CN113888670A - Method for generating high-precision two-dimensional terrain of one-way river channel based on deep learning - Google Patents
Method for generating high-precision two-dimensional terrain of one-way river channel based on deep learning Download PDFInfo
- Publication number
- CN113888670A CN113888670A CN202111212259.9A CN202111212259A CN113888670A CN 113888670 A CN113888670 A CN 113888670A CN 202111212259 A CN202111212259 A CN 202111212259A CN 113888670 A CN113888670 A CN 113888670A
- Authority
- CN
- China
- Prior art keywords
- river
- river channel
- terrain
- target
- deep learning
- 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
- 238000013135 deep learning Methods 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000011144 upstream manufacturing Methods 0.000 claims abstract description 48
- 238000012549 training Methods 0.000 claims abstract description 32
- 238000001514 detection method Methods 0.000 claims abstract description 24
- 238000013507 mapping Methods 0.000 claims abstract description 24
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 20
- 238000009499 grossing Methods 0.000 claims abstract description 13
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims description 3
- 238000010008 shearing Methods 0.000 claims description 3
- 230000008685 targeting Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000009418 renovation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/20—Drawing from basic elements, e.g. lines or circles
- G06T11/203—Drawing of straight lines or curves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
- Y02A90/30—Assessment of water resources
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Remote Sensing (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a method for generating a high-precision two-dimensional terrain of a one-way river channel based on deep learning, which comprises the steps of preprocessing the existing river channel measurement data according to the characteristics of a target river channel based on the deep learning according to the existing other river channel measurement data and target river channel images and upstream and downstream section data, establishing a one-to-one mapping relation of an image set, a river channel data set and the existing river channel data set of the existing river channel data, providing a method for classifying river channel quadrants, classifying the existing river channel image set by using a convolutional neural network under the deep learning, training a river channel shore line image by using key point detection CPM, identifying key points of the target river channel by using a trained model, matching a plurality of alternative river channel sets similar to the target river channel form, matching two most similar river channels in the alternative river channel sets in the section form, and performing scaling reconstruction on terrain scatter data corresponding to the two river channels by using a similar proportion, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the target river channel.
Description
Technical Field
The invention relates to the technical field of river dynamic simulation, in particular to a method for generating a one-way river high-precision two-dimensional terrain based on deep learning.
Background
Throughout the national economic development pattern, the areas around the river are often densely populated and highly economic, and the important construction for promoting economic development is the use of river resources. The method is a prerequisite for developing river channel dynamic related scientific research, planning and making, engineering demonstration and other works, and is an indispensable step for better utilizing and protecting river channel resources.
River course section data is the basic data that river course landform characteristic discerned, and general data source has two: designing section data aiming at river regulation; secondly, aiming at the measurement data of the natural river channel or the river channel which is longer in renovation, the measurement is a time-consuming and labor-consuming work, and a surveying and mapping department usually only measures the section which is more characteristic in landform and landform. In any way, under the condition of only river channel section data, river channel simulation calculation is carried out, terrain interpolation is carried out by directly utilizing a small number of measuring points of the river channel section, often obtained river channel terrain simulation files are distorted or a large number of distorted points exist, and for the terrain difference phenomenon, a common method is to artificially modify the terrain, so that the method has great subjectivity. Therefore, under the condition of mastering the river channel design section data, how to obtain the river channel terrain scatter data meeting the precision requirement and then use the river channel terrain scatter data for the interpolation simulation of the river channel terrain is a point which is often ignored in the basic work.
Disclosure of Invention
The invention provides a method for generating a high-precision two-dimensional terrain of a one-way river channel based on deep learning, aiming at overcoming the defects in the prior art.
In order to solve the technical problems, the technical scheme of the invention is as follows:
the invention provides a method for generating a high-precision two-dimensional terrain of a one-way river channel based on deep learning, which comprises the following steps:
s1, acquiring a target river R0Calculating the ratio of the average width W to the length L of the shoreline to obtain a ratio parameter a, wherein a is L/W;
s2, performing geometric enhancement on the existing other river channel data based on the average width W, the shoreline length L and the proportion parameter a characteristics to obtain a river channel terrain image set IMAGER and a river channel data set DR aiming at the parameter characteristics, and establishing a mapping relation between the river channel terrain image set IMAGER and the river channel data set DR;
s3, splitting the river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to the IMAGER range of the river channel terrain image set, and ensuring the existence of the mapping relation among DR, DL and IMAGER;
s4, carrying out image classification based on Convolutional Neural Network (CNN) under deep learning on the river channel terrain image set IMAGER, dividing four types of river channels according to the trend and the shoreline curvature of the river channels to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)；
S5 for IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i＝1、2、3、4)；
S6, identifying the target river channel by using the key point detection CPM model, and selecting a proper training set model result DTi(i is 1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a set threshold valuej(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
S7, extracting terrain scatter database DBj(j is 1, 2, …, n) and target river R0River R with highest upstream similarityg1And with the target river R0River R with highest downstream similarityg2And calculating parameter proportions of scale1, scale2, ratio1 and ratio 2;
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S02H02)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)
wherein: sg1To indicate river course Rg1Cross sectional area, Sg2To indicate river course Rg2Cross sectional area, S01Representing a target river R0Upstream cross-sectional area, S02Representing a target river R0Upstream cross-sectional area, Hg1To indicate river course Rg1Elevation of lowest point of section, Hg2To indicate river course Rg2Elevation of lowest point of section, H01Representing a target river R0Elevation of lowest point of upstream section, H02Representing a target river R0The elevation of the lowest point of the upstream section;
s8 river channel RgIs represented by Rg1、Rg2River course formed by river course RgIs used as a substrate and is prepared according to the terrain elevation ratios scale1 and scale2Step zoom Rg1、Rg2Landform scatter data to target river R0And splicing in the shoreline range according to ratio1 and ratio2, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the river channel.
