CN117288283B - River flow rate monitoring method and system based on video - Google Patents
River flow rate monitoring method and system based on video Download PDFInfo
- Publication number
- CN117288283B CN117288283B CN202311585610.8A CN202311585610A CN117288283B CN 117288283 B CN117288283 B CN 117288283B CN 202311585610 A CN202311585610 A CN 202311585610A CN 117288283 B CN117288283 B CN 117288283B
- Authority
- CN
- China
- Prior art keywords
- flow rate
- river
- sample point
- flow velocity
- obtaining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000012544 monitoring process Methods 0.000 title claims abstract description 26
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 25
- 238000005259 measurement Methods 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims description 24
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 6
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 4
- OGHNVEJMJSYVRP-UHFFFAOYSA-N carvedilol Chemical compound COC1=CC=CC=C1OCCNCC(O)COC1=CC=CC2=C1C1=CC=CC=C1N2 OGHNVEJMJSYVRP-UHFFFAOYSA-N 0.000 description 8
- 229940069210 coreg Drugs 0.000 description 7
- 208000034699 Vitreous floaters Diseases 0.000 description 6
- 230000000694 effects Effects 0.000 description 3
- 238000006124 Pilkington process Methods 0.000 description 2
- 230000000903 blocking effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 239000012530 fluid Substances 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000011144 upstream manufacturing Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000001373 regressive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01F—MEASURING VOLUME, VOLUME FLOW, MASS FLOW OR LIQUID LEVEL; METERING BY VOLUME
- G01F1/00—Measuring the volume flow or mass flow of fluid or fluent solid material wherein the fluid passes through a meter in a continuous flow
- G01F1/704—Measuring the volume flow or mass flow of fluid or fluent solid material wherein the fluid passes through a meter in a continuous flow using marked regions or existing inhomogeneities within the fluid stream, e.g. statistically occurring variations in a fluid parameter
- G01F1/708—Measuring the time taken to traverse a fixed distance
- G01F1/7086—Measuring the time taken to traverse a fixed distance using optical detecting arrangements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01P—MEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
- G01P5/00—Measuring speed of fluids, e.g. of air stream; Measuring speed of bodies relative to fluids, e.g. of ship, of aircraft
- G01P5/26—Measuring speed of fluids, e.g. of air stream; Measuring speed of bodies relative to fluids, e.g. of ship, of aircraft by measuring the direct influence of the streaming fluid on the properties of a detecting optical wave
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/766—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using regression, e.g. by projecting features on hyperplanes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/70—Labelling scene content, e.g. deriving syntactic or semantic representations
-
- 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)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Remote Sensing (AREA)
- Computational Linguistics (AREA)
- Aviation & Aerospace Engineering (AREA)
- Fluid Mechanics (AREA)
- Measuring Volume Flow (AREA)
Abstract
The invention relates to the technical field of video measurement, in particular to a river flow rate monitoring method and a river flow rate monitoring system based on video, which are used for collecting local river surface flow rates of measuring points in river video data and obtaining a river surface flow rate sequence; acquiring water depth data, and drawing a river cross section according to the water depth data; acquiring a sample point and a riverbed line on a river cross section; obtaining a flow velocity attenuation factor according to the distance from a sample point to a river bed line, and further obtaining a predicted flow velocity by combining a river surface flow velocity sequence; obtaining a predicted flow rate of a sample point according to the distance from the measuring point to the riverbed line and the local river surface flow rate; obtaining a confidence predicted flow rate according to the predicted flow rate and the predicted flow rate, and further obtaining a corrected flow rate of the sample point by combining the predicted flow rate; and obtaining the river flow rate and the average river flow rate of the river cross section according to the corrected flow rate of the sample point. The river flow and velocity monitoring method solves the problem that the existing river flow and velocity monitoring method has larger error.
Description
Technical Field
The invention relates to the technical field of video measurement, in particular to a river flow rate monitoring method and system based on video.
Background
Rivers are important components in human activities since ancient times, play important roles in running cities, along with the rapid development of economy, accurate measurement of water resources and reasonable allocation of water resources are important issues in modern society. The detection of river flow velocity is one of the tasks of water resource management and hydrologic research, and the devices commonly used for monitoring river flow velocity are an electric wave radar flow velocity meter, a flow velocity measurer, an acoustic Doppler flow velocity profiler and the like. In addition, more traditional float methods and float methods are used to calculate the flow rate of water by observing the distance and time an object moves over the water surface. The flow of the river is equal to the average flow velocity multiplied by the cross-sectional area, and can be simply estimated with a known cross-sectional area. River flow monitoring plays an important role in water resource management, flood early warning, ecological protection and energy utilization.
