CN115861816A - Three-dimensional low vortex identification method and device, storage medium and terminal - Google Patents
Three-dimensional low vortex identification method and device, storage medium and terminal Download PDFInfo
- Publication number
- CN115861816A CN115861816A CN202211578597.9A CN202211578597A CN115861816A CN 115861816 A CN115861816 A CN 115861816A CN 202211578597 A CN202211578597 A CN 202211578597A CN 115861816 A CN115861816 A CN 115861816A
- Authority
- CN
- China
- Prior art keywords
- low
- vortex
- low vortex
- target
- model
- 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
Images
Classifications
-
- 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
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention discloses a three-dimensional low vortex identification method, a device, a storage medium and a terminal, wherein the method comprises the following steps: processing a target standard weather map of an area to be identified in a time period to be predicted according to a pre-trained three-dimensional low vortex identification model to generate a central point of each low vortex in layers with different heights; constructing a target matrix according to the distance between the central points of any two low vortexes in different height layers on the adjacent height layers; determining a plurality of potential matching pairs according to the matrix, and calculating the maximum influence range of two low vortexes in each potential matching pair; when the maximum influence ranges of the two low vortexes have an overlapping part and the central point is located in the overlapping part, determining each potential match as a true match, and generating a stereo low vortex based on each potential match pair of the true match. Because carry out analysis quantification through to standard weather map to there is pseudo-matching's low vortex between the filtering adjacent high layer, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching was finally established higher, promoted recognition efficiency.
Description
Technical Field
The invention relates to the technical field of automatic identification of weather systems, in particular to a three-dimensional low-vortex identification method, a three-dimensional low-vortex identification device, a storage medium and a terminal.
Background
Low vortices, in meteorological terms, refer to cyclonic vortices on a weather map where the central air pressure tends to be lower than the surrounding, i.e. low pressure vortices that occur in the atmosphere in the lower layers of the troposphere in small horizontal and vertical extent. A weather system, mainly in relation to air pressure fields, is most prominent on different height levels, 500hPa,700hPa and 850 hPa.
The low vortex has stronger convergent ascending air current, and cloud and rain weather can be generated. After the formation of partial low vortex, the low vortex is weakened and disappeared in situ, and only the weather change of the source area and the nearby area is caused. And some low vortexes guide airflow to move along with the low grooves or the high altitude, and are continuously enhanced and developed, rain areas are enlarged, precipitation is enhanced, and rainstorm is often formed. Therefore, the low vortex is used as a main weather influence system of extreme weather such as rainstorm, and the weather condition can be accurately forecasted by identifying and tracking the low vortex.
The existing calculation result of the low vortex based on the weather theory and a simple mathematical formula is low in accuracy and cannot replace manual identification of abundant experience accumulated by weather workers according to the weather theory, so that the accuracy of the finally generated three-dimensional low vortex is low, and the identification efficiency of the three-dimensional low vortex is reduced.
Disclosure of Invention
The embodiment of the application provides a three-dimensional low vortex identification method and device, a storage medium and a terminal. The following presents a simplified summary in order to provide a basic understanding of some aspects of the disclosed embodiments. This summary is not an extensive overview and is intended to neither identify key/critical elements nor delineate the scope of such embodiments. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is presented later.
In a first aspect, an embodiment of the present application provides a method for identifying a three-dimensional low vortex, where the method includes:
processing a target standard weather map of an area to be identified in a time period to be predicted according to a pre-trained three-dimensional low vortex identification model, and generating a central point of each low vortex in layers with different heights in the target standard weather map;
constructing a target matrix according to the target distance between the central points of any two low vortexes on the adjacent height layers in different height layers;
determining a plurality of potential matching pairs according to the target matrix, and calculating the maximum influence range of two low vortexes in each potential matching pair;
when the maximum influence ranges of the two low vortexes have an overlapping portion and the central points of the two low vortexes are located in the overlapping portion, determining each potential match as a true match, and generating a stereo low vortex based on each potential matching pair of the true matches.
Optionally, determining a plurality of potential matching pairs according to the target matrix includes:
traversing to obtain a row minimum value and a column minimum value in a target matrix;
when the positions of the row minimum value and the column minimum value are the same, determining two low vortex numbers of different height layers corresponding to the positions as a potential matching pair;
and after the target matrix traversal is finished, generating a plurality of potential matching pairs.
Optionally, the pre-trained stereo low vortex recognition model includes a pre-trained positioning model, a pre-trained classification model and a pre-trained low vortex central point determination model;
processing a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, and generating a central point of each low vortex in different height layers in the target standard weather map, wherein the steps comprise:
acquiring a target standard weather map of an area to be identified in a time period to be predicted;
inputting the target standard weather map into a pre-trained positioning model, and outputting target low vortex or target high pressure corresponding to the target standard weather map;
determining a target influence range diagram of target low vortex or target high pressure in a target standard weather diagram;
inputting the target influence range diagram into a pre-trained classification model, and outputting a classification result corresponding to a target standard weather diagram;
removing all high pressure according to the classification result, and generating a low vortex list corresponding to the target standard weather map;
and performing regression processing on each low vortex in the low vortex list according to a pre-trained low vortex central point determination model to obtain the central point of each low vortex in different height layers in the target standard weather map.
Optionally, before obtaining the target standard weather map of the area to be identified in the time period to be predicted, the method further includes:
generating a sample standard weather map according to historical lattice point data of the area to be identified;
receiving a marking instruction aiming at the sample standard weather map, marking low vortex and high pressure positions in a preset time period on the sample standard weather map based on the marking instruction, and generating a model training sample;
constructing a positioning model, inputting a model training sample into the positioning model for model training, and generating a pre-trained positioning model after training is finished;
generating a pre-trained classification model according to the labeled sample standard weather map;
and generating a pre-trained low vortex central point determination model according to the model training sample.