Further, the specific process of step S2 is:
s2.1, collecting other river channel data, generating an initial high-definition image set and an initial river channel data set, establishing a mapping relation between the initial high-definition image set and the initial river channel data set, and ensuring that subsequent changes are reflected to the river channel data set;
s2.2, according to the parameter characteristics involved in the step S1, utilizing a deep learning preprocessing network to carry out image preprocessing, including image scaling, image rotation and image shearing, to obtain a classification river channel image set IMAGER and a river channel data set DR, wherein the river channel image passing through the preprocessing network has the characteristics that:
the river channel is a one-way river channel without branch;
secondly, the sizes of all river channel images are consistent;
thirdly, the image areas occupied by the river channels in the single river channel image are approximately similar;
the river channel data set DR comprises initial and terminal 2 plane straight lines of a one-way river channel, 2 plane curves of the left and right banks of the river channel and river channel terrain elevation scattered point data in the range of the river channel.
Further, the specific process of step S3 is:
for the river data set DR obtained by processing in step S2, the front line segment is extracted and split into the shoreline data set DL, the river terrain elevation scatter data is extracted as the terrain data set DB, and it is ensured that the mapping relationship between DR, DL and IMAGER still exists.
Further, the specific process of step S4 is:
s4.1, dividing four quadrants, limiting the flow direction of the river channel to be from west to east, enabling the river channels to cross over 3 quadrants in total, dividing the river channels into four categories according to the difference of the quadrants crossed by the river channels in sequence, and marking the four categories as IR1, IR2, IR3 and IR4 respectively, wherein the specific classification rule is as follows:
IR 1: the river channel sequentially passes through a second quadrant, a third quadrant and a fourth quadrant;
IR 2: the river channel sequentially passes through a second quadrant, a first quadrant and a fourth quadrant;
IR 3: the river channel sequentially passes through a third quadrant, a second quadrant and a first quadrant;
IR 4: the river channel sequentially passes through a third quadrant, a fourth quadrant and a first quadrant;
s4.2, carrying out image classification based on Convolutional Neural Network (CNN) under deep learning on the classified river channel terrain image set IMAGER by utilizing the provided river channel quadrant classification rule to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)。
Further, the specific process of step S5 is:
s5.1, definition of IMAGERiThe identifying key point features in (i ═ 1, 2, 3, and 4) include: starting point 1 of a left bank; a point 2 with the largest vertical distance from a connecting line from the starting point to the end point in the left bank line; left bank end point 3; starting point 4 of right bank; a point 5 with the largest vertical distance from the starting point to the end point in the right bank line; right bank end point 6;
s5.2, IMAGER with key point characteristicsi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using a key point detection CPM model under deep learning, training and testing, and establishing a training set model result DTi(i＝1、2、3、4)。
Further, the specific process of step S6 is:
s6.1, performing key point identification on the target river channel by using the established key point detection CPM model;
s6.2, selecting a proper training set model result DTi(i-1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a preset threshold (90%, 95%,. 99% and the like)j(j is 1, 2, …, n), extracting the existing corresponding terrain scatter database DBj(j＝1、2、…、n)。
Further, the specific process of step S7 is:
s7.1, extracting the terrain scatter database DBj(j is 1, 2, …, n) carrying out triangulation interpolation of the landforms of the starting point section and the ending point section;
s7.2, extracting a target river R0Cross-sectional area S of the cross-section at the upstream and downstream 201、S02Elevation of lowest point of cross section H01、H02；
S7.3, extracting and targeting river R0River R with highest similarity between upstream section area and lowest section point elevation parametersg1And river channel R with highest similarity of elevation parameters of downstream section area and lowest point of sectiong2Counting the corresponding parameter Sg1、Sg2Elevation of lowest point of cross section Hg1、Hg2And calculating the parameter proportion:
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S01H01)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)。
further, the specific process of step S8 is:
s8.1, river channel RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1And Rg2Topographic scatter elevation data;
s8.2, according to the target river R0Shore range scaling Rg1And Rg2A data range;
s8.3, taking Rg1Upstream to downstream ratio1 of proportional length of channel and Rg2Splicing river channels with proportional lengths from the downstream to the upstream ratio 2;
and S8.4, smoothing the spliced topographic scattered point elevation data to obtain the high-precision two-dimensional topographic scattered point data of the river.