In summary, in the conventional river flow rate measurement method, devices such as a flow rate meter or a drift are required, and manual intervention or a monitoring device is often required to be placed at a specific position. The flow rate of the river is not uniformly distributed in the whole section, the flow rate of the water flow near the bank is smaller, and the flow rate of a single position is difficult to represent the average flow rate of the whole section of the river. Therefore, the existing method for monitoring river flow rate through a single position has large errors.
Disclosure of Invention
The invention provides a river flow rate monitoring method and a river flow rate monitoring system based on video, which are used for solving the problem of larger error existing in the existing river flow rate monitoring method.
In a first aspect, an embodiment of the present invention provides a video-based river discharge flow rate monitoring method comprising the steps of:
collecting local river surface flow velocity of a measuring point in river video data, and obtaining a river surface flow velocity sequence;
connecting the measuring points to obtain a measuring line; collecting water depths at all points on a measuring line, acquiring water depth data, and drawing a river cross section according to the water depth data; acquiring a sample point and a riverbed line on a river cross section; obtaining a flow velocity attenuation factor of the sample point according to the distance from the sample point to the riverbed line; obtaining the maximum measured flow rate according to the river surface flow rate sequence, and further obtaining the estimated flow rate of the sample point by combining the flow rate attenuation factors;
obtaining a flow velocity prediction model according to the distance from the measuring point to the riverbed line and the local river surface flow velocity, and further obtaining a predicted flow velocity of the sample point; obtaining a river flow velocity prediction confidence coefficient according to the predicted flow velocity and the predicted flow velocity; obtaining a predicted flow rate of a measurement point according to the flow rate prediction model, and further obtaining a flow rate comprehensive confidence coefficient by combining the local river surface flow rate and the river flow rate prediction confidence coefficient; obtaining a confidence prediction flow rate of the sample point according to the flow rate comprehensive confidence coefficient, and further obtaining a correction flow rate of the sample point by combining the prediction flow rate;
connecting adjacent sample points to obtain a closed area; and obtaining the regional average flow velocity according to the corrected flow velocity of all sample points on the edge of the closed region, and further obtaining the river flow and the river average flow velocity of the river cross section.
Further, the specific method for obtaining the sample points and the riverbed lines on the cross section of the river comprises the following steps:
sample points are arranged on the river cross section at intervals of a preset length and a preset width;
the bottom edge line of the river cross section is denoted as the riverbed line.
Further, the method for obtaining the flow velocity attenuation factor of the sample point according to the distance from the sample point to the riverbed line comprises the following specific steps:
the minimum value of the horizontal distance from the sample point to the river bed line is recorded as the nearest horizontal distance of the sample point;
the vertical distance from the sample point to the river bed line is recorded as the vertical distance from the sample point;
the maximum value of the nearest horizontal distance of all sample points is recorded as the maximum horizontal distance;
the maximum value of the vertical distances of all sample points is recorded as the maximum vertical distance;
the product of the nearest horizontal distance and the vertical distance of the sample point is recorded as the comprehensive distance of the sample point;
the product of the maximum horizontal distance and the maximum vertical distance is recorded as a maximum distance value;
the ratio of the comprehensive distance to the maximum distance value of the sample point is recorded as a flow velocity influence factor of the sample point;
the flow rate decay factor of the sample point is raised to a power of the exponent of the inverse of the flow rate influencing factor, with the natural constant as the base.
Further, the method for obtaining the maximum measured flow rate according to the river surface flow rate sequence and further obtaining the estimated flow rate of the sample point by combining the flow rate attenuation factors comprises the following specific steps:
recording the maximum element value in the river surface flow velocity sequence as the maximum measured flow velocity;
the product of the flow rate decay factor of the sample point and the maximum measured flow rate is recorded as the estimated flow rate of the sample point.
Further, the method for obtaining the flow velocity prediction model according to the distance from the measuring point to the riverbed line and the local river surface flow velocity, and further obtaining the predicted flow velocity of the sample point comprises the following specific steps:
acquiring the vertical distance and the nearest horizontal distance of a measuring point;
taking the vertical distance and the nearest horizontal distance as characteristic values, taking the local river surface flow velocity of the measuring point as a training set, training a machine learning model based on the adjacent points, and obtaining a flow velocity prediction model;
and obtaining the predicted flow rate of the sample point by using a flow rate prediction model according to the vertical distance and the nearest horizontal distance of the sample point.
Further, the method for obtaining the river flow velocity prediction confidence coefficient according to the predicted flow velocity and the predicted flow velocity comprises the following specific steps:
recording the absolute value of the difference value between the predicted flow rate and the predicted flow rate of the sample point as the flow rate prediction difference of the sample point;
and recording the inverse of the sum of the flow velocity prediction difference of the sample point and the preset value as a river flow velocity prediction confidence coefficient of the sample point.