Optionally, generating a sample standard weather map according to historical grid point data of the area to be identified, including:
constructing an electronic map according to a preset longitude range and a preset latitude range;
drawing a plurality of isolines with different heights on the electronic map by adopting an isoline drawing method and combining preset height parameters and spacing parameters to obtain an isoline map;
acquiring historical grid point data of an area to be identified, and indicating the direction of a wind field in the historical grid point data by adopting the direction of an arrow of a preset pixel to obtain a wind field arrow mark;
and projecting the arrow mark of the wind field onto the contour map to generate a sample standard weather map.
Optionally, the generating a pre-trained classification model according to the labeled sample standard weather map includes:
determining and screening an influence range diagram of each low vortex and high pressure position marked in a sample standard weather diagram;
classifying data according to the screened influence range diagram of each low vortex and high pressure position to obtain low vortex data and high pressure data;
constructing a classification model;
and respectively inputting the low vortex data and the high pressure data into the classification model for model training, and generating a pre-trained classification model after the training is finished.
Optionally, generating a pre-trained low vortex central point determination model according to the model training sample, including:
determining a plurality of single-center low vortexes present in the model training sample;
labeling the center of each single-center low vortex to generate a labeled training sample;
and inputting the marked training samples into a preset neural network for network training, and generating a pre-trained low-vortex central point determination model after the training is finished.
In a second aspect, an embodiment of the present application provides a stereoscopic low vortex identification device, where the device includes:
the central point generation module is used for processing a target standard weather map of the area to be identified in the time period to be predicted according to a pre-trained three-dimensional low vortex identification model to generate the central point of each low vortex in different height layers in the target standard weather map;
the target matrix construction module is used for constructing a target matrix according to the target distance between the central points of any two low vortexes in different height layers on the adjacent height layers;
the maximum influence range determining module is used for determining a plurality of potential matching pairs according to the target matrix and calculating the maximum influence ranges of two low vortexes in each potential matching pair;
and the stereo low vortex generating module is used for determining each potential match as a true match when the maximum influence ranges of the two low vortices have an overlapping part and the central points of the two low vortices are positioned in the overlapping part, and generating the stereo low vortex based on each potential match pair of the true match.
In a third aspect, embodiments of the present application provide a computer storage medium having stored thereon a plurality of instructions adapted to be loaded by a processor and to perform the above-mentioned method steps.
In a fourth aspect, an embodiment of the present application provides a terminal, which may include: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to perform the above-mentioned method steps.
The technical scheme provided by the embodiment of the application can have the following beneficial effects:
in the embodiment of the application, the three-dimensional low vortex recognition device firstly obtains a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, generates a central point of each low vortex in different height layers, then constructs a target matrix according to the distance between the central points of any two low vortices in different height layers on adjacent height layers, secondly determines a plurality of potential matching pairs according to the matrix, calculates the maximum influence range of the two low vortices in each potential matching pair, and finally determines that each potential matching is true matching when the maximum influence range of the two low vortices has an overlapping part and the central point is located in the overlapping part, and generates the three-dimensional low vortex based on each potential matching pair of true matching. Because this application is through carrying out analysis quantization to standard weather map to there is the low vortex of pseudo-matching between the adjacent high layer of filtering, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching was finally established higher, and promoted the recognition efficiency of three-dimensional low vortex.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 is a schematic flow chart of a method for identifying a three-dimensional low vortex provided in an embodiment of the present application;
FIG. 2 is a schematic diagram of a standard weather map provided by an embodiment of the present application;
FIG. 3 is a schematic diagram of another standard weather map provided by an embodiment of the present application;
FIG. 4 is a schematic flow chart illustrating the generation of a stereo low vortex identification model according to an embodiment of the present disclosure;
FIG. 5 is a schematic diagram of another standard weather map provided by an embodiment of the present application;
FIG. 6 is a schematic view of a low vortex weather map provided by an embodiment of the present application;
FIG. 7 is a schematic diagram of a high pressure weather map provided by an embodiment of the present application;
FIG. 8 is a schematic illustration of a low vortex center point marking provided by an embodiment of the present application;
fig. 9 is a schematic structural diagram of a three-dimensional low vortex identification device according to an embodiment of the present application;
fig. 10 is a schematic structural diagram of a terminal according to an embodiment of the present application.
Detailed Description
The following description and the drawings sufficiently illustrate specific embodiments of the invention to enable those skilled in the art to practice them.
It should be understood that the described embodiments are only some embodiments of the invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
In the description of the present invention, it is to be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art. In addition, in the description of the present invention, "a plurality" means two or more unless otherwise specified. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
The application provides a three-dimensional low vortex identification method, a three-dimensional low vortex identification device, a storage medium and a terminal, which are used for solving the problems in the related art. Among the technical scheme that this application provided, because this application is through carrying out analysis quantification to standard weather map to there is pseudo-matching's low vortex between the adjacent high layer of filtering, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching finally established higher, and promoted the recognition efficiency of three-dimensional low vortex, adopt the exemplary embodiment to carry out the detailed description below.
The following describes in detail a method for identifying a three-dimensional low vortex provided by an embodiment of the present application with reference to fig. 1 to 8. The method may be implemented in dependence on a computer program, operable on a stereo low vortex identification device based on the von neumann architecture. The computer program may be integrated into the application or may run as a separate tool-like application.