The invention provides a system for generating a high-precision two-dimensional terrain of a one-way river based on deep learning, which comprises a memory and a processor, wherein the memory comprises a method program for generating the high-precision two-dimensional terrain of the one-way river based on deep learning, and the method program for generating the high-precision two-dimensional terrain of the one-way river based on deep learning realizes the following steps when being executed by the processor:
s1, acquiring a target river R0Calculating the ratio of the average width W to the length L of the shoreline to obtain a ratio parameter a, wherein a is L/W;
s2, performing geometric enhancement on the existing other river channel data based on the average width W, the shoreline length L and the proportion parameter a characteristics to obtain a river channel terrain image set IMAGER and a river channel data set DR aiming at the parameter characteristics, and establishing a mapping relation between the river channel terrain image set IMAGER and the river channel data set DR;
s3, splitting the river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to the IMAGER range of the river channel terrain image set, and ensuring the existence of the mapping relation among DR, DL and IMAGER;
s4, carrying out image classification based on Convolutional Neural Network (CNN) under deep learning on the river channel terrain image set IMAGER, dividing four types of river channels according to the trend and the shoreline curvature of the river channels to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)；
S5 for IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i＝1、2、3、4)；
S6, identifying the target river channel by using the key point detection CPM model, and selecting a proper training set model result DTi(i is 1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a set threshold valuej(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
S7, extracting terrain scatter database DBj(j is 1, 2, …, n) and target river R0River R with highest upstream similarityg1And with the target river R0River R with highest downstream similarityg2And calculating parameter proportions of scale1, scale2, ratio1 and ratio 2;
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S02H02)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)
wherein: sg1To indicate river course Rg1Cross sectional area, Sg2To indicate river course Rg2Cross sectional area, S01Representing a target river R0Upstream cross-sectional area, S02Representing a target river R0Upstream cross-sectional area, Hg1To indicate river course Rg1Elevation of lowest point of section, Hg2To indicate river course Rg2Elevation of lowest point of section, H01Representing a target river R0Elevation of lowest point of upstream section, H02Representing a target river R0The elevation of the lowest point of the upstream section;
s8 river channel RgIs represented by Rg1、Rg2River course formed by river course RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1、Rg2Landform scatter data to target river R0And splicing in the shoreline range according to ratio1 and ratio2, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the river channel.
The third aspect of the invention provides a readable storage medium, wherein the readable storage medium comprises a program of a method for generating a high-precision two-dimensional terrain of a one-way river based on deep learning, and when the program of the method for generating the high-precision two-dimensional terrain of the one-way river based on deep learning is executed by a processor, the steps of the method for generating the high-precision two-dimensional terrain of the one-way river based on deep learning are realized.
Compared with the prior art, the technical scheme of the invention has the beneficial effects that: the invention provides a method for generating a high-precision two-dimensional terrain of a one-way river channel based on deep learning, which comprises the steps of preprocessing the existing river channel measurement data according to the characteristics of a target river channel based on the deep learning according to the existing other river channel measurement data, target river channel images and upstream and downstream section data, establishing a one-to-one mapping relation of an image set, river channel data sets and the two of the existing river channel data, proposing a method for classifying river channel quadrants, classifying the image set of the existing river channel data by using a convolutional neural network under the deep learning, then training a river channel shoreline image carrying the characteristics of key points by using key point detection CPM, identifying the key points of the target river channel by using a trained model, matching a plurality of alternative river channel sets similar to the target river channel shape, and finally matching the two closest river channels in the alternative river channel sets by using section shapes, and carrying out scaling reconstruction on the terrain scattered point data corresponding to the two river channels by utilizing the similarity proportion, and obtaining the high-precision two-dimensional terrain scattered point data of the target river channel after smoothing.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention.
Fig. 2 is a schematic diagram of river channel classification according to different quadrants spanned by the present invention.
FIG. 3 shows the target river R in example 40Schematic representation.
Fig. 4 is a schematic diagram of the existing river image set IMAGER in example 4.
Fig. 5 is a schematic diagram of a shoreline data set DL and a topographic data set DB in example 4.
Fig. 6 is a diagram showing the detection of the key points in example 4.
Fig. 7 is a schematic cross-sectional view of the upstream river channel Rg1 with the highest similarity of the target river channel in example 4.
Fig. 8 is a schematic cross-sectional view of the downstream channel Rg2 with the highest similarity of the target channel in example 4.
Detailed Description
The present invention will be further described with reference to the accompanying drawings, and it should be noted that the present embodiment is based on the technical solution, and gives a detailed implementation and a specific operation procedure. But is for illustrative purposes only, and the scope of the present invention is not limited to the present embodiment and is not to be construed as being limited thereto.