Further, the method for obtaining the predicted flow rate of the measurement point according to the flow rate prediction model, and further obtaining the flow rate comprehensive confidence coefficient by combining the local river surface flow rate and the river flow rate prediction confidence coefficient comprises the following specific steps:
each sample point is marked as a sample point to be analyzed;
marking a preset number of measurement points closest to the sample point to be analyzed in the training set as neighborhood sample points of the sample point to be analyzed;
respectively marking each neighborhood sample point of the sample points to be analyzed as a neighborhood point to be analyzed;
adding sample points to be analyzed into a training set, training a machine learning model based on adjacent points, and obtaining a confidence flow rate prediction model;
obtaining a model predicted flow rate of the neighborhood point to be analyzed by using a confidence flow rate prediction model according to the characteristic value of the neighborhood point to be analyzed;
the square of the difference between the local river surface flow velocity and the predicted flow velocity of the neighborhood point to be analyzed is recorded as the predicted flow velocity difference of the neighborhood point to be analyzed;
marking the square of the difference between the local river surface flow velocity of the neighborhood point to be analyzed and the model predicted flow velocity as the model predicted flow velocity difference of the neighborhood point to be analyzed;
recording the difference value between the predicted flow velocity difference of the neighborhood point to be analyzed and the predicted flow velocity difference of the model as the predicted flow velocity confidence degree of the neighborhood point to be analyzed;
marking the average value of the confidence degrees of the predicted flow rates of all the neighborhood points to be analyzed as the model prediction confidence degree of the sample points to be analyzed;
the sum of the model prediction confidence coefficient and the river flow velocity prediction confidence coefficient of the sample point to be analyzed is recorded as the flow velocity confidence coefficient of the sample point to be analyzed;
and marking the normalized value of the flow velocity confidence of the sample point to be analyzed as the flow velocity comprehensive confidence coefficient of the sample point to be analyzed.
Further, the method for obtaining the confidence prediction flow rate of the sample point according to the flow rate comprehensive confidence coefficient and further obtaining the correction flow rate of the sample point by combining the prediction flow rate comprises the following specific steps:
taking the flow velocity comprehensive confidence coefficient as the confidence coefficient of the semi-supervised learning collaborative regression algorithm, and obtaining the confidence prediction flow velocity of each sample point by using the semi-supervised learning collaborative regression algorithm;
the flow rate comprehensive confidence coefficient is recorded as the confidence prediction flow rate weight of the sample point, and the difference value between the number 1 and the flow rate comprehensive confidence coefficient is recorded as the prediction flow rate weight of the sample point;
and (3) recording the weighted average of the estimated flow rate and the confidence predicted flow rate of the sample point as the corrected flow rate of the sample point.
Further, the method for obtaining the area average flow velocity according to the corrected flow velocity of all the sample points on the edge of the closed area, and further obtaining the river flow rate and the river average flow velocity of the cross section of the river comprises the following specific steps:
each closed region is respectively marked as a region to be analyzed;
the average value of the corrected flow rates of all sample points on the edge of the area to be analyzed is recorded as the area average flow rate of the area to be analyzed;
the product of the preset width and the preset length is recorded as the area of the area to be analyzed;
the product of the area to be analyzed and the average flow rate of the area is recorded as the local flow of the area to be analyzed;
the sum of the local flows of all the closed areas is recorded as the river flow of the cross section of the river;
the sum of the areas of all the closed areas is recorded as the cross-sectional area of the river;
the ratio of the river discharge of the river cross section to the river cross section area is taken as the average river flow velocity.
In a second aspect, an embodiment of the present invention further provides a system for monitoring river flow and velocity based on video, including a memory, a processor, and a computer program stored in the memory and running on the processor, where the processor implements the steps of any one of the methods described above when the processor executes the computer program.
The beneficial effects of the invention are as follows: according to the method, the condition that the flow velocity of different positions of the river is hindered by the river bed when the river flows is analyzed, the cross section of the river is drawn, the sample points are set, the flow velocity attenuation factors are constructed according to the space positions of the sample points, the estimated flow velocity is further obtained, and the influence of the friction force of the river bed on the flow velocity of the sample points at different positions is considered; according to the spatial position of the sample point, a machine learning method based on the adjacent point is used for obtaining the predicted flow rate of the sample point, and the influence of the spatial position of the sample point on the river flow rate at the sample point is considered; the influence of the difference value of the estimated flow rate and the predicted flow rate on the reliability of the predicted result is analyzed, the confidence parameters in the COREG algorithm are modified according to the estimated flow rate and the predicted flow rate, the flow rate comprehensive confidence coefficient of the sample point is obtained, the corrected flow rate of the sample point is further obtained, and the accuracy of river flow rate prediction at the sample point is improved; and finally, calculating the river flow rate of the cross section of the river and the average river flow rate by the corrected flow rates of all sample points, thereby solving the problem of larger error in the existing method for monitoring the river flow rate by a single position.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a flow chart of an embodiment of a video-based river flow rate monitoring method of the present invention;
FIG. 2 is a schematic view of a projectile apparatus;
FIG. 3 is a schematic illustration of a unmanned aerial vehicle route;
FIG. 4 is a schematic representation of a cross-sectional river section.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
An embodiment of a video-based river flow rate monitoring method of the present invention, as shown in fig. 1, includes:
step S1, acquiring video data of a river by using an unmanned plane, and acquiring local river surface flow velocity of a measuring point, thereby obtaining a river surface flow velocity sequence.