Referring to fig. 1, a schematic flow chart of a stereo low vortex identification method is provided in an embodiment of the present application.
As shown in fig. 1, the method of the embodiment of the present application may include the following steps:
s101, processing a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, and generating a central point of each low vortex in layers with different heights in the target standard weather map;
the pre-trained three-dimensional low vortex recognition model is a mathematical model capable of determining the central point of each low vortex in the target standard weather map, and comprises a pre-trained positioning model, a pre-trained classification model and a pre-trained low vortex central point determination model.
In the embodiment of the application, when a pre-trained stereo low vortex recognition model is generated, firstly, a sample standard weather map is generated according to historical lattice point data of an area to be recognized, then, a marking instruction for the sample standard weather map is received, low vortex and high pressure positions in a preset time period are marked on the sample standard weather map based on the marking instruction, a model training sample is generated, secondly, a positioning model is built, the model training sample is input into the positioning model for model training, a pre-trained positioning model is generated after training is finished, a pre-trained classification model is generated according to the marked sample standard weather map, finally, a pre-trained low vortex center point determination model is generated according to the model training sample, and the finally generated pre-trained positioning model, the pre-trained classification model and the pre-trained low vortex center point determination model are integrated to generate the pre-trained stereo low vortex recognition model.
In a possible implementation manner, when the central point of each low vortex in different height layers in a target standard weather map is generated, a target standard weather map of an area to be identified in a time period to be predicted is obtained, the target standard weather map is input into a pre-trained positioning model, target low vortices or target high pressure corresponding to the target standard weather map are output, a target influence range map of the target low vortices or the target high pressure is determined in the target standard weather map, the target influence range map is input into a pre-trained classification model, a classification result corresponding to the target standard weather map is output, all high pressures are removed according to the classification result, a low vortex list corresponding to the target standard weather map is generated, and finally, regression processing is performed on each low vortex in the low vortex list according to the pre-trained low vortex central point determination model, so that the central point of each low vortex in different height layers in the target standard weather map is obtained.
Specifically, the area to be identified is a certain area in which the three-dimensional low vortex identification needs to be performed, and the time period to be predicted is a set period of time to be identified, for example, one day.
Further, the target standard weather map is input into the pre-trained positioning model, and a target low vortex or a target high pressure corresponding to the target standard weather map is output, for example, as shown in fig. 2, a range covered by each positioning black frame of the low vortex or the high pressure is an influence range of the low vortex or the high pressure. And judging whether one low vortex or high pressure interior contains at least two low vortex or high pressure centers according to the positioning result, if so, determining that the low vortex or high pressure is a composite low vortex or high pressure, otherwise, determining that the low vortex or high pressure is a single-center low vortex or high pressure, and if the low vortex 1 in the figure 2 contains two low vortices, determining that the low vortex 1 is a composite low vortex. The low vortex and the high pressure are only distinguished by different wind field directions, namely a counterclockwise wind field and a clockwise wind field.
After obtaining the low vortices and high pressure, the characteristics of the low vortices or high pressure for each layer may be stored as a list, each of the low vortices or high pressure characteristics including: extent of influence, whether it is compound low eddy, if it is compound low eddy, it contains the extent of influence of each single central low eddy.
Further, specifically, when the low vortex list is generated, the influence range of the target low vortex or the target high pressure corresponding to the target standard weather map can be extracted from the standard weather map, the influence range is input into a classification model trained in advance, if the classification result is high pressure, the target low vortex or the target high pressure is removed, if the target low vortex or the target high pressure is low vortex, the target high vortex or the target high pressure is retained, and finally the low vortex list corresponding to the target standard weather map can be obtained. For example, taking the low vortex and the high pressure in fig. 2 as an example, the list of low vortex after high pressure rejection is shown in fig. 3.
Further, after the low vortex list is obtained, traversing each low vortex in the low vortex list, and judging whether the low vortex is a composite low vortex; if the compound low vortex exists, the influence range of each single-center low vortex is extracted from the marked weather picture, the extracted pictures are regressed to obtain the central point of each single-center low vortex, and then the geometric center of the central point of the single-center low vortex contained in the compound low vortex is calculated and used as the center of the compound low vortex; and if the single-center low vortex is present, the influence range of the single-center low vortex is extracted from the marked weather picture, and regression is carried out to obtain the center of the single-center low vortex. Each low vortex in different height layers in the final target standard weather map adds a center point attribute.
S102, constructing a target matrix according to the target distance between the central points of any two low vortexes in different height layers on adjacent height layers;
in one possible implementation manner, after the central point of each low vortex in different height layers in the target standard weather map is obtained, the target matrix can be constructed according to the target distance between the central points of any two low vortices in different height layers on the adjacent height layers. For example, in the case of low vortices at three height levels of 500hpa,700hpa and 850hpa, the horizontal distance between the center of each low vortex and all the low vortices at 700hpa is calculated by traversing all the low vortices at 500hpa height levels, and assuming that the number of low vortices at 500hpa height is m and the number of low vortices at 700hpa height is n, a matrix of m rows and n columns can be generated.
S103, determining a plurality of potential matching pairs according to the target matrix, and calculating the maximum influence range of two low vortexes in each potential matching pair;
in the embodiment of the application, when a plurality of potential matching pairs are determined according to a target matrix, firstly, a row minimum value and a column minimum value in the target matrix are obtained in a traversing mode, then when the positions of the row minimum value and the column minimum value are the same, two low vortex numbers of layers with different heights corresponding to the positions are determined to be one potential matching pair, and finally, after the target matrix is traversed, the plurality of potential matching pairs are generated.