Example 1
As shown in fig. 1, the present embodiment provides a method for generating a high-precision two-dimensional terrain of a one-way river based on deep learning, including the following steps:
s1, acquiring a target river R0The average width W and the shoreline length L of the first and second images are calculated, and a ratio parameter a is obtainedL/W；
S2, performing geometric enhancement on the existing other river channel data based on the average width W, the shoreline length L and the proportion parameter a characteristics to obtain a river channel terrain image set IMAGER and a river channel data set DR aiming at the parameter characteristics, and establishing a mapping relation between the river channel terrain image set IMAGER and the river channel data set DR;
s3, splitting the river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to the IMAGER range of the river channel terrain image set, and ensuring the existence of the mapping relation among DR, DL and IMAGER;
s4, carrying out image classification based on Convolutional Neural Network (CNN) under deep learning on the river channel terrain image set IMAGER, dividing four types of river channels according to the trend and the shoreline curvature of the river channels to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)；
S5 for IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i＝1、2、3、4)；
S6, identifying the target river channel by using the key point detection CPM model, and selecting a proper training set model result DTi(i is 1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a set threshold valuej(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
S7, extracting terrain scatter database DBj(j is 1, 2, …, n) and target river R0River R with highest upstream similarityg1And with the target river R0River R with highest downstream similarityg2And calculating parameter proportions of scale1, scale2, ratio1 and ratio 2;
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S02H02)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)
wherein: sg1To indicate river course Rg1Cross sectional area, Sg2To indicate river course Rg2Cross sectional area, S01Representing a target river R0Upstream cross-sectional area, S02Representing a target river R0Upstream cross-sectional area, Hg1To indicate river course Rg1Elevation of lowest point of section, Hg2To indicate river course Rg2Elevation of lowest point of section, H01Representing a target river R0Elevation of lowest point of upstream section, H02Representing a target river R0The elevation of the lowest point of the upstream section;
s8 river channel RgIs represented by Rg1、Rg2River course formed by river course RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1、Rg2Landform scatter data to target river R0And splicing in the shoreline range according to ratio1 and ratio2, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the river channel.
The embodiment preprocesses the existing river channel measurement data according to the existing other river channel measurement data, a target river channel image and upstream and downstream section data based on deep learning and target river channel characteristics, establishes a one-to-one mapping relation of an image set of the existing river channel data, a river channel data set and the two, provides a method for river channel quadrant classification, classifies the image set of the existing river channel data by using a convolutional neural network under deep learning, trains river channel shoreline images carrying key point characteristics by using key point detection CPM, identifies key points of the target river channel by using a trained model, matches a plurality of alternative river channel sets similar to the target river channel shape, matches two river channels closest to the target river channel shape in the alternative river channel sets in the section shape, and reconstructs terrain scatter data corresponding to the two river channels by using a similar proportion, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the target river channel.
Further, the specific process of step S2 is:
s2.1, collecting other river channel data, generating an initial high-definition image set and an initial river channel data set, establishing a mapping relation between the initial high-definition image set and the initial river channel data set, and ensuring that subsequent changes are reflected to the river channel data set;
s2.2, according to the parameter characteristics involved in the step S1, utilizing a deep learning preprocessing network to carry out image preprocessing, including image scaling, image rotation and image shearing, to obtain a classification river channel image set IMAGER and a river channel data set DR, wherein the river channel image passing through the preprocessing network has the characteristics that:
the river channel is a one-way river channel without branch;
secondly, the sizes of all river channel images are consistent;
thirdly, the image areas occupied by the river channels in the single river channel image are approximately similar;
the river channel data set DR comprises initial and terminal 2 plane straight lines of a one-way river channel, 2 plane curves of the left and right banks of the river channel and river channel terrain elevation scattered point data in the range of the river channel.
Further, the specific process of step S3 is:
for the river data set DR obtained by processing in step S2, the front line segment is extracted and split into the shoreline data set DL, the river terrain elevation scatter data is extracted as the terrain data set DB, and it is ensured that the mapping relationship between DR, DL and IMAGER still exists.
Further, as shown in fig. 2, the specific process of step S4 is:
s4.1, dividing four quadrants, limiting the flow direction of the river channel to be from west to east, enabling the river channels to cross over 3 quadrants in total, dividing the river channels into four categories according to the difference of the quadrants crossed by the river channels in sequence, and marking the four categories as IR1, IR2, IR3 and IR4 respectively, wherein the specific classification rule is as follows:
IR 1: the river channel sequentially passes through a second quadrant, a third quadrant and a fourth quadrant;
IR 2: the river channel sequentially passes through a second quadrant, a first quadrant and a fourth quadrant;
IR 3: the river channel sequentially passes through a third quadrant, a second quadrant and a first quadrant;
IR 4: the river channel sequentially passes through a third quadrant, a fourth quadrant and a first quadrant;
s4.2, carrying out deep learning on the classified river channel terrain image set IMAGER by utilizing the provided river channel quadrant classification ruleObtaining a river channel terrain image classification set IMAGER based on image classification of a Convolutional Neural Network (CNN)i(i＝1、2、3、4)。
Further, the specific process of step S5 is:
s5.1, definition of IMAGERiThe identifying key point features in (i ═ 1, 2, 3, and 4) include: starting point 1 of a left bank; a point 2 with the largest vertical distance from a connecting line from the starting point to the end point in the left bank line; left bank end point 3; starting point 4 of right bank; a point 5 with the largest vertical distance from the starting point to the end point in the right bank line; right bank end point 6;
s5.2, IMAGER with key point characteristicsi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using a key point detection CPM model under deep learning, training and testing, and establishing a training set model result DTi(i＝1、2、3、4)。
Further, the specific process of step S6 is:
s6.1, performing key point identification on the target river channel by using the established key point detection CPM model;
s6.2, selecting a proper training set model result DTi(i-1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a preset threshold (90%, 95%,. 99% and the like)j(j is 1, 2, …, n), extracting the existing corresponding terrain scatter database DBj(j＝1、2、…、n)。
Further, the specific process of step S7 is:
s7.1, extracting the terrain scatter database DBj(j is 1, 2, …, n) carrying out triangulation interpolation of the landforms of the starting point section and the ending point section;
s7.2, extracting a target river R0Cross-sectional area S of the cross-section at the upstream and downstream 201、S02Elevation of lowest point of cross section H01、H02；
S7.3, extracting and targeting river R0River R with highest similarity between upstream section area and lowest section point elevation parametersg1And river channel R with highest similarity of elevation parameters of downstream section area and lowest point of sectiong2Counting the corresponding parameter Sg1、Sg2Elevation of lowest point of cross section Hg1、Hg2And calculating the parameter proportion:
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S01H01)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)。
further, the specific process of step S8 is:
s8.1, river channel RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1And Rg2Topographic scatter elevation data;
s8.2, according to the target river R0Shore range scaling Rg1And Rg2A data range;
s8.3, taking Rg1Upstream to downstream ratio1 of proportional length of channel and Rg2Splicing river channels with proportional lengths from the downstream to the upstream ratio 2;
and S8.4, smoothing the spliced topographic scattered point elevation data to obtain the high-precision two-dimensional topographic scattered point data of the river.