As shown in fig. 2, an unmanned aerial vehicle carrying a high-resolution CCD digital camera is hovered at a fixed height right above a river near the shoreThe empirical value is 10 meters, and the length of the unmanned aerial vehicle can be observed at the moment is +.>Width is->Is defined in the drawings. In the figure, a rectangular frame is a throwing device which is arranged at one end of the upstream of the unmanned aerial vehicle and can throw floaters, and the floaters are required to be bright in color and have larger difference with the water body environment. Firstly, throwing a floater towards the upstream direction of a river, and simultaneously, the unmanned plane starts to record river information. As shown in fig. 3, after the floaters fly out of the field of view of the unmanned aerial vehicle, the photographing is stopped, and then the floaters advance in the direction opposite to the shore of the river>Hovering again. The above operation is continued until the unmanned aerial vehicle arrives at the river to shore, and the co-measurement is obtained>Video data.
First, a training set is created by labeling floats using a LabelImg labeling tool. And training the YOLO network by using the manufactured training set, and storing the trained model. For acquisition and getAnd respectively executing a YOLO target detection algorithm on the video data to detect floaters in the video. When the float is detected for the first time in the video, the recording time is +.>Marking the intersection of float with field of view +.>The method comprises the steps of carrying out a first treatment on the surface of the The floating object is floating out of view, i.e. the recording time is +.>Marking the intersection of float with field of view +.>The method comprises the steps of carrying out a first treatment on the surface of the Will->And->The Euclidean distance between them is noted as floating distance +.>The floating object appearance period length is +.>. The center point of each video is recorded as the measurement point +.>The detected +.>Local river surface flow at the individual measuring points +.>Is the ratio of the floating distance to the length of the period of occurrence of the float. Co-acquisition->Detected +.>Local river surface flow velocity of each measuring point to obtain river surface flow velocity sequence +.>。
So far, the local river surface flow velocity and the river surface flow velocity sequence of the measuring point are obtained.
And S2, setting sample points, acquiring flow rate attenuation factors according to the distances between the sample points and the river bed, and acquiring estimated flow rate and predicted flow rate of the sample points through flow rate attenuation factors and river surface flow rate sequences respectively.
Due to the dynamic nature of river flow and the influence of geological topography, the friction force of the river bed and the bank has an effect of blocking the speed of the water flow, so that the closer to the bank, the smaller the flow speed, and the larger the flow speed of the river in the central area of the river. In general, river flow is also affected by the depth of the water, with greater depth resulting in wider channels of water flow, which will create less resistance to water flow. In consideration of the influence of the above factors, as shown in fig. 4, the present invention uses a sonar depth finder to measure the water depth of a river cross section at the measuring line by connecting measuring points to obtain the measuring line, obtains water depth data and depicts the river cross section, and the bottom edge line of the river cross section is the riverbed line. And by combiningDistance of (a) is interval (empirical value is +.>=1，/>=1) sample points are set on the cross section of the river, wherein each dotted line junction is a sample point, and +.>And a number of sample points. Sample dot->The minimum value of the horizontal distance to both sides of the river bed is recorded as the sample point +.>Is>Sample Point +.>The vertical distance to the river bed is noted as sample point +.>Is>. The maximum value of the nearest horizontal distance of all sample points is recorded as the maximum horizontal distance +.>The maximum value of the vertical distances from all sample points to the river bed is recorded as the maximum vertical distance +.>. Constructing a flow rate attenuation factor according to the spatial position of each sample point in the river section>Wherein, sample point->Flow rate attenuation factor->The calculation formula is as follows:
in the method, in the process of the invention,for sample points in river cross section>A flow rate decay factor at; />Representative sample point->Is a vertical distance of (2); />Representing sample points->Is the closest horizontal distance to the ground; />Is a natural constant; />Is the maximum vertical distance; />Is the maximum horizontal distance.
The flow rate attenuation factor of the sample points which are far from the river bed in the horizontal distance and the vertical distance is smaller, which means that the flow rate is reduced less, and the flow rate attenuation factor is smaller; and the flow rate attenuation factor of the sample point which is closer to the horizontal distance and the vertical distance of the river bed is larger, and the flow rate is reduced to a large extent. Often the place of greatest flow rate is also the place furthest from the offshore side and the riverbed. The friction force on the water flow exists on the river bank of the river bed, namely, the friction force has an effect on the speed of the water flow, so that the closer to the bank, the smaller the flow speed is, and the larger the flow speed of the river is in the central area of the river.