In a possible implementation manner, for example, when a matrix of m rows and n columns between any adjacent layers is analyzed, the position of the row minimum value and the position of the column minimum value of the m row and n column matrix are searched first, whether the two are overlapped is judged, that is, a certain number is the row minimum value and the column minimum value, if the two are overlapped, the low vortex of the 500hpa height layer represented by the row number where the value is located is matched with the low vortex of the 700hpa height layer represented by the column number where the value is located, and a potential matching pair is obtained. If the remaining low vortexes with the height of 500hpa can not find the matching low vortexes with the height of 700hpa, the single-layer isolated low vortexes are considered.
Further, after obtaining a plurality of potential matching pairs of a matrix of m rows and n columns, the maximum influence range of the two low vortexes in each potential matching pair can be determined.
And S104, when the maximum influence ranges of the two low vortexes have an overlapping part and the central points of the two low vortexes are positioned in the overlapping part, determining each potential match as a true match, and generating the three-dimensional low vortexes based on each potential match pair of the true match.
In one possible implementation, after obtaining the maximum influence ranges of the two low vortexes in each potential matching pair, when there is an overlapping portion of the maximum influence ranges of the two low vortexes and the central points of the two low vortexes are located in the overlapping portion, determining each potential matching as a true matching, and generating a stereo low vortex based on each potential matching pair of the true matching. For example, when whether the maximum influence ranges of the two low vortexes overlap is judged, if the maximum influence ranges of the two low vortexes do not overlap, the potential matching pair is a false match, if the maximum influence ranges of the two low vortexes overlap, whether the central points of the two low vortexes are within the overlap range is judged, if the central points of the two low vortexes are both within the overlap range, the potential matching pair is a true match, otherwise, the potential matching pair is a false match, and the low vortexes with the height of 500hpa in the false match are considered as single-layer isolated low vortexes. The method can filter out potential false matches, so that the identified three-dimensional low vortex accuracy is higher.
Furthermore, true matching between adjacent layers among three height layers of 500hpa,700hpa and 850hpa can be obtained according to the above mode, and finally a three-dimensional low vortex can be constructed according to the low vortex identification of true matching.
In the embodiment of the application, the three-dimensional low vortex recognition device firstly obtains a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, generates a central point of each low vortex in different height layers, then constructs a target matrix according to the distance between the central points of any two low vortices in different height layers on adjacent height layers, secondly determines a plurality of potential matching pairs according to the matrix, calculates the maximum influence range of the two low vortices in each potential matching pair, and finally determines that each potential matching is true matching when the maximum influence range of the two low vortices has an overlapping part and the central point is located in the overlapping part, and generates the three-dimensional low vortex based on each potential matching pair of true matching. Because this application is through carrying out analysis quantification to standard weather map to there is the low vortex of pseudo-matching between the adjacent high layer of filtering, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching finally established higher, and promoted the recognition efficiency of three-dimensional low vortex.
Referring to fig. 4, a schematic flow chart of generating a three-dimensional low vortex identification model is provided for an embodiment of the present application. As shown in fig. 4, the method of the embodiment of the present application may include the following steps:
s201, generating a sample standard weather map according to historical lattice point data of an area to be identified;
in the embodiment of the application, when a sample standard weather map is generated according to historical grid point data of an area to be identified, an electronic map is firstly constructed according to a preset longitude range and a preset latitude range, then a contour line drawing method is adopted, a plurality of contour lines with different heights are drawn on the electronic map by combining preset height parameters and interval parameters, a contour line map is obtained, historical grid point data of the area to be identified are obtained, an arrow point of a preset pixel is adopted to indicate a wind field direction in the historical grid point data, a wind field arrow mark is obtained, and finally the wind field arrow mark is projected onto the contour line map to generate the sample standard weather map.
In particular, the contour is a uniform shape and number, and the process of drawing the contour is a process of mathematically interpolating a large number of discrete geometric or physical quantities with a certain rule and transforming points with the same quantity into a graph. The contour line is widely applied in engineering and technical fields such as geology, water conservancy, civil engineering and the like, and in the geological field, the thickness distribution of geological blocks, the data distribution of seismic sections and the like can be visually displayed by using the contour line technology. The conventional contour drawing usually adopts a grid method, and the drawing steps are generally as follows: gridding discrete data; digitizing the grid points; calculating an equivalent point; tracking the contour line; smooth and mark contours.
In one possible implementation, an electronic map with a longitude range of 40-170 and a latitude range of 5-80 is drawn first, contour lines with a height of 500hpa,700hpa and 850hpa and a distance of 1dagpm are obtained by adopting a contour line drawing method, the directions of wind fields on grid points are projected onto the electronic map by using arrows with a length of 5 pixels to form a composite set, and the composite set is determined to be a sample standard weather map, as shown in fig. 5.
S202, receiving a marking instruction aiming at the sample standard weather map, marking low vortex and high pressure positions in a preset time period on the sample standard weather map based on the marking instruction, and generating a model training sample;
in a possible implementation manner, after a sample labeled weather map is obtained, a training sample can be manually labeled, and low vortex and high pressure positions are labeled on a standard weather map, as shown in fig. 2, wherein black frames are low vortex or high pressure, and finally a model training sample is obtained after the labeling is finished.
It should be noted that the compound low vortex means that at least two low vortex centers are located inside one large low vortex, and if two low vortex centers are located inside one large low vortex, the large low vortex is one low vortex, and the two low vortex centers inside are counted as two low vortices, so that a total of three low vortices are formed.
