CN116645618B - Agricultural data processing method, system and storage medium - Google Patents
Agricultural data processing method, system and storage medium Download PDFInfo
- Publication number
- CN116645618B CN116645618B CN202310657390.9A CN202310657390A CN116645618B CN 116645618 B CN116645618 B CN 116645618B CN 202310657390 A CN202310657390 A CN 202310657390A CN 116645618 B CN116645618 B CN 116645618B
- Authority
- CN
- China
- Prior art keywords
- data
- remote sensing
- task
- data processing
- classification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 55
- 230000008447 perception Effects 0.000 claims abstract description 12
- 230000011218 segmentation Effects 0.000 claims abstract description 12
- 238000013145 classification model Methods 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims abstract description 8
- 230000001953 sensory effect Effects 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 description 8
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 6
- 241000607479 Yersinia pestis Species 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 241000238631 Hexapoda Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 239000000575 pesticide Substances 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- 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
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Remote Sensing (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Astronomy & Astrophysics (AREA)
- Image Processing (AREA)
Abstract
The application discloses an agricultural data processing method, an agricultural data processing system and a storage medium, wherein the method comprises the following steps: acquiring satellite remote sensing data, processing the satellite remote sensing data based on a first remote sensing data processing model, outputting a segmentation result based on a geographic position, and outputting a first classification result of each segmentation area; determining suspicious classification results with classification probability smaller than a threshold value in the first classification results, and acquiring a region range corresponding to each suspicious classification result; acquiring sensing data of each area range in a preset time interval before and after the acquisition time from a sensing system of an access platform according to the acquisition time of the remote sensing data and each area range; and determining the correctness of the first classification result by using a second remote sensing data classification model according to the perception data.
Description
Technical Field
The application relates to a remote sensing technology and an artificial intelligence technology, in particular to an agricultural data processing method and system and a storage medium.
Background
With the development of remote sensing technology, agricultural research can analyze the agricultural development condition of a certain area outside of the thousand miles by using remote sensing data. Current remote sensing technologies, such as remote sensing satellites, can analyze the condition of a certain area, such as the growth condition of crops, whether crops have insect pests, the area where the crops are located, and the like, through multispectral shooting.
However, for the remote sensing satellite, the data of a large area can be acquired at one time, but the periodicity of the data of a certain position photographed by the remote sensing satellite is longer. If the fixed wing aircraft, the unmanned aerial vehicle and other modes are adopted to acquire the remote sensing data, the cost is higher than that of a remote sensing satellite.
Modern analysis of remote sensing data has changed from manual analysis in the past to intelligent analysis of AI models, but AI model analysis also has certain problems. AI classifies objects in an image based on features based on training, but in some cases AI has classification edges that do not classify well for some features. This will lead to some uncertainty in the analysis of the remotely sensed satellite data.
Disclosure of Invention
The present application aims to solve at least one of the technical problems existing in the prior art. To this end, the present application provides an agricultural data processing method and system and a storage medium.
In one aspect, an embodiment of the present application provides an agricultural data processing method, including:
acquiring satellite remote sensing data, processing the satellite remote sensing data based on a first remote sensing data processing model, outputting a segmentation result based on a geographic position, and outputting a first classification result of each segmentation area;
determining suspicious classification results with classification probability smaller than a threshold value in the first classification results, and acquiring a region range corresponding to each suspicious classification result;
acquiring sensing data of each area range in a preset time interval before and after the acquisition time from a sensing system of an access platform according to the acquisition time of the remote sensing data and each area range;
and determining the correctness of the first classification result by using a second remote sensing data classification model according to the perception data.
In some embodiments, according to the obtaining time of the remote sensing data and each area range, obtaining, from a sensing system of an access platform, sensing data of each area range within a preset time interval before and after the obtaining time, including:
issuing a query request to a perception system of an access platform, requesting to acquire the existing aerial photographing data in a preset time interval before and after the acquisition time, wherein the aerial photographing data relate to each regional range;
when the related partial area range does not have corresponding aerial photographing data, a data acquisition task is issued to a terminal of the accessed platform so as to obtain sensing data of each area range.
In some embodiments, the determining the correctness of the first classification result using the second telemetry data classification model based on the sensory data comprises:
and selecting a corresponding second remote sensing data processing model for classification based on the obtained data type of the number and the classification type of the first remote sensing data processing model according to the obtained existing aerial photographing data or the perception data obtained through the issuing task, obtaining a second classification result, and verifying the correctness of the first classification result according to the second classification result.