Based on the analysis, each element in the obtained river surface flow velocity sequence represents the river surface flow velocity of a local area, and the main reason for the difference between the river surface flow velocities of the local areas is the blocking effect of the river bed and the river bank on the water flow, and a flow velocity attenuation factor is constructed based on the reasons. In this embodiment, the maximum element value in the river surface flow velocity sequence is used as the maximum measured flow velocity with minimum obstructionFurther estimating estimated flow rates at all sample points according to the flow rate attenuation factors, and marking the maximum element value in the river surface flow rate sequence as the maximum measured flow rate +.>Sample Point->Estimated flow rate +.>Expressed as sample points +.>Flow rate attenuation factor->And maximum measured flow rate->Is a product of (a) and (b).
The vertical distance and the nearest horizontal distance of all the measuring points are obtained, the vertical distance and the nearest horizontal distance are taken as characteristic values, and the local river surface flow velocity of the measuring points is taken as a training set and is called as annotation dataUse +.>Training a machine learning model KNN regressor based on adjacent points to obtain a trained model, and marking the trained model as a flow velocity prediction model +.>Wherein->Representing the use of a flow prediction model +.>The obtained sample points->Is provided for the flow rate. And predicting the predicted flow rates at all the sample points according to the characteristic values of the sample points by using a flow rate prediction model. KNN regressor algorithm model by +.>The nearest neighbor samples predict the regression value of the new sample, set +.>=4, i.e. each sample point has 4 adjacent sample points, and the flow velocity of a certain point is considered to be most affected by the four adjacent areas, i.e. the KNN algorithm is a known technique, and is not described in detail.
Thus, the estimated flow velocity and the predicted flow velocity of the sample point are obtained.
And S3, modifying confidence parameters in the COREG algorithm according to the estimated flow rate and the predicted flow rate, constructing flow rate confidence coefficients, obtaining the flow rate comprehensive confidence coefficient of each sample point, and further obtaining the corrected flow rate of the sample point.
When the semi-supervised learning collaborative regression algorithm COREG is used for predicting the flow velocity of the sample point, the confidence coefficient in the COREG algorithm is difficult to determine, and the confidence parameters in the COREG algorithm are modified by the estimated velocity and the predicted velocity of the sample point which are respectively obtained by the two methods. First, the flow rate can be estimated according to the estimated flow rateSample Point is acquired->Confidence coefficient of river flow rate prediction +.>：
In the method, in the process of the invention,for sample dot->A river flow rate prediction confidence coefficient; />Sample Point for use with flow prediction model +.>Is provided for the flow rate of the fluid; />For sample dot->Estimated flow rate at the location; />The empirical value is 0.005, which is a very small constant value, prevents the case of zero denominator, and also defines the maximum value of the confidence coefficient.
The greater the confidence coefficient for the flow rate at that point, the greater the confidence when the flow rates estimated by the two methods described herein are the closer.
The collaborative training in the semi-supervised learning collaborative regression algorithm COREG requires that the sample with the highest confidence coefficient is continuously added into the training set of the regressive for retraining until the termination condition is met. In order to improve the accuracy of flow rate prediction, the method further selects the sample points added into the training set by analyzing the influence of the sample addition before and after the sample addition on the prediction result of the sample adjacent points. And if the initial training set is a measurement point, the adjacent sample points of the sample points are all measurement points. Thus, according to the sample pointsObtaining sample point +.>Flow rate confidence +.>：
In the method, in the process of the invention,for sample dot->Flow rate confidence of (2); />For sample dot->In marking data +.>A number of adjacent sample points in the sample; />A flow velocity prediction model; />Sample Point for use with flow prediction model +.>Is>Predicted flow rates for adjacent sample points; />For sample dot->Is>Local river surface flow rates adjacent to the sample points; />To add sample points->Adding a flow velocity prediction model obtained after the training set; />For use +.>The obtained sample points->Is>Predicted flow rates for adjacent sample points; />For sample dot->Is a river flow rate prediction confidence coefficient.
If the sample is to be takenThe smaller the difference between the predicted flow velocity of the adjacent sample points calculated by using the flow velocity prediction model and the measured local river surface flow velocity is before and after the training set, the description of the sample pointsThe smaller the degree of influence on the predicted flow rate of the neighboring sample points after addition to the training set, the greater the flow rate confidence, indicating the addition of sample points +.>The greater the benefit to the training set, the more reliable the predicted outcome.