Example 2
The embodiment provides a system for generating a high-precision two-dimensional terrain of a one-way river based on deep learning, which comprises a memory and a processor, wherein the memory comprises a method program for generating the high-precision two-dimensional terrain of the one-way river based on deep learning, and when the processor executes the method program, the following steps are realized:
s1, acquiring a target river R0Calculating the ratio of the average width W to the length L of the shoreline to obtain a ratio parameter a, wherein a is L/W;
s2, performing geometric enhancement on the existing other river channel data based on the average width W, the shoreline length L and the proportion parameter a characteristics to obtain a river channel terrain image set IMAGER and a river channel data set DR aiming at the parameter characteristics, and establishing a mapping relation between the river channel terrain image set IMAGER and the river channel data set DR;
s3, splitting the river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to the IMAGER range of the river channel terrain image set, and ensuring the existence of the mapping relation among DR, DL and IMAGER;
s4, carrying out image classification based on Convolutional Neural Network (CNN) under deep learning on the river channel terrain image set IMAGER, dividing four types of river channels according to the trend and the shoreline curvature of the river channels to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)；
S5 for IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i＝1、2、3、4)；
S6, identifying the target river channel by using the key point detection CPM model, and selecting a proper training set model result DTi(i is 1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a set threshold valuej(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
S7, extracting terrain scatter database DBj(j is 1, 2, …, n) and target river R0River R with highest upstream similarityg1And with the target river R0River R with highest downstream similarityg2And calculating parameter proportions of scale1, scale2, ratio1 and ratio 2;
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S02H02)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)
wherein: sg1To indicate river course Rg1Cross sectional area, Sg2To indicate river course Rg2Cross sectional area, S01Representing a target river R0Upstream cross-sectional area, S02Representing a target river R0Upstream ofCross-sectional area, Hg1To indicate river course Rg1Elevation of lowest point of section, Hg2To indicate river course Rg2Elevation of lowest point of section, H01Representing a target river R0Elevation of lowest point of upstream section, H02Representing a target river R0The elevation of the lowest point of the upstream section;
s8 river channel RgIs represented by Rg1、Rg2River course formed by river course RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1、Rg2Landform scatter data to target river R0And splicing in the shoreline range according to ratio1 and ratio2, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the river channel.
Example 3
The embodiment provides a readable storage medium, the readable storage medium comprises a program of a method for generating a high-precision two-dimensional terrain of a one-way river based on deep learning, and when the program of the method for generating the high-precision two-dimensional terrain of the one-way river based on deep learning is executed by a processor, the steps of the method for generating the high-precision two-dimensional terrain of the one-way river based on deep learning are realized.
Example 4
This example illustrates the method of the present invention through a set of experiments, which include the following steps:
1. as shown in fig. 3, the target river R0And (3) only one remote sensing base map, left and right hand-drawn shorelines and upstream and downstream actual measurement section points exist, and the calculation process a is 0.1 according to the step S1 after the average width 600m and the length 6000m of the shoreline of the river are counted.
2. And (3) collecting the measured data of other riverways, preprocessing the riverway data by using geometric enhancement to obtain the existing riverway image set IMAGER and the riverway data set DR, specifically referring to FIG. 4, and establishing a mapping relation between the two sets.
3. Splitting a river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to an IMAGER range of a river channel image set, specifically referring to FIG. 5, and ensuring that a mapping relation among DR, DL and IMAGER exists;
4. for the classified river terrain image set IThe method comprises the steps that image classification based on a Convolutional Neural Network (CNN) is carried out on the MAGER under deep learning, four types of river channels are divided according to the trend and the curvature of a shoreline of the river channel, and a river channel terrain image classification set IMAGER is obtainedi(i＝1、2、3、4)；
5. For IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i ═ 1, 2, 3, 4), the key point identification schematic is shown in fig. 6;
6. selecting proper training set model result DT1Identifying the target river channel by using a key point detection CPM model, and finding out a plurality of known river channels R with the similarity of 90% to the bank line of the target river channelj(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
7. In a terrain scatter database DBj(j ═ 1, 2, …, n) of the matching target river R0River R with highest upstream similarity and highest downstream similarityg1、Rg2The section morphology, as shown in fig. 7 and 8, and the parameter ratios scale1 of 1.05, scale2 of 0.90, ratio1 of 53.8% and ratio2 of 46.2% were calculated;
8. synchronously scaling R according to terrain elevation proportion of 1.05 and 0.90g1、Rg2Landform scatter data to target river R0Taking R within the range of shoreline according to 53.8% and 46.2%g1、Rg2And splicing the data, and smoothing to obtain the high-precision two-dimensional terrain scattered point data of the river.