S203, constructing a positioning model, inputting a model training sample into the positioning model for model training, and generating a pre-trained positioning model after training is finished;
in the embodiment of the application, after the model training sample is obtained, the yolo-v5 algorithm can be used as the positioning model, the model training sample is used for training the positioning model, and the positioning model which is trained in advance can be obtained after the training is finished. The sample size may be sample data from 2017 to 2021 year 5.
S204, generating a pre-trained classification model according to the labeled sample standard weather map;
in the embodiment of the application, when a pre-trained classification model is generated according to a labeled sample standard weather map, firstly, an influence range map of each low vortex and high pressure position labeled in the sample standard weather map is determined and screened, then, data classification is performed according to the influence range map of each low vortex and high pressure position screened, low vortex data and high pressure data are obtained, secondly, the classification model is constructed, finally, the low vortex data and the high pressure data are respectively input into the classification model for model training, and the pre-trained classification model is generated after the training is finished.
In one possible implementation, when classification is completed after determining and screening the influence range map of each low vortex and high pressure position marked in the sample standard weather map, the obtained low vortex is one type, and the obtained high pressure is one type. Taking the low vortex and high pressure of the northern hemisphere as examples (the low vortex of the northern hemisphere is oriented counterclockwise and the high pressure is clockwise under the influence of the yaw bias), the low vortex is shown in fig. 6 and the high pressure is shown in fig. 7. The classification model may be constructed using resnet 18.
S205, generating a pre-trained low vortex central point determination model according to the model training sample;
in the embodiment of the application, when the pre-trained low vortex central point determination model is generated according to the model training sample, a plurality of single-center low vortices existing in the model training sample are determined firstly, then the center of each single-center low vortex is labeled to generate a labeled training sample, finally the labeled training sample is input into a preset neural network for network training, and the pre-trained low vortex central point determination model is generated after the training is finished.
Specifically, the center of each single-center low vortex is labeled, one low vortex has two labels, x and y respectively, as shown in fig. 8, the position of the gray point is the center of the low vortex, the coordinate of the gray point is x and y, and finally the gray point can be input into a traditional neural network for network training based on the labeled sample of the coordinate point. Practice shows that the non-traditional neural network does not perform well on the task of returning to the low vortex center.
And S206, integrating the pre-trained positioning model, the pre-trained classification model and the pre-trained low vortex central point determination model to obtain the three-dimensional low vortex identification model.
In the embodiment of the application, the three-dimensional low vortex recognition device firstly obtains a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, generates a central point of each low vortex in different height layers, then constructs a target matrix according to the distance between the central points of any two low vortices in different height layers on adjacent height layers, secondly determines a plurality of potential matching pairs according to the matrix, calculates the maximum influence range of the two low vortices in each potential matching pair, and finally determines that each potential matching is true matching when the maximum influence range of the two low vortices has an overlapping part and the central point is located in the overlapping part, and generates the three-dimensional low vortex based on each potential matching pair of true matching. Because this application is through carrying out analysis quantization to standard weather map to there is the low vortex of pseudo-matching between the adjacent high layer of filtering, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching was finally established higher, and promoted the recognition efficiency of three-dimensional low vortex.
The following are embodiments of the apparatus of the present invention that may be used to perform embodiments of the method of the present invention. For details which are not disclosed in the embodiments of the apparatus of the present invention, reference is made to the embodiments of the method of the present invention.
Referring to fig. 9, a schematic structural diagram of a stereoscopic low vortex identification device according to an exemplary embodiment of the present invention is shown. The stereoscopic low vortex identification device can be realized by software, hardware or a combination of the two to form all or part of the terminal. The device 1 comprises a central point generation module 10, a target matrix construction module 20, a maximum influence range determination module 30 and a three-dimensional low vortex generation module 40.
The central point generation module 10 is configured to process a target standard weather map of the area to be identified in the time period to be predicted according to a pre-trained three-dimensional low vortex identification model, and generate a central point of each low vortex in different height layers in the target standard weather map;
the target matrix building module 20 is configured to build a target matrix according to a target distance between center points of any two low vortexes in different height layers on adjacent height layers;
a maximum influence range determining module 30, configured to determine multiple potential matching pairs according to the target matrix, and calculate maximum influence ranges of two low vortexes in each potential matching pair;
and the stereoscopic low vortex generating module 40 is used for determining each potential match as a true match when the maximum influence ranges of the two low vortices have an overlapping part and the central points of the two low vortices are positioned in the overlapping part, and generating the stereoscopic low vortex based on each potential match pair of the true match.
It should be noted that, when the stereo low vortex identification device provided in the foregoing embodiment executes the stereo low vortex identification method, only the division of the above functional modules is taken as an example, and in practical applications, the above functions may be distributed by different functional modules according to needs, that is, the internal structure of the device may be divided into different functional modules to complete all or part of the functions described above. In addition, the three-dimensional low vortex identification device and the three-dimensional low vortex identification method provided by the above embodiments belong to the same concept, and details of implementation processes are shown in the method embodiments, which are not described herein again.
The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.
In the embodiment of the application, the three-dimensional low vortex recognition device firstly obtains a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, generates a central point of each low vortex in different height layers, then constructs a target matrix according to the distance between the central points of any two low vortices in different height layers on adjacent height layers, secondly determines a plurality of potential matching pairs according to the matrix, calculates the maximum influence range of the two low vortices in each potential matching pair, and finally determines that each potential matching is true matching when the maximum influence range of the two low vortices has an overlapping part and the central point is located in the overlapping part, and generates the three-dimensional low vortex based on each potential matching pair of true matching. Because this application is through carrying out analysis quantization to standard weather map to there is the low vortex of pseudo-matching between the adjacent high layer of filtering, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching was finally established higher, and promoted the recognition efficiency of three-dimensional low vortex.