Spot the sampleNormalization of flow rate confidence of (c)The value is recorded as sample point->Is>And acquiring flow velocity comprehensive confidence coefficients of all the sample points, and adding the sample point with the maximum flow velocity comprehensive confidence coefficient into the training set. Obtaining a confidence predicted flow rate for each sample point using a modified COREG algorithm, substituting the sample point +.>Is recorded as +.>。
Defining sample pointsIs a modified flow rate of->The method comprises the following steps:
wherein,for sample dot->Is provided for the flow rate of the fluid; />For sample dot->Is a flow rate integrated confidence coefficient; />Is the sample point calculated by the flow rate decay factor +.>Estimated flow rate at->Sample points obtained by using semi-supervised learning collaborative regression algorithm COREG>Is a confidence predicted flow rate.
Sample pointThe higher the flow rate integrated confidence coefficient, the greater the weight of the predicted flow rate in the corrected flow rate, and the smaller the weight of the predicted flow rate in the corrected flow rate.
To this end, a corrected flow rate of the sample point is obtained.
And S4, obtaining the river flow of the cross section of the river according to the corrected flow velocity of the sample point, and further obtaining the average flow velocity of the river.
All adjacent sample points are connected, and the river cross section is divided intoA closed region, will be->Individual closure areas->The mean value of the corrected flow rates for all sample points on the edge of (a) is recorded as the area mean flow rate +.>，/>The value of (2) is 1 to +.>An integer therebetween.
Further, calculating river cross section from the area average flow velocityRiver dischargeThe unit is->River flow of river cross section>The calculation formula of (2) is as follows:
in the method, in the process of the invention,river flow is the cross section of the river; />Sampling interval of the sample point, namely area of the closed region; />Is the number of closed regions; />Is the>Individual closure areas->Is a region average flow rate of (c).
The sum of the areas of all the closed areas is recorded as the area of the cross section of the river, and the ratio of the flow rate of the river to the area of the cross section of the river is further used as the average flow rate of the river, wherein the unit is。
So far, the river flow rate and the average river flow velocity of the cross section of the river are obtained.
Based on the same inventive concept as the above method, the embodiment of the present invention further provides a video-based river flow rate monitoring system, which includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above video-based river flow rate monitoring methods when executing the computer program.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (6)
1. The river flow rate monitoring method based on the video is characterized by comprising the following steps of:
collecting local river surface flow velocity of a measuring point in river video data, and obtaining a river surface flow velocity sequence;
connecting the measuring points to obtain a measuring line; collecting water depths at all points on a measuring line, acquiring water depth data, and drawing a river cross section according to the water depth data; acquiring a sample point and a riverbed line on a river cross section; obtaining a flow velocity attenuation factor of the sample point according to the horizontal distance and the vertical distance from the sample point to the riverbed line; obtaining the maximum measured flow rate according to the river surface flow rate sequence, and further obtaining the estimated flow rate of the sample point by combining the flow rate attenuation factors;
obtaining a flow velocity prediction model according to the distance from the measuring point to the riverbed line and the local river surface flow velocity, and further obtaining a predicted flow velocity of the sample point; obtaining a river flow velocity prediction confidence coefficient according to the predicted flow velocity and the predicted flow velocity; obtaining a predicted flow rate of a measurement point according to the flow rate prediction model, and further obtaining a flow rate comprehensive confidence coefficient by combining the local river surface flow rate and the river flow rate prediction confidence coefficient; obtaining a confidence prediction flow rate of the sample point according to the flow rate comprehensive confidence coefficient, and further obtaining a correction flow rate of the sample point by combining the prediction flow rate;
connecting adjacent sample points to obtain a closed area; obtaining the regional average flow velocity according to the corrected flow velocity of all sample points on the edge of the closed region, and further obtaining the river flow and the river average flow velocity of the river cross section;
the method for obtaining the flow velocity attenuation factor of the sample point according to the horizontal distance and the vertical distance from the sample point to the riverbed line comprises the following specific steps:
the minimum value of the horizontal distance from the sample point to the river bed line is recorded as the nearest horizontal distance of the sample point;
the vertical distance from the sample point to the river bed line is recorded as the vertical distance from the sample point;
the maximum value of the nearest horizontal distance of all sample points is recorded as the maximum horizontal distance;
the maximum value of the vertical distances of all sample points is recorded as the maximum vertical distance;
the product of the nearest horizontal distance and the vertical distance of the sample point is recorded as the comprehensive distance of the sample point;
the product of the maximum horizontal distance and the maximum vertical distance is recorded as a maximum distance value;
the ratio of the comprehensive distance to the maximum distance value of the sample point is recorded as a flow velocity influence factor of the sample point;