It should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.
Claims (10)
1. A method for generating a high-precision two-dimensional terrain of a one-way river channel based on deep learning is characterized by comprising the following steps:
s1, acquiring a target river R0Calculating the ratio of the average width W to the length L of the shoreline to obtain a ratio parameter a, wherein a is L/W;
s2, performing geometric enhancement on the existing other river channel data based on the average width W, the shoreline length L and the proportion parameter a characteristics to obtain a river channel terrain image set IMAGER and a river channel data set DR aiming at the parameter characteristics, and establishing a mapping relation between the river channel terrain image set IMAGER and the river channel data set DR;
s3, splitting the river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to the IMAGER range of the river channel terrain image set, and ensuring the existence of the mapping relation among DR, DL and IMAGER;
s4, carrying out image classification based on the convolutional neural network under deep learning of the river channel terrain image set IMAGER, dividing four types of river channels according to the trend and the bank curvature of the river channels to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)；
S5 for IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i＝1、2、3、4)；
S6, identifying the target river channel by using the key point detection CPM model, and selecting a proper training set model result DTi(i is 1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a set threshold valuej(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
S7, extracting terrain scatter database DBj(j is 1, 2, …, n) and target river R0River R with highest upstream similarityg1And with the target river R0River R with highest downstream similarityg2And calculating parameter proportions of scale1, scale2, ratio1 and ratio 2;
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S02H02)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)
wherein: sg1To indicate river course Rg1Cross sectional area, Sg2To indicate river course Rg2Cross sectional area, S01Representing a target river R0Upstream cross-sectional area, S02Representing a target river R0Upstream cross-sectional area, Hg1To indicate river course Rg1Elevation of lowest point of section, Hg2To indicate river course Rg2Elevation of lowest point of section, H01Representing a target river R0Elevation of lowest point of upstream section, H02Representing a target river R0The elevation of the lowest point of the upstream section;
s8 river channel RgIs represented by Rg1、Rg2River course formed by river course RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1、Rg2Landform scatter data to target river R0And splicing in the shoreline range according to ratio1 and ratio2, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the river channel.
2. The method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 1, wherein the specific process of the step S2 is as follows:
s2.1, collecting other river channel data, generating an initial high-definition image set and an initial river channel data set, establishing a mapping relation between the initial high-definition image set and the initial river channel data set, and ensuring that subsequent changes are reflected to the river channel data set;
s2.2, according to the parameter characteristics involved in the step S1, utilizing a deep learning preprocessing network to carry out image preprocessing, including image scaling, image rotation and image shearing, to obtain a classification river channel image set IMAGER and a river channel data set DR, wherein the river channel image passing through the preprocessing network has the characteristics that:
the river channel is a one-way river channel without branch;
secondly, the sizes of all river channel images are consistent;
thirdly, the image areas occupied by the river channels in the single river channel image are approximately similar;
the river channel data set DR comprises initial and terminal 2 plane straight lines of a one-way river channel, 2 plane curves of the left and right banks of the river channel and river channel terrain elevation scattered point data in the range of the river channel.
3. The method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 2, wherein the specific process of the step S3 is as follows:
for the river data set DR obtained by processing in step S2, the front line segment is extracted and split into the shoreline data set DL, the river terrain elevation scatter data is extracted as the terrain data set DB, and it is ensured that the mapping relationship between DR, DL and IMAGER still exists.