In some embodiments, the perceived data includes image data captured by both unmanned and unmanned aerial vehicles, field captured image data, fixed-point camera captured image data, and data acquired by non-image sensors.
In some embodiments, the first telemetry data processing model is configured to output a geographic location based segmentation result and to output a planting type result for each segmented region;
the second remote sensing data processing model is used for outputting classification results of whether the perception data belong to a certain planting type.
In some embodiments, a data acquisition task is issued to a terminal of an accessed platform to obtain perceived data of each region range, specifically:
when a user terminal of an access platform invokes a platform interface to process user perceived data, acquiring user terminal positioning information or an acquisition position of the perceived data, and defining an active area of the user terminal according to the terminal positioning information or the acquisition position of the perceived data;
when the active area of the user terminal overlaps with the area range corresponding to the suspicious classification result or the distance is smaller than a threshold value, creating task information aiming at the area range corresponding to the classification result;
task information corresponding to the region range corresponding to each suspicious classification result is put into a task information set;
establishing an information collection table, wherein whether the task is completed is recorded in the information collection table;
creating a plurality of delay tasks with different delay time into a message queue;
and when the delay task is consumed, pulling incomplete task information in the task information set according to the completion condition of the task in the information collection table, and issuing task pushing to a user terminal.
In some embodiments, when pushing a task to a user terminal, generating a task information acquisition template according to an optional second remote sensing data processing model, wherein the template comprises at least one data template unit required by the second remote sensing data processing model;
wherein the optional second remote sensing data processing model refers to: has been trained and resides in the server and can output a model for determining whether the first classification result is correct based on the input data type that is different from the first telemetry data processing model.
In some embodiments, the task completion in the information collection table means that the second classification result obtained by the sensory data obtained by the task is consistent with the first classification result, or the second classification result obtained by the sensory data obtained by the task is inconsistent with the first classification result and the classification probability is greater than a preset value.
In another aspect, an embodiment of the present application provides an agricultural data processing system, including:
a memory for storing a program;
and the processor is used for loading the program to execute the agricultural data processing method.
In another aspect, an embodiment of the present application provides a computer-readable storage medium storing a program that, when executed, implements an agricultural data processing method as described.
When satellite remote sensing data is processed, the region with lower classification probability, namely the region with easily confused model, is screened out according to the classification probability of the model, and then classification results of other sensing data on the regions are secondarily confirmed by utilizing other remote sensing data and a second remote sensing data processing model different from the first remote sensing data processing model; according to the scheme, the characteristic of low satellite data acquisition frequency is considered, the defects of satellite data and models are made up by using other perception data and classification models, and the effectiveness of analysis conclusion of regional remote sensing data is improved; meanwhile, the fact that the agricultural information generally does not generate mutation is considered, so that certain asynchronism exists between the sensing data and the satellite remote sensing data, the sensing data before satellite data processing can be acquired by the system, the sensing data after satellite data processing can be acquired by the issuing task, and the probability of data acquisition is increased.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the description of the embodiments will be briefly described.
FIG. 1 is a flow chart of a method for processing agricultural data according to an embodiment of the present application;
FIG. 2 is a schematic diagram of the present solution model selection provided by an embodiment of the present application;
fig. 3 is a block diagram of a system for processing agricultural data according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be clearly and completely described by means of implementation examples with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Referring to fig. 1, an embodiment of the application discloses an agricultural data processing method, which comprises the following steps:
s1, acquiring satellite remote sensing data, processing the satellite remote sensing data based on a first remote sensing data processing model, outputting a segmentation result based on a geographic position, and outputting a first classification result of each segmentation area.