the power of the inverse number of the flow velocity influence factor with the natural constant as a base number is marked as a flow velocity attenuation factor of the sample point;
the method for obtaining the flow velocity prediction model according to the distance from the measuring point to the riverbed line and the local river surface flow velocity, and further obtaining the predicted flow velocity of the sample point comprises the following specific steps:
acquiring the vertical distance and the nearest horizontal distance of a measuring point;
taking the vertical distance and the nearest horizontal distance as characteristic values, taking the local river surface flow velocity of the measuring point as a training set, training a machine learning model based on the adjacent points, and obtaining a flow velocity prediction model;
obtaining a predicted flow rate of the sample point by using a flow rate prediction model according to the vertical distance and the nearest horizontal distance of the sample point;
the method for obtaining the flow rate comprehensive confidence coefficient by combining the local river surface flow rate and the river flow rate prediction confidence coefficient comprises the following specific steps:
each sample point is marked as a sample point to be analyzed;
marking a preset number of measurement points closest to the sample point to be analyzed in the training set as neighborhood sample points of the sample point to be analyzed;
respectively marking each neighborhood sample point of the sample points to be analyzed as a neighborhood point to be analyzed;
adding sample points to be analyzed into a training set, training a machine learning model based on adjacent points, and obtaining a confidence flow rate prediction model;
obtaining a model predicted flow rate of the neighborhood point to be analyzed by using a confidence flow rate prediction model according to the characteristic value of the neighborhood point to be analyzed;
the square of the difference between the local river surface flow velocity and the predicted flow velocity of the neighborhood point to be analyzed is recorded as the predicted flow velocity difference of the neighborhood point to be analyzed;
marking the square of the difference between the local river surface flow velocity of the neighborhood point to be analyzed and the model predicted flow velocity as the model predicted flow velocity difference of the neighborhood point to be analyzed;
recording the difference value between the predicted flow velocity difference of the neighborhood point to be analyzed and the predicted flow velocity difference of the model as the predicted flow velocity confidence degree of the neighborhood point to be analyzed;
marking the average value of the confidence degrees of the predicted flow rates of all the neighborhood points to be analyzed as the model prediction confidence degree of the sample points to be analyzed;
the sum of the model prediction confidence coefficient and the river flow velocity prediction confidence coefficient of the sample point to be analyzed is recorded as the flow velocity confidence coefficient of the sample point to be analyzed;
marking the normalized value of the flow velocity confidence coefficient of the sample point to be analyzed as the flow velocity comprehensive confidence coefficient of the sample point to be analyzed;
the method for obtaining the confidence prediction flow rate of the sample point according to the flow rate comprehensive confidence coefficient and further obtaining the correction flow rate of the sample point by combining the prediction flow rate comprises the following specific steps:
taking the flow velocity comprehensive confidence coefficient as the confidence coefficient of the semi-supervised learning collaborative regression algorithm, and obtaining the confidence prediction flow velocity of each sample point by using the semi-supervised learning collaborative regression algorithm;
the flow rate comprehensive confidence coefficient is recorded as the confidence prediction flow rate weight of the sample point, and the difference value between the number 1 and the flow rate comprehensive confidence coefficient is recorded as the prediction flow rate weight of the sample point;
and (3) recording the weighted average of the estimated flow rate and the confidence predicted flow rate of the sample point as the corrected flow rate of the sample point.
2. The method for monitoring the flow rate of a river based on video according to claim 1, wherein the step of obtaining the sample points and the riverbed line on the cross section of the river comprises the following specific steps:
sample points are arranged on the river cross section at intervals of a preset length and a preset width;
the bottom edge line of the river cross section is denoted as the riverbed line.
3. The method for monitoring river discharge and flow rate based on video as claimed in claim 1, wherein the method for obtaining the maximum measured flow rate according to the river surface flow rate sequence and further obtaining the estimated flow rate of the sample point by combining the flow rate attenuation factor comprises the following specific steps:
recording the maximum element value in the river surface flow velocity sequence as the maximum measured flow velocity;
the product of the flow rate decay factor of the sample point and the maximum measured flow rate is recorded as the estimated flow rate of the sample point.
4. The method for monitoring river flow rate based on video of claim 1, wherein the obtaining the predicted confidence coefficient of the river flow rate according to the predicted flow rate and the predicted flow rate comprises the following specific steps:
recording the absolute value of the difference value between the predicted flow rate and the predicted flow rate of the sample point as the flow rate prediction difference of the sample point;
and recording the inverse of the sum of the flow velocity prediction difference of the sample point and the preset value as a river flow velocity prediction confidence coefficient of the sample point.