4. The method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 3, wherein the specific process of the step S4 is as follows:
s4.1, dividing four quadrants, limiting the flow direction of the river channel to be from west to east, enabling the river channels to cross over 3 quadrants in total, dividing the river channels into four categories according to the difference of the quadrants crossed by the river channels in sequence, and marking the four categories as IR1, IR2, IR3 and IR4 respectively, wherein the specific classification rule is as follows:
IR 1: the river channel sequentially passes through a second quadrant, a third quadrant and a fourth quadrant;
IR 2: the river channel sequentially passes through a second quadrant, a first quadrant and a fourth quadrant;
IR 3: the river channel sequentially passes through a third quadrant, a second quadrant and a first quadrant;
IR 4: the river channel sequentially passes through a third quadrant, a fourth quadrant and a first quadrant;
s4.2, carrying out image classification based on a convolutional neural network under deep learning on the classified river channel terrain image set IMAGER by utilizing the provided river channel quadrant classification rule to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)。
5. The method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 4, wherein the specific process of the step S5 is as follows:
s5.1, definition of IMAGERiThe identifying key point features in (i ═ 1, 2, 3, and 4) include: starting point 1 of a left bank; a point 2 with the largest vertical distance from a connecting line from the starting point to the end point in the left bank line; left bank end point 3; starting point 4 of right bank; a point 5 with the largest vertical distance from the starting point to the end point in the right bank line; right bank end point 6;
s5.2, IMAGER with key point characteristicsi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using a key point detection CPM model under deep learning, training and testing, and establishing a training set model result DTi(i＝1、2、3、4)。
6. The method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 5, wherein the specific process of the step S6 is as follows:
s6.1, performing key point identification on the target river channel by using the established key point detection CPM model;
s6.2, selecting a proper training set model result DTi(i-1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a preset threshold valuej(j is 1, 2, …, n), extracting the existing corresponding terrain scatter database DBj(j＝1、2、…、n)。
7. The method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 6, wherein the specific process of the step S7 is as follows:
s7.1, extracting the terrain scatter database DBj(j is 1, 2, …, n) carrying out triangulation interpolation of the landforms of the starting point section and the ending point section;
s7.2, extracting a target river R0Cross-sectional area S of the cross-section at the upstream and downstream 201、S02Elevation of lowest point of cross section H01、H02；
S7.3, extracting and targeting river R0River R with highest similarity between upstream section area and lowest section point elevation parametersg1And river channel R with highest similarity of elevation parameters of downstream section area and lowest point of sectiong2Counting the corresponding parameter Sg1、Sg2Elevation of lowest point of cross section Hg1、Hg2And calculating the parameter proportion:
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S01H01)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)。
8. the method for generating the high-precision two-dimensional terrain of the unidirectional river channel based on the deep learning of claim 7, wherein the specific process of the step S8 is as follows:
s8.1, river channel RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1And Rg2Topographic scatter elevation data;
s8.2, according to the target river R0Shore range scaling Rg1And Rg2A data range;
s8.3, taking Rg1Upstream to downstream ratio1 of proportional length of channel and Rg2Splicing river channels with proportional lengths from the downstream to the upstream ratio 2;
and S8.4, smoothing the spliced topographic scattered point elevation data to obtain the high-precision two-dimensional topographic scattered point data of the river.
9. A system for generating a high-precision two-dimensional terrain of a one-way river channel based on deep learning is characterized by comprising a memory and a processor, wherein the memory comprises a method program for generating the high-precision two-dimensional terrain of the one-way river channel based on deep learning, and the method program for generating the high-precision two-dimensional terrain of the one-way river channel based on deep learning realizes the following steps when being executed by the processor:
s1, acquiring a target river R0Calculating the ratio of the average width W to the length L of the shoreline to obtain a ratio parameter a, wherein a is L/W;
s2, performing geometric enhancement on the existing other river channel data based on the average width W, the shoreline length L and the proportion parameter a characteristics to obtain a river channel terrain image set IMAGER and a river channel data set DR aiming at the parameter characteristics, and establishing a mapping relation between the river channel terrain image set IMAGER and the river channel data set DR;
s3, splitting the river channel data set DR into a shoreline data set DL and a terrain data set DB corresponding to the IMAGER range of the river channel terrain image set, and ensuring the existence of the mapping relation among DR, DL and IMAGER;
s4, carrying out image classification based on the convolutional neural network under deep learning of the river channel terrain image set IMAGER, dividing four types of river channels according to the trend and the bank curvature of the river channels to obtain a river channel terrain image classification set IMAGERi(i＝1、2、3、4)；
S5 for IMAGERi(i is 1, 2, 3 and 4), performing river course shoreline feature recognition by using key point detection CPM under deep learning, training and verifying, and establishing a training set model result DTi(i＝1、2、3、4)；
S6, identifying the target river channel by using the key point detection CPM model, and selecting a proper training set model result DTi(i is 1, 2, 3, 4) and finding out a plurality of known river channels R with the similarity to the target river channel bank reaching a set threshold valuej(j is 1, 2, …, n), and extracting the existing corresponding terrain scatter-point database DBj(j＝1、2、…、n)；
S7, extracting terrain scatter database DBj(j is 1, 2, …, n) and target river R0River R with highest upstream similarityg1And with the target river R0River R with highest downstream similarityg2And calculating parameter proportions of scale1, scale2, ratio1 and ratio 2;
scale1＝(Sg1*Hg1)/(S01H01)
scale2＝(Sg2*Hg2)/(S02H02)
ratio1＝scale1/(scale1+scale2)
ratio2＝scale2/(scale1+scale2)
wherein: sg1To indicate river course Rg1Cross sectional area, Sg2To indicate river course Rg2Cross sectional area, S01Representing a target river R0Upstream cross-sectional area, S02Representing a target river R0Upstream cross-sectional area, Hg1To indicate river course Rg1Elevation of lowest point of section, Hg2To indicate river course Rg2Elevation of lowest point of section, H01Representing a target river R0Elevation of lowest point of upstream section, H02Representing a target river R0The elevation of the lowest point of the upstream section;
s8 river channel RgIs represented by Rg1、Rg2River course formed by river course RgSynchronously scaling R according to terrain elevation scale1 and scale2 as a baseg1、Rg2Landform scatter data to target river R0And splicing in the shoreline range according to ratio1 and ratio2, and smoothing to obtain the high-precision two-dimensional terrain scatter data of the river channel.
