, and the third vision acquisition module (3) acquires at least one complete crop picture.
5. The visual analysis-based agricultural pest identification system according to claim 4, wherein: the first vision acquisition module (1), the second vision acquisition module (2) and the third vision acquisition module (3) are vision cameras, and the flying body (7) is a four-rotor unmanned aerial vehicle.
6. The visual analysis-based agricultural pest identification system according to claim 5, wherein: the visual analysis module (4) comprises a data input module (41), a visual data memory (42) and a visual analysis computer (43), wherein the data input module (41) is respectively in communication connection with the data output ends of the first visual acquisition module (1), the second visual acquisition module (2) and the third visual acquisition module (3), the acquired visual data are stored in the visual data memory (42), the visual analysis computer (43) is respectively in communication connection with the visual data memory (42), the database module (5) and the analysis result output module (6), and the visual data and the database data of the visual data memory (42) are called to carry out pest visual analysis and the analysis result is output through the analysis result output module (6).
7. The visual analysis-based agricultural pest identification system according to claim 5, wherein: the analysis result output module (6) is a display screen.
8. The agricultural pest identification method based on visual analysis is characterized by comprising the following steps of: the visual analysis-based agricultural pest identification system according to claim 6, for visual identification of crop pests, comprising the steps of:
step 1, a flying body (7) flies above a target crop, a shooting control chip (92) controls a movable part of an electric universal cradle head (8) to rotate, so that a third vision acquisition module (3) shoots the whole view right above the target crop, and shooting is carried out for the first time to obtain a picture which is recorded as Pca; controlling the flying body (7) to descend the flying height again, controlling the movable part of the electric universal cradle head (8) to rotate, enabling the second vision acquisition module (2) to be opposite to the leaf surface of the target crop, and shooting for the second time to obtain a picture which is recorded as Pba; finally, controlling the movable part of the electric universal holder (8) to rotate, so that the shooting angles of the first vision acquisition module (1) and the second vision acquisition module (2) are the same, and shooting for the third time to obtain a picture which is recorded as Paa;
and 2, inputting Pca, pba and Paa into a visual data memory (42) through a data input module (41), firstly selecting and identifying the insect body characteristics of Paa based on a visual characteristic identification algorithm by a visual analysis computer (43), calculating the characteristic similarity based on a pest characteristic database, recording the pest characteristic matching result with the maximum characteristic similarity as Sa, setting a pest characteristic similarity threshold Sax, judging that the data based on the pest characteristic database outputs a preliminary pest judging result if Sax is more than Sax, continuously analyzing Pba, performing leaf surface characteristic frame selection and identification on Pba based on a visual characteristic identification algorithm by a visual analysis computer (43), performing similarity calculation based on the pest leaf surface characteristic database, recording the pest characteristic matching result with the maximum characteristic similarity as Sb, setting a pest leaf surface characteristic similarity threshold Sbx, outputting the pest judging result based on the feeding characteristics according to the pest leaf surface characteristic database, comparing the pest judging result based on the feeding characteristics with the preliminary judging result if Sb is more than Sbx, and displaying the identification result through an analysis output module (6) if the pest judging result is consistent.
9. The visual analysis-based agricultural pest identification method according to claim 8, wherein: in the step 1, after the third vision acquisition module (3) finishes shooting for the first time, the Pca is sent to the vision analysis chip (91), the vision analysis chip (91) analyzes the Pca based on a vision feature recognition algorithm, if the image of the target crop is judged to be complete, the analysis result is interacted with the shooting control chip (92), if the image of the target crop is judged to be incomplete, the shooting control chip (92) controls the flying body (7) to change the flying height, shooting is performed again until the image of the target crop is judged to be complete, the second vision acquisition module (2) is controlled to perform shooting for the second time, and shooting control logic of the second vision acquisition module (2) and the shooting control logic of the first vision acquisition module (1) are consistent with the third vision acquisition module (3).