The satellite remote sensing data requires that the region can be acquired only in the satellite shooting range, and therefore, the region is not in a state of being available at any time. The obtained time interval is long and uncertain, but the data area obtained by the remote sensing satellite is large, and compared with other obtaining modes, the average cost is lower and is relatively comprehensive. By means of the AI model, the data may be processed, including but not limited to dividing the areas into areas of different properties, identifying the type of crop, distinguishing the pest situation, etc. It is understood that these AI models may intelligently divide and categorize regions. It will be appreciated that there are many relatively sophisticated AI models that can be classified for the problem of processing remote sensing data. However, AI models may have classification edges, and sample models near the classification edges often cannot be classified accurately. Therefore, in classification, there are some classification results with a lower classification probability. For example, assuming that the model divides the ground into a road, a farmland and a water surface, in general, assuming that the model classifies the farmland as a result of the farmland, the classification probability of the farmland will be much greater than that of the road and the water surface, but when the sample is at the classification edges of the farmland and the water surface, the model will possibly give a conclusion that the classification is that of the farmland, but the classification probabilities of the farmland and the water surface are very close, it will be understood that the sum of the classification probabilities of the three classification results is 1, and therefore, at this time, the classification probability of the farmland is about 40 to 50%, for example, 45% of the farmland, 40% of the water surface, and 15% of the road surface.
S2, determining suspicious classification results with classification probability smaller than a threshold value in the first classification results, and acquiring a region range corresponding to each suspicious classification result.
As explained in step 1, the present step determines the results of which the classification probability is smaller than the threshold value among the classification results as suspicious classification results, and then selects the areas corresponding to these results.
S3, acquiring the sensing data of each area range in a preset time interval before and after the acquisition time from a sensing system of an access platform according to the acquisition time of the remote sensing data and each area range.
It will be appreciated that the sensing data referred to in this step refers to other telemetry data in addition to satellite data. In some embodiments, the perceived data includes image data captured by both unmanned and unmanned aerial vehicles, field captured image data, fixed-point camera captured image data, and data acquired by non-image sensors. It will be appreciated that the platform wishes to learn these data in a low cost manner to refine data in satellite telemetry that may not be accurately analyzed. Therefore, the platform provides the service, and when the user accesses the service to use the platform data, the authorized platform knows some data, so that the sharing of the data is realized. It should be understood that the satellite remote sensing data acquisition time may refer to the time of satellite data acquisition, or may refer to the time of platform acquisition. Some other type of sensory data may be acquired before and after this point in time to further confirm the more suspicious regions in the classification results. The data acquired in this step may be obtained before the satellite remote sensing data is acquired, or may be obtained after the satellite remote sensing data is acquired by a passive reporting mode or an active reporting mode of the user. The data provided by the user when using the platform service can be used, or the publishing task can be obtained by the user.
It will be appreciated that these fragmented areas, due to the smaller area, may be presented to local users for collection by unmanned aerial vehicles or collection.
S4, determining the correctness of the first classification result by using a second remote sensing data classification model according to the perception data.
It will be appreciated that the use of non-satellite telemetry data may require a secondary validation of portions of satellite telemetry data that may not give the correct classification result, and that the models used are different. For example, for a remote sensing satellite, the data it collects may be a multispectral image. For a drone, it may only acquire a traditional red, green and blue three-channel image. The resolution of the remote sensing satellite is different from that of the unmanned aerial vehicle, and the remote sensing satellite shoots at a overlook angle, and if the remote sensing satellite shoots in a collecting mode, the remote sensing satellite shoots in a head-up mode at a relatively short distance. Different data acquisition modes require different classification models for processing. It will be appreciated that the platform has stored therein models that categorize different types of data, subject to the type of data. Of course, the input and output data types of the model are also limited, and the output of the first remote sensing data processing model needs to be associated when issuing tasks and acquiring remote sensing data. For example, in some embodiments, the first telemetry data processing model is used to divide the surface into fields, buildings, roads and water, assuming that the classification probability of the appearance of a classification result into fields is relatively low. When a task is issued, unmanned aerial vehicle aerial photographing data can be obtained aiming at the area, and photographs taken by a user on site can also be obtained. At this time, the classification model used for taking the aerial photograph and the photograph taken by the user at a near point may be a special model for inputting the picture and outputting whether the classification is of a farmland, not necessarily a model for distinguishing the types of sites. It only needs the model to distinguish whether the result classified by the first telemetry data processing model is correct.
In some embodiments, according to the obtaining time of the remote sensing data and each area range, obtaining, from a sensing system of an access platform, sensing data of each area range within a preset time interval before and after the obtaining time, including:
issuing a query request to a perception system of an access platform, requesting to acquire the existing aerial photographing data in a preset time interval before and after the acquisition time, wherein the aerial photographing data relate to each regional range; typically, aerial data as it results herein refers to periodic aerial missions, such as fixed-wing aircraft, that are periodically aerial. Or the agricultural unmanned opportunities to make a tour, irrigate, and when the pesticide is applied, the data can be recorded. These data can be used as image recognition.