The present invention also provides a computer readable medium having stored thereon program instructions which, when executed by a processor, implement the method of stereo low vortex identification provided by the various method embodiments described above.
The present invention also provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of stereo low vortex identification of the various method embodiments described above.
Please refer to fig. 10, which provides a schematic structural diagram of a terminal according to an embodiment of the present application. As shown in fig. 10, terminal 1000 can include: at least one processor 1001, at least one network interface 1004, a user interface 1003, memory 1005, at least one communication bus 1002.
Wherein a communication bus 1002 is used to enable connective communication between these components.
The user interface 1003 may include a Display screen (Display) and a Camera (Camera), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.
The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface), among others.
The Memory 1005 may include a Random Access Memory (RAM) or a Read-Only Memory (Read-Only Memory). Optionally, the memory 1005 includes a non-transitory computer-readable medium. The memory 1005 may be used to store an instruction, a program, code, a set of codes, or a set of instructions. The memory 1005 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the various method embodiments described above, and the like; the storage data area may store data and the like referred to in the above respective method embodiments. The memory 1005 may optionally be at least one memory device located remotely from the processor 1001. As shown in fig. 10, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a stereoscopic low vortex recognition application.
In the terminal 1000 shown in fig. 10, the user interface 1003 is mainly used as an interface for providing input for a user, and acquiring data input by the user; and the processor 1001 may be configured to invoke the stereoscopic low vortex identification application stored in the memory 1005, and specifically perform the following operations:
processing a target standard weather map of an area to be identified in a time period to be predicted according to a pre-trained three-dimensional low vortex identification model, and generating a central point of each low vortex in layers with different heights in the target standard weather map;
constructing a target matrix according to the target distance between the central points of any two low vortexes in different height layers on the adjacent height layers;
determining a plurality of potential matching pairs according to the target matrix, and calculating the maximum influence range of two low vortexes in each potential matching pair;
when the maximum influence ranges of the two low vortexes have an overlapping portion and the central points of the two low vortexes are located in the overlapping portion, determining each potential match as a true match, and generating a stereo low vortex based on each potential matching pair of the true matches.
In one embodiment, when determining a plurality of potential matching pairs according to the target matrix, the processor 1001 specifically performs the following operations:
traversing to obtain a row minimum value and a column minimum value in a target matrix;
when the positions of the row minimum value and the column minimum value are the same, determining two low vortex numbers of different height layers corresponding to the positions as a potential matching pair;
and after the target matrix traversal is finished, generating a plurality of potential matching pairs.
In one embodiment, when the processor 1001 executes processing on a target standard weather map of a to-be-identified area in a to-be-predicted time period according to a pre-trained stereo low vortex identification model to generate a central point of each low vortex in different height layers in the target standard weather map, the following operations are specifically executed:
acquiring a target standard weather map of an area to be identified in a time period to be predicted;
inputting the target standard weather map into a pre-trained positioning model, and outputting target low vortex or target high pressure corresponding to the target standard weather map;
determining a target influence range diagram of target low vortex or target high pressure in a target standard weather diagram;
inputting the target influence range diagram into a pre-trained classification model, and outputting a classification result corresponding to a target standard weather diagram;
removing all high pressure according to the classification result, and generating a low vortex list corresponding to the target standard weather map;
and performing regression processing on each low vortex in the low vortex list according to a pre-trained low vortex central point determination model to obtain the central point of each low vortex in different height layers in the target standard weather map.
In one embodiment, the processor 1001, when performing the operation of obtaining the target standard weather map of the to-be-identified region in the to-be-predicted time period, further performs the following operation:
generating a sample standard weather map according to historical grid point data of the area to be identified;
receiving a marking instruction aiming at the sample standard weather map, marking low vortex and high pressure positions in a preset time period on the sample standard weather map based on the marking instruction, and generating a model training sample;
constructing a positioning model, inputting a model training sample into the positioning model for model training, and generating a pre-trained positioning model after training is finished;
generating a pre-trained classification model according to the labeled sample standard weather map;
and generating a pre-trained low vortex central point determination model according to the model training sample.
In one embodiment, when the processor 1001 generates the sample standard weather map according to the historical grid point data of the area to be identified, the following operations are specifically performed:
constructing an electronic map according to a preset longitude range and a preset latitude range;
drawing a plurality of isolines with different heights on the electronic map by adopting an isoline drawing method and combining preset height parameters and spacing parameters to obtain an isoline map;
acquiring historical grid point data of an area to be identified, and adopting an arrow direction of a preset pixel to represent a wind field direction in the historical grid point data to obtain a wind field arrow mark;
and projecting the arrow mark of the wind field onto the contour map to generate a sample standard weather map.
In one embodiment, the processor 1001, when executing the generation of the pre-trained classification model according to the labeled sample standard weather map, specifically performs the following operations:
determining and screening an influence range diagram of each low vortex and high pressure position marked in a sample standard weather diagram;
classifying data according to the screened influence range diagram of each low vortex and high pressure position to obtain low vortex data and high pressure data;
constructing a classification model;
and respectively inputting the low vortex data and the high pressure data into the classification model for model training, and generating a pre-trained classification model after the training is finished.
In one embodiment, the processor 1001, when executing the generation of the pre-trained low vortex center point determination model from the model training samples, specifically performs the following operations:
determining a plurality of single-center low vortexes present in the model training sample;
labeling the center of each single-center low vortex to generate a labeled training sample;
and inputting the marked training samples into a preset neural network for network training, and generating a pre-trained low vortex central point determination model after the training is finished.