10. A readable storage medium, wherein the readable storage medium includes a program of a method for generating a high-precision two-dimensional terrain of a one-way river based on deep learning, and when the program of the method for generating a high-precision two-dimensional terrain of a one-way river based on deep learning is executed by a processor, the steps of the method for generating a high-precision two-dimensional terrain of a one-way river based on deep learning according to any one of claims 1 to 8 are implemented.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111212259.9A CN113888670B (en) | 2021-10-18 | 2021-10-18 | Method for generating high-precision two-dimensional terrain of one-way river channel based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111212259.9A CN113888670B (en) | 2021-10-18 | 2021-10-18 | Method for generating high-precision two-dimensional terrain of one-way river channel based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113888670A true CN113888670A (en) | 2022-01-04 |
| CN113888670B CN113888670B (en) | 2022-05-20 |
Family
ID=79003545
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111212259.9A Active CN113888670B (en) | 2021-10-18 | 2021-10-18 | Method for generating high-precision two-dimensional terrain of one-way river channel based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113888670B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385684A (en) * | 2023-04-10 | 2023-07-04 | 长江水利委员会水文局 | Riverway section automatic extraction method based on reservoir point cloud |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110298211A (en) * | 2018-03-21 | 2019-10-01 | 北京大学 | A kind of Methods Deriving Drainage Network based on deep learning and high-resolution remote sensing image |
| CN111681316A (en) * | 2020-06-24 | 2020-09-18 | 河南省水利勘测设计研究有限公司 | High-precision river terrain interpolation method |
| CN112148820A (en) * | 2020-08-24 | 2020-12-29 | 中国地质大学（武汉） | Underwater terrain data identification and service method and system based on deep learning |
| CN112967354A (en) * | 2021-03-17 | 2021-06-15 | 武汉大学 | Method for generating two-dimensional terrain based on limited fixed-section terrain and remote sensing image data |
-
2021
- 2021-10-18 CN CN202111212259.9A patent/CN113888670B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110298211A (en) * | 2018-03-21 | 2019-10-01 | 北京大学 | A kind of Methods Deriving Drainage Network based on deep learning and high-resolution remote sensing image |
| CN111681316A (en) * | 2020-06-24 | 2020-09-18 | 河南省水利勘测设计研究有限公司 | High-precision river terrain interpolation method |
| CN112148820A (en) * | 2020-08-24 | 2020-12-29 | 中国地质大学（武汉） | Underwater terrain data identification and service method and system based on deep learning |
| CN112967354A (en) * | 2021-03-17 | 2021-06-15 | 武汉大学 | Method for generating two-dimensional terrain based on limited fixed-section terrain and remote sensing image data |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116385684A (en) * | 2023-04-10 | 2023-07-04 | 长江水利委员会水文局 | Riverway section automatic extraction method based on reservoir point cloud |
| CN116385684B (en) * | 2023-04-10 | 2023-11-03 | 长江水利委员会水文局 | Riverway section automatic extraction method based on reservoir point cloud |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113888670B (en) | 2022-05-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110276732B (en) | Mountain area point cloud cavity repairing method considering topographic characteristic line elements | |
| CN104732553B (en) | A kind of Feature Points Extraction based on many laser assisted targets | |
| CN103810473B (en) | A kind of target identification method of human object based on HMM | |
| CN109948547A (en) | Urban green space landscape evaluation method, device, storage medium and terminal device | |
| WO2011023245A1 (en) | Realigning road networks in a digital map based on reliable road existence probability data | |
| CN112990183B (en) | Method, system and device for extracting homonymous strokes of offline handwritten Chinese characters | |
| CN112084869A (en) | Compact quadrilateral representation-based building target detection method | |
| CN107944437B (en) | A kind of Face detection method based on neural network and integral image | |
| CN109657614B (en) | Automatic road identification method in aerial photography road traffic accident scene investigation | |
| CN109408985A (en) | The accurate recognition methods in bridge steel structure crack based on computer vision | |
| CN108288269A (en) | Bridge pad disease automatic identifying method based on unmanned plane and convolutional neural networks | |
| CN112150804B (en) | City multi-type intersection identification method based on MaskRCNN algorithm | |
| CN111695431A (en) | Face recognition method, face recognition device, terminal equipment and storage medium | |
| CN113888670B (en) | Method for generating high-precision two-dimensional terrain of one-way river channel based on deep learning | |
| CN107895140A (en) | Porny identification method based on face complexion | |
| CN109635726A (en) | A kind of landslide identification method based on the symmetrical multiple dimensioned pond of depth network integration | |
| CN102542543A (en) | Block similarity-based interactive image segmenting method | |
| CN115345881A (en) | Pavement disease detection method based on computer vision | |
| CN109460761A (en) | Bank card number detection and recognition methods based on dimension cluster and multi-scale prediction | |
| CN107657262A (en) | A kind of computer automatic sorting Accuracy Assessment | |
| CN114463720A (en) | Lane line detection method based on line segment intersection-to-parallel ratio loss function | |
| CN115937708A (en) | High-definition satellite image-based roof information automatic identification method and device | |
| CN102663761A (en) | Linear vector and remote-sensing image automatic registration method for photographic map | |
| CN112001359B (en) | Heuristic multi-fork line repairing method based on pattern recognition and optimal allocation | |
| CN106652020B (en) | Model-based tower three-dimensional reconstruction method |
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 |