When the related partial area range does not have corresponding aerial photographing data, a data acquisition task is issued to a terminal of the accessed platform so as to obtain sensing data of each area range.
It can be understood that the sensing system of the access platform comprises various internet of things systems, unmanned aerial vehicle systems and an organic machine data acquisition system. Platforms are contacted by providing services to these systems, including image recognition, data statistics analysis, etc. The platform does not need to acquire all data of the systems, only needs to collect necessary data, such as abstracts of data contents, acquisition time of the data contents, acquisition places of the data contents, acquisition modes of the data contents and the like, when the platforms call services, so that the data mastered by the platforms can be approximately known, and request call can be issued when required.
In some embodiments, the determining the correctness of the first classification result using the second telemetry data classification model based on the sensory data comprises:
and selecting a corresponding second remote sensing data processing model for classification based on the obtained data type of the number and the classification type of the first remote sensing data processing model according to the obtained existing aerial photographing data or the perception data obtained through the issuing task, obtaining a second classification result, and verifying the correctness of the first classification result according to the second classification result.
Referring to fig. 2, in this embodiment, a model that can be used is selected based on the obtained type of perceived data that is different from satellite telemetry data, and the classification type of the first telemetry data processing model. For example, the classification type of the first remote sensing data processing model is classified into farmland, and the farmland of the area is obtained by unmanned aerial vehicle aerial photographing, assuming that there is an a model based on whether the aerial photographing picture is resolved in the platform, a b model based on whether the photographed picture is resolved in what type of area is collected, and a c model based on what type of image is resolved in the current area (the number of classification types may be different from that of the first remote sensing data processing model), model c and model a may be selected. The diversified mode can be suitable for occasions where the obtained data have uncertainty, and the accessed internet of things platform has the characteristics.
Of course, the above data is merely an example of an image, and in essence, in applications for identifying pests and the like, other data may be used for analysis, such as non-image recognition data such as pest characteristic voice recognition devices and the like.
In some embodiments, the first telemetry data processing model is configured to output a geographic location based segmentation result and to output a planting type result for each segmented region;
the second remote sensing data processing model is used for outputting classification results of whether the perception data belong to a certain planting type.
It will be appreciated that the above scheme may be applied to distinguish between planting types in a farm.
In some embodiments, a data acquisition task is issued to a terminal of an accessed platform to obtain perceived data of each region range, specifically:
when a user terminal of an access platform invokes a platform interface to process user perceived data, acquiring user terminal positioning information or an acquisition position of the perceived data, and defining an active area of the user terminal according to the terminal positioning information or the acquisition position of the perceived data. It can be understood that when the user terminal of the access platform invokes the platform interface to process the user perceived data, the positioning information of the user terminal can be obtained, and the positioning information can reflect the position where the user is located or the position where the user collects the data, so that the area where the user is usually active can be known. Then the task is issued in a targeted manner and such information can be more readily obtained. It will be appreciated that users of the access platform, typically in the relevant field, may increase the likelihood of task completion based on geographic location matching.
And when the active area of the user terminal is overlapped with the area range corresponding to the suspicious classification result or the distance is smaller than a threshold value, creating task information aiming at the area range corresponding to the classification result.
Task information corresponding to the region range corresponding to each suspicious classification result is put into a task information set; at this time, a plurality of users and a plurality of areas generate a set of task information. All tasks generated at the present time are collected in the collection.
An information collection table is established, and whether the task is completed is recorded in the information collection table. After at least one task corresponding to one area range is completed, all tasks associated with the area range are marked as completed.
A plurality of delay tasks with different delay times are created in a message queue. It will be appreciated that the release of tasks to completion is not typically performed at a single kick, and therefore, task pushing is typically performed on a half-day or once-a-day basis on the longest allowed days of tasks to alert the user.
And when the delay task is consumed, pulling incomplete task information in the task information set according to the completion condition of the task in the information collection table, and issuing task pushing to a user terminal. It is to be understood that the timing task plays a role in determining whether to push the task issued to the user according to the completion condition of the task in the information collection table, so that the mode can realize multiple pushing, meanwhile, the interference to the user can be reduced, and the system occupies less resources.