In the embodiment of the application, the three-dimensional low vortex recognition device firstly obtains a target standard weather map of an area to be recognized in a time period to be predicted according to a pre-trained three-dimensional low vortex recognition model, generates a central point of each low vortex in different height layers, then constructs a target matrix according to the distance between the central points of any two low vortices in different height layers on adjacent height layers, secondly determines a plurality of potential matching pairs according to the matrix, calculates the maximum influence range of the two low vortices in each potential matching pair, and finally determines that each potential matching is true matching when the maximum influence range of the two low vortices has an overlapping part and the central point is located in the overlapping part, and generates the three-dimensional low vortex based on each potential matching pair of true matching. Because this application is through carrying out analysis quantization to standard weather map to there is the low vortex of pseudo-matching between the adjacent high layer of filtering, make the three-dimensional low vortex degree of accuracy that the low vortex based on really matching was finally established higher, and promoted the recognition efficiency of three-dimensional low vortex.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program to instruct associated hardware, and the program for stereoscopic low vortex recognition can be stored in a computer readable storage medium, and when executed, the program can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a read-only memory or a random access memory.
The above disclosure is only for the purpose of illustrating the preferred embodiments of the present application and is not to be construed as limiting the scope of the present application, so that the present application is not limited thereto, and all equivalent variations and modifications can be made to the present application.
Claims (10)
1. A method for identifying three-dimensional low vortices, the method comprising:
processing a target standard weather map of an area to be identified in a time period to be predicted according to a pre-trained three-dimensional low vortex identification model, and generating a central point of each low vortex in layers with different heights in the target standard weather map;
constructing a target matrix according to the target distance between the central points of any two low vortexes in different height layers on the adjacent height layers;
determining a plurality of potential matching pairs according to the target matrix, and calculating the maximum influence range of two low vortexes in each potential matching pair;
when the maximum influence ranges of the two low vortexes have an overlapping part and the central point of the two low vortexes is located in the overlapping part, determining that each potential match is a true match, and generating a stereo low vortex based on each potential match pair of the true match.
2. The method of claim 1, wherein determining a plurality of potential matching pairs from the target matrix comprises:
traversing to obtain a row minimum value and a column minimum value in the target matrix;
when the positions of the row minimum value and the column minimum value are the same, determining two low vortex numbers of different height layers corresponding to the positions as a potential matching pair;
and after the target matrix traversal is finished, generating a plurality of potential matching pairs.
3. The method of claim 1, wherein the pre-trained stereo low vortex recognition model comprises a pre-trained localization model, a pre-trained classification model, and a pre-trained low vortex centroiding model;
the processing of the target standard weather map of the area to be recognized in the time period to be predicted according to the pre-trained three-dimensional low vortex recognition model to generate the central point of each low vortex in different height layers in the target standard weather map comprises the following steps:
acquiring a target standard weather map of an area to be identified in a time period to be predicted;
inputting the target standard weather map into a pre-trained positioning model, and outputting target low vortex or target high pressure corresponding to the target standard weather map;
determining a target influence range map of the target low vortex or the target high pressure in the target standard weather map;
inputting the target influence range diagram into a pre-trained classification model, and outputting a classification result corresponding to the target standard weather diagram;
removing all high pressure according to the classification result, and generating a low vortex list corresponding to the target standard weather map;
and performing regression processing on each low vortex in the low vortex list according to a pre-trained low vortex central point determination model to obtain the central point of each low vortex in different height layers in the target standard weather map.
4. The method of claim 1, wherein the obtaining the target standard weather map for the area to be identified in the time period to be predicted further comprises:
generating a sample standard weather map according to historical lattice point data of the area to be identified;
receiving a marking instruction aiming at the sample standard weather map, marking low vortex and high pressure positions in a preset time period on the sample standard weather map based on the marking instruction, and generating a model training sample;
constructing a positioning model, inputting the model training sample into the positioning model for model training, and generating a pre-trained positioning model after training is finished;
generating a pre-trained classification model according to the labeled sample standard weather map;
and generating a pre-trained low vortex central point determination model according to the model training sample.
5. The method of claim 4, wherein generating a sample standard weather map from historical grid point data for the area to be identified comprises:
constructing an electronic map according to a preset longitude range and a preset latitude range;
drawing a plurality of contour lines with different heights on the electronic map by adopting a contour line drawing method and combining preset height parameters and interval parameters to obtain a contour line map;
acquiring historical grid point data of an area to be identified, and adopting an arrow direction of a preset pixel to represent a wind field direction in the historical grid point data to obtain a wind field arrow mark;
and projecting the wind field arrow mark on the contour map to generate a sample standard weather map.
6. The method of claim 4, wherein generating a pre-trained classification model from the labeled sample standard weather map comprises:
determining and screening an influence range diagram of each low vortex and high pressure position marked in a sample standard weather diagram;
classifying data according to the screened influence range diagram of each low vortex and high pressure position to obtain low vortex data and high pressure data;
constructing a classification model;
and respectively inputting the low vortex data and the high pressure data into the classification model for model training, and generating a pre-trained classification model after the training is finished.
7. The method of claim 4, wherein generating a pre-trained low-vortex centroiding model from the model training samples comprises:
determining a plurality of single-center low vortices present in the model training sample;
labeling the center of each single-center low vortex to generate a labeled training sample;
and inputting the marked training samples into a preset neural network for network training, and generating a pre-trained low vortex central point determination model after the training is finished.