10. The visual analysis-based agricultural pest identification method according to claim 8, wherein: in the step 2, if the pest judgment result based on the gnawing characteristics is consistent with the preliminary pest judgment result, continuously analyzing the Pca, performing crop characteristic frame selection and recognition on the Pca based on a visual characteristic recognition algorithm by a visual analysis computer (43), and performing matching based on common pest types of corresponding crops stored in a crop characteristic database, displaying the pest recognition result through an analysis result output module (6) if the matching is successful, and displaying recognition abnormality if the matching is unsuccessful.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311765601.7A CN117437385B (en) | 2023-12-21 | 2023-12-21 | Agricultural pest identification system and method based on visual analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311765601.7A CN117437385B (en) | 2023-12-21 | 2023-12-21 | Agricultural pest identification system and method based on visual analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117437385A true CN117437385A (en) | 2024-01-23 |
| CN117437385B CN117437385B (en) | 2024-04-02 |
Family
ID=89558683
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311765601.7A Active CN117437385B (en) | 2023-12-21 | 2023-12-21 | Agricultural pest identification system and method based on visual analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117437385B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109032093A (en) * | 2018-08-14 | 2018-12-18 | 荆州农业科学院 | A kind of rice pest monitoring system |
| CN109409170A (en) * | 2017-08-18 | 2019-03-01 | 广州极飞科技有限公司 | The identifying pest method and apparatus of crops |
| CN109726700A (en) * | 2019-01-07 | 2019-05-07 | 武汉南博网络科技有限公司 | A kind of identifying pest method for early warning and device based on multiple features |
| CN109840549A (en) * | 2019-01-07 | 2019-06-04 | 武汉南博网络科技有限公司 | A kind of pest and disease damage recognition methods and device |
| CN115082706A (en) * | 2022-06-29 | 2022-09-20 | 山东华丰数字科技有限公司 | Method, device and equipment for identifying plant diseases and insect pests and readable storage medium |
| US20220358742A1 (en) * | 2019-10-11 | 2022-11-10 | Hangzhou Glority Software Limited | Insect identification method and system |
-
2023
- 2023-12-21 CN CN202311765601.7A patent/CN117437385B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109409170A (en) * | 2017-08-18 | 2019-03-01 | 广州极飞科技有限公司 | The identifying pest method and apparatus of crops |
| CN109032093A (en) * | 2018-08-14 | 2018-12-18 | 荆州农业科学院 | A kind of rice pest monitoring system |
| CN109726700A (en) * | 2019-01-07 | 2019-05-07 | 武汉南博网络科技有限公司 | A kind of identifying pest method for early warning and device based on multiple features |
| CN109840549A (en) * | 2019-01-07 | 2019-06-04 | 武汉南博网络科技有限公司 | A kind of pest and disease damage recognition methods and device |
| US20220358742A1 (en) * | 2019-10-11 | 2022-11-10 | Hangzhou Glority Software Limited | Insect identification method and system |
| CN115082706A (en) * | 2022-06-29 | 2022-09-20 | 山东华丰数字科技有限公司 | Method, device and equipment for identifying plant diseases and insect pests and readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 蔡清;何东健;: "基于图像分析的蔬菜食叶害虫识别技术", 计算机应用, no. 07, 1 July 2010 (2010-07-01) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117437385B (en) | 2024-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110148120B (en) | Intelligent disease identification method and system based on CNN and transfer learning | |
| RU2764872C2 (en) | Weed detection in natural environment | |
| CN109726700B (en) | Insect pest recognition early warning method and device based on multiple features | |
| JP2019520631A (en) | Weed recognition in the natural environment | |
| CN110796011B (en) | Rice ear recognition method, system, device and medium based on deep learning | |
| CN109117857B (en) | Biological attribute identification method, device and equipment | |
| KR102393265B1 (en) | System for detecting pests of shiitake mushrooms | |
| KR101846301B1 (en) | Crops grow and management system using automatic control device | |
| KR102002632B1 (en) | Fruit monitoring system and method at the same | |
| CN110070571B (en) | Phyllostachys pubescens morphological parameter detection method based on depth camera | |
| CN115410192A (en) | Citrus huanglongbing image acquisition and identification method, storage medium and system | |
| KR20200077808A (en) | Video-based monitoring apparatus and operation method | |
| CN111339906A (en) | Image processing device and image processing system | |
| CN115661650A (en) | Farm management system based on data monitoring of Internet of things | |
| CN113627248A (en) | Method, system, lawn mower and storage medium for automatically selecting recognition model | |
| CN115376125A (en) | Target detection method based on multi-modal data fusion and in-vivo fruit picking method based on target detection model | |
| Pramudhita et al. | Strawberry Plant Diseases Classification Using CNN Based on MobileNetV3-Large and EfficientNet-B0 Architecture | |
| CN209983105U (en) | Harvester | |
| CN115861686A (en) | Litchi key growth period identification and detection method and system based on edge deep learning | |
| Mangaonkar et al. | Fruit harvesting robot using computer vision | |
| CN117437385B (en) | Agricultural pest identification system and method based on visual analysis | |
| CN110689022A (en) | Leaf matching-based image extraction method for each crop | |
| CN108090910A (en) | It is a kind of that tomato plant image segmentation algorithm outside the night room of PCNN models is simplified based on comentropy gradient | |
| CN112381028A (en) | Target feature detection method and device | |
| CN113792715B (en) | Granary pest monitoring and early warning method, device, equipment and 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 |