In the foregoing embodiments, the uncertainty of the data that can be obtained has been described, and therefore, a diversified data acquisition manner is provided as much as possible. Therefore, in these embodiments, when pushing a task to the user terminal, a task information acquisition template is generated according to the optional second remote sensing data processing model, where the template includes at least one data template unit required by the second remote sensing data processing model;
wherein the optional second remote sensing data processing model refers to: has been trained and resides in the server and can output a model for determining whether the first classification result is correct based on the input data type that is different from the first telemetry data processing model.
It will be appreciated that in this embodiment, it is necessary to analyze, based on the classification result of the first telemetry data processing model, which of the second telemetry data processing models can determine whether the classification result is correct. Task templates are then generated based on the input requirements of these models. In order to expand the possibility of data acquisition, all possible task templates are put in the task, and only one task is needed to be selected when a user finishes the task.
In some embodiments, the task completion in the information collection table means that the second classification result obtained by the sensory data obtained by the task is consistent with the first classification result, or the second classification result obtained by the sensory data obtained by the task is inconsistent with the first classification result and the classification probability is greater than a preset value.
It should be understood that the task is completed, which means that the data provided by the user has a certain degree of reliability. For example, in the first classification result, the classification probability that a certain area is farmland is 45% and the classification probability of a highway is 40% less different. The classification obtained by the photograph provided by the user by using the second remote sensing data processing model is also farmland, and the region is indicated to be farmland with high probability. If the classification result obtained by the photo provided by the user is not farmland, the two results are different, and the result is doubtful. If the classification probability of the result obtained by the second remote sensing data processing model is relatively large, the result of the model can be acquired, and at the moment, the task is also judged to be completed. It can be appreciated that by the scheme, the task pushing can be terminated only when a sufficiently reliable result is obtained, and the balance between obtaining the result and reducing the pushing to the user can be achieved.
Referring to FIG. 3, an embodiment of the present application provides an agricultural data processing system, comprising:
a memory for storing a program;
and the processor is used for loading the program to execute the agricultural data processing method.
An embodiment of the present application provides a computer-readable storage medium storing a program that, when executed, implements the agricultural data processing method as described.
Note that the above is only a preferred embodiment of the present application and the technical principle applied. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, while the application has been described in connection with the above embodiments, the application is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the application, which is set forth in the following claims.
Claims (7)
1. A method of agricultural data processing, comprising:
acquiring satellite remote sensing data, processing the satellite remote sensing data based on a first remote sensing data processing model, outputting a segmentation result based on a geographic position, and outputting a first classification result of each segmentation area;
determining suspicious classification results with classification probability smaller than a threshold value in the first classification results, and acquiring a region range corresponding to each suspicious classification result;
acquiring sensing data of each area range in a preset time interval before and after the acquisition time from a sensing system of an access platform according to the acquisition time of the remote sensing data and each area range;
determining the correctness of a first classification result by using a second remote sensing data classification model according to the perceived data, wherein the method comprises the steps of selecting a corresponding second remote sensing data processing model to classify according to the acquired existing aerial photographing data or the perceived data acquired through a release task, based on the acquired data type of the number and the classification type of the first remote sensing data processing model, acquiring a second classification result, and verifying the correctness of the first classification result according to the second classification result;
the first remote sensing data processing model is used for outputting a segmentation result based on the geographic position and outputting a planting type result of each segmentation area; the second remote sensing data processing model is used for outputting a classification result of whether the sensing data belongs to a certain planting type;
the perception data comprise image data of unmanned aerial vehicle aerial photography and unmanned aerial vehicle aerial photography, image data of field photography, image data of fixed-point camera photography and data acquired by a non-image sensor.
2. The agricultural data processing method according to claim 1, wherein, according to the acquisition time of the remote sensing data and each region range, acquiring the sensing data of each region range in a preset time interval before and after the acquisition time from the sensing system of the access platform, specifically comprising:
issuing a query request to a perception system of an access platform, requesting to acquire the existing aerial photographing data in a preset time interval before and after the acquisition time, wherein the aerial photographing data relate to each regional range;
when the related partial area range does not have corresponding aerial photographing data, a data acquisition task is issued to a terminal of the accessed platform so as to obtain sensing data of each area range.