8. A stereoscopic low vortex identification device, the device comprising:
the central point generation module is used for processing a target standard weather map of the area to be identified in the time period to be predicted according to a pre-trained three-dimensional low vortex identification model to generate the central point of each low vortex in different height layers in the target standard weather map;
the target matrix construction module is used for constructing a target matrix according to the target distance between the central points of any two low vortexes in different height layers on the adjacent height layers;
the maximum influence range determining module is used for determining a plurality of potential matching pairs according to the target matrix and calculating the maximum influence ranges of two low vortexes in each potential matching pair;
and the stereo low vortex generating module is used for determining that each potential match is a true match when the maximum influence ranges of the two low vortices have an overlapping part and the central points of the two low vortices are positioned in the overlapping part, and generating the stereo low vortex based on each potential match pair of the true match.
9. A computer storage medium storing a plurality of instructions adapted to be loaded by a processor and to perform the method of any of claims 1-7.
10. A terminal, comprising: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to perform the method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211578597.9A CN115861816B (en) | 2022-12-06 | 2022-12-06 | Three-dimensional low-vortex recognition method and device, storage medium and terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211578597.9A CN115861816B (en) | 2022-12-06 | 2022-12-06 | Three-dimensional low-vortex recognition method and device, storage medium and terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115861816A true CN115861816A (en) | 2023-03-28 |
| CN115861816B CN115861816B (en) | 2023-07-18 |
Family
ID=85671454
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211578597.9A Active CN115861816B (en) | 2022-12-06 | 2022-12-06 | Three-dimensional low-vortex recognition method and device, storage medium and terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861816B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116759009A (en) * | 2023-06-09 | 2023-09-15 | 中科三清科技有限公司 | Ozone forecasting method, system and medium matched with weather through machine learning |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114626458B (en) * | 2022-03-15 | 2022-10-21 | 中科三清科技有限公司 | High-voltage rear part identification method and device, storage medium and terminal |
| CN114565057B (en) * | 2022-03-15 | 2022-10-21 | 中科三清科技有限公司 | Machine learning-based grading field identification method and device, storage medium and terminal |
| CN114565056B (en) * | 2022-03-15 | 2022-09-20 | 中科三清科技有限公司 | Machine learning-based cold-front identification method and device, storage medium and terminal |
| CN114863418B (en) * | 2022-04-29 | 2023-03-07 | 中科三清科技有限公司 | Voltage-sharing field identification method and device, storage medium and terminal |
-
2022
- 2022-12-06 CN CN202211578597.9A patent/CN115861816B/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116759009A (en) * | 2023-06-09 | 2023-09-15 | 中科三清科技有限公司 | Ozone forecasting method, system and medium matched with weather through machine learning |
| CN116759009B (en) * | 2023-06-09 | 2024-02-02 | 中科三清科技有限公司 | Ozone forecasting method, system and medium matched with weather through machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115861816B (en) | 2023-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110136170B (en) | Remote sensing image building change detection method based on convolutional neural network | |
| CN114565056B (en) | Machine learning-based cold-front identification method and device, storage medium and terminal | |
| US10115165B2 (en) | Management of tax information based on topographical information | |
| CN114863418B (en) | Voltage-sharing field identification method and device, storage medium and terminal | |
| US7778808B2 (en) | Geospatial modeling system providing data thinning of geospatial data points and related methods | |
| CN112990086A (en) | Remote sensing image building detection method and device and computer readable storage medium | |
| CN112818925A (en) | Urban building and crown identification method | |
| CN114565057B (en) | Machine learning-based grading field identification method and device, storage medium and terminal | |
| CN115861816B (en) | Three-dimensional low-vortex recognition method and device, storage medium and terminal | |
| Tarsha Kurdi et al. | Automatic evaluation and improvement of roof segments for modelling missing details using Lidar data | |
| CN116168246A (en) | Method, device, equipment and medium for identifying waste slag field for railway engineering | |
| CN108287845A (en) | A kind of Automatic extraction method for road information and device and hybrid navigation system | |
| CN115731560B (en) | Deep learning-based slot line identification method and device, storage medium and terminal | |
| CN116071651B (en) | Voltage equalizing field identification method and device, storage medium and terminal | |
| CN116051777B (en) | Super high-rise building extraction method, apparatus and readable storage medium | |
| CN115937690B (en) | Slotline generation method and device, storage medium and terminal | |
| CN114626458B (en) | High-voltage rear part identification method and device, storage medium and terminal | |
| CN115861811A (en) | Wind shear region identification method and device, storage medium and terminal | |
| CN113867410A (en) | Unmanned aerial vehicle aerial photography data acquisition mode identification method and system | |
| CN116681959B (en) | Machine learning-based frontal line identification method and device, storage medium and terminal | |
| CN116759009B (en) | Ozone forecasting method, system and medium matched with weather through machine learning | |
| CN115830380B (en) | Wind shear line identification method and device based on artificial intelligence, storage medium and terminal | |
| CN115859628B (en) | Bezier curve-based wind shear point identification method, device, storage medium and terminal | |
| Yu et al. | A cue line based method for building modeling from LiDAR and satellite imagery | |
| CN115861479A (en) | Slot line generation method and device, storage medium and terminal |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20230328 Assignee: Beijing Zhongke Sanqing Environmental Technology Co.,Ltd. Assignor: 3CLEAR TECHNOLOGY Co.,Ltd. Contract record no.: X2023980042173 Denomination of invention: A three-dimensional vortex recognition method, device, storage medium, and terminal Granted publication date: 20230718 License type: Common License Record date: 20230920 |
|
| EE01 | Entry into force of recordation of patent licensing contract |