5. The method for monitoring the river discharge and the flow rate based on the video as claimed in claim 2, wherein the method for obtaining the area average flow rate according to the corrected flow rates of all the sample points on the edge of the closed area, and further obtaining the river discharge and the river average flow rate of the river cross section comprises the following specific steps:
each closed region is respectively marked as a region to be analyzed;
the average value of the corrected flow rates of all sample points on the edge of the area to be analyzed is recorded as the area average flow rate of the area to be analyzed;
the product of the preset width and the preset length is recorded as the area of the area to be analyzed;
the product of the area to be analyzed and the average flow rate of the area is recorded as the local flow of the area to be analyzed;
the sum of the local flows of all the closed areas is recorded as the river flow of the cross section of the river;
the sum of the areas of all the closed areas is recorded as the cross-sectional area of the river;
the ratio of the river discharge of the river cross section to the river cross section area is taken as the average river flow velocity.
6. A video-based river flow rate monitoring system comprising a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the method according to any one of claims 1-5 when the computer program is executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311585610.8A CN117288283B (en) | 2023-11-27 | 2023-11-27 | River flow rate monitoring method and system based on video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311585610.8A CN117288283B (en) | 2023-11-27 | 2023-11-27 | River flow rate monitoring method and system based on video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117288283A CN117288283A (en) | 2023-12-26 |
| CN117288283B true CN117288283B (en) | 2024-02-06 |
Family
ID=89258929
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311585610.8A Active CN117288283B (en) | 2023-11-27 | 2023-11-27 | River flow rate monitoring method and system based on video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117288283B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101788692A (en) * | 2009-11-24 | 2010-07-28 | 西北电网有限公司 | Short-time wind speed forecasting method based on neural network |
| CN103698553A (en) * | 2013-12-26 | 2014-04-02 | 天津大学 | Novel surface flow field velocity measurement system and velocity measurement method on basis of infrared image pickup |
| CN113155107A (en) * | 2021-05-25 | 2021-07-23 | 华北水利水电大学 | Irregular river channel section flow measuring device and method |
| CN114397476A (en) * | 2021-11-15 | 2022-04-26 | 河海大学 | Flow velocity effectiveness identification and correction method for frequency domain space-time image velocity measurement |
| DE102020134966A1 (en) * | 2020-12-28 | 2022-06-30 | Okeanos Consulting GbR (vertretungsberechtigte Gesellschafter: Dr. Henning Oppel, 44799 Bochum; Dr. Benjamin Mewes, 44795 Bochum) | Metrological arrangement for evaluating water level measurement data for non-invasive water level-discharge quantity correlation in running waters, as well as method and use |
| CN115600044A (en) * | 2022-11-28 | 2023-01-13 | 湖南大学（Cn） | River section flow calculation method, device, equipment and storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8339584B2 (en) * | 2010-05-21 | 2012-12-25 | Teledyne Technologies Incorporated | Velocity measuring system |
| US8538738B2 (en) * | 2011-03-22 | 2013-09-17 | Aerion Corporation | Predicting transition from laminar to turbulent flow over a surface |
-
2023
- 2023-11-27 CN CN202311585610.8A patent/CN117288283B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101788692A (en) * | 2009-11-24 | 2010-07-28 | 西北电网有限公司 | Short-time wind speed forecasting method based on neural network |
| CN103698553A (en) * | 2013-12-26 | 2014-04-02 | 天津大学 | Novel surface flow field velocity measurement system and velocity measurement method on basis of infrared image pickup |
| DE102020134966A1 (en) * | 2020-12-28 | 2022-06-30 | Okeanos Consulting GbR (vertretungsberechtigte Gesellschafter: Dr. Henning Oppel, 44799 Bochum; Dr. Benjamin Mewes, 44795 Bochum) | Metrological arrangement for evaluating water level measurement data for non-invasive water level-discharge quantity correlation in running waters, as well as method and use |
| CN113155107A (en) * | 2021-05-25 | 2021-07-23 | 华北水利水电大学 | Irregular river channel section flow measuring device and method |
| CN114397476A (en) * | 2021-11-15 | 2022-04-26 | 河海大学 | Flow velocity effectiveness identification and correction method for frequency domain space-time image velocity measurement |
| CN115600044A (en) * | 2022-11-28 | 2023-01-13 | 湖南大学（Cn） | River section flow calculation method, device, equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| Chengpeng Wan ; Shilong Guo ; Jinfen Zhang ; Liang Fan ; Yanyi Chen ; Ziqiang Li.Identification of accident black spots in inland rivers based on DBSCAN with adaptive parameters determination method.《2023 7th International Conference on Transportation Information and Safety (ICTIS)》.2023,574-582. * |
| 甘肃北山钻孔地下水流速流向测定及地下水流数值模拟;李仁海;《中国优秀硕士学位论文全文数据库 (基础科学辑)》;A011-557 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117288283A (en) | 2023-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113124941B (en) | Non-contact type river channel flow measuring and accurate calculating method | |
| CN112711899B (en) | Fusion prediction method for height of evaporation waveguide | |
| CN115311828B (en) | Intelligent underwater engineering structure detection method and system based on digital twins | |
| CN105651265B (en) | The element of