3. The agricultural data processing method according to claim 1, wherein a data acquisition task is issued to a terminal of an accessed platform to obtain perceived data of each area range, specifically:
when a user terminal of an access platform invokes a platform interface to process user perceived data, acquiring user terminal positioning information or an acquisition position of the perceived data, and defining an active area of the user terminal according to the terminal positioning information or the acquisition position of the perceived data;
when the active area of the user terminal overlaps with the area range corresponding to the suspicious classification result or the distance is smaller than a threshold value, creating task information aiming at the area range corresponding to the classification result;
task information corresponding to the region range corresponding to each suspicious classification result is put into a task information set;
establishing an information collection table, wherein the information collection table records whether tasks are completed or not, and after at least one task corresponding to an area range is completed, all tasks associated with the area range are marked as completed;
creating a plurality of delay tasks with different delay time into a message queue;
and when the delay task is consumed, pulling incomplete task information in the task information set according to the completion condition of the task in the information collection table, and issuing task pushing to a user terminal.
4. The agricultural data processing method according to claim 3, wherein when pushing a task to the user terminal, a task information acquisition template is generated according to the optional second remote sensing data processing model, and the template comprises at least one data template unit required by the second remote sensing data processing model;
wherein the optional second remote sensing data processing model refers to: has been trained and resides in the server and can output a model for determining whether the first classification result is correct based on the input data type that is different from the first telemetry data processing model.
5. The agricultural data processing method of claim 4, wherein the completion of the task in the information collection table means that the second classification result obtained by the sensory data obtained by the task is identical to the first classification result, or the second classification result obtained by the sensory data obtained by the task is inconsistent with the first classification result and the classification probability is greater than a preset value.
6. An agricultural data processing system, comprising:
a memory for storing a program;
a processor for loading the program to perform the agricultural data processing method of any one of claims 1 to 5.
7. A computer-readable storage medium, characterized in that a program is stored, which when executed, implements the agricultural data processing method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310657390.9A CN116645618B (en) | 2023-06-05 | 2023-06-05 | Agricultural data processing method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310657390.9A CN116645618B (en) | 2023-06-05 | 2023-06-05 | Agricultural data processing method, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116645618A CN116645618A (en) | 2023-08-25 |
| CN116645618B true CN116645618B (en) | 2023-12-08 |
Family
ID=87639764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310657390.9A Active CN116645618B (en) | 2023-06-05 | 2023-06-05 | Agricultural data processing method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116645618B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108182454A (en) * | 2018-01-18 | 2018-06-19 | 苏州大学 | Safety check identifying system and its control method |
| WO2018214195A1 (en) * | 2017-05-25 | 2018-11-29 | 中国矿业大学 | Remote sensing imaging bridge detection method based on convolutional neural network |
| CN110479638A (en) * | 2019-08-08 | 2019-11-22 | 南京英锐祺科技有限公司 | A kind of method and system of garbage classification |
| CN111968078A (en) * | 2020-07-28 | 2020-11-20 | 北京恒通智控机器人科技有限公司 | Appearance detection method, device, equipment and storage medium for power transformation equipment |
| CN113167779A (en) * | 2018-12-10 | 2021-07-23 | 克莱米特公司 | Mapping field anomalies using digital images and machine learning models |
| WO2021169473A1 (en) * | 2020-02-28 | 2021-09-02 | 深圳前海微众银行股份有限公司 | Model performance optimization method, apparatus and device, and storage medium |
| CN113807219A (en) * | 2021-09-06 | 2021-12-17 | 苏州中科蓝迪软件技术有限公司 | Method for identifying types of grain and oil crops in planting land by steps |
| CN114782815A (en) * | 2022-04-26 | 2022-07-22 | 平安普惠企业管理有限公司 | Method, device and equipment for automatically supervising greenbelt and storage medium |
| CN114881127A (en) * | 2022-04-21 | 2022-08-09 | 厦门天卫科技有限公司 | Crop fine classification method based on high-resolution remote sensing satellite image |
| CN116071600A (en) * | 2023-02-17 | 2023-05-05 | 中国科学院地理科学与资源研究所 | Crop remote sensing identification method and device based on multi-classification probability |
-
2023
- 2023-06-05 CN CN202310657390.9A patent/CN116645618B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018214195A1 (en) * | 2017-05-25 | 2018-11-29 | 中国矿业大学 | Remote sensing imaging bridge detection method based on convolutional neural network |
| CN108182454A (en) * | 2018-01-18 | 2018-06-19 | 苏州大学 | Safety check identifying system and its control method |
| CN113167779A (en) * | 2018-12-10 | 2021-07-23 | 克莱米特公司 | Mapping field anomalies using digital images and machine learning models |
| CN110479638A (en) * | 2019-08-08 | 2019-11-22 | 南京英锐祺科技有限公司 | A kind of method and system of garbage classification |
| WO2021169473A1 (en) * | 2020-02-28 | 2021-09-02 | 深圳前海微众银行股份有限公司 | Model performance optimization method, apparatus and device, and storage medium |
| CN111968078A (en) * | 2020-07-28 | 2020-11-20 | 北京恒通智控机器人科技有限公司 | Appearance detection method, device, equipment and storage medium for power transformation equipment |
| CN113807219A (en) * | 2021-09-06 | 2021-12-17 | 苏州中科蓝迪软件技术有限公司 | Method for identifying types of grain and oil crops in planting land by steps |
| CN114881127A (en) * | 2022-04-21 | 2022-08-09 | 厦门天卫科技有限公司 | Crop fine classification method based on high-resolution remote sensing satellite image |
| CN114782815A (en) * | 2022-04-26 | 2022-07-22 | 平安普惠企业管理有限公司 | Method, device and equipment for automatically supervising greenbelt and storage medium |
| CN116071600A (en) * | 2023-02-17 | 2023-05-05 | 中国科学院地理科学与资源研究所 | Crop remote sensing identification method and device based on multi-classification probability |
Non-Patent Citations (1)
| Title |
|---|
| 标签聚类损失在遥感影像分类中的应用;苏 赋， 于海鹏， 朱威西;自 然 资 源 遥 感;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116645618A (en) | 2023-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2018225069A1 (en) | Digitizing and mapping the public space using collaborative networks of mobile agents and cloud nodes | |
| DE112015006255T5 (en) | Object recognition device, object recognition method, and program | |
| KR102333143B1 (en) | System for providing people counting service | |
| CN112770265B (en) | Pedestrian identity information acquisition method, system, server and storage medium | |
| WO2021114985A1 (en) | Companionship object identification method and apparatus, server and system | |
| CN112749589A (en) | Method and device for determining routing inspection path and storage medium | |
| Minakshi et al. | High-accuracy detection of malaria mosquito habitats using drone-based multispectral imagery and Artificial Intelligence (AI) algorithms in an agro-village peri-urban pastureland intervention site (Akonyibedo) in Unyama SubCounty, Gulu District, Northern Uganda | |
| WO2023023272A1 (en) | Vision-based system training with simulated content | |
| Xie et al. | Recognition of big mammal species in airborne thermal imaging based on YOLO V5 algorithm | |
| CN111369152A (en) | Agricultural land value evaluation optimization method, device and equipment and readable storage medium | |
| CN111125290A (en) | Intelligent river patrol method and device based on river growth system and storage medium | |
| CN107578003A (en) | A kind of remote sensing images transfer learning method based on GEOGRAPHICAL INDICATION image | |
| CN111753610A (en) | Weather identification method and device | |
| CN116645618B (en) | Agricultural data processing method, system and storage medium | |
| CN109960965A (en) | Methods, devices and systems based on unmanned plane identification animal behavior | |
| AU2020254128A1 (en) | System and method for determining location and orientation of an object in a space | |
| CN108551473B (en) | Agricultural product communication method and device based on visual agriculture | |
| CN113255531B (en) | Method and device for processing living body detection model, computer equipment and storage medium | |
| Ogura et al. | Ground object recognition and segmentation from aerial image‐based 3D point cloud | |
| CN108399413A (en) | A kind of picture shooting region recognition and geographic positioning and device | |
| CN111291597B (en) | Crowd situation analysis method, device, equipment and system based on image | |
| JP6922447B2 (en) | Information processing system, server and communication method | |
| Bashir et al. | Efficient CNN-based disaster events classification using UAV-aided images for emergency response application | |
| US20210081784A1 (en) | Device and computer implemented method for training an artificial neural network | |
| CN113642671B (en) | Semi-supervised meta learning method and device based on task distribution change |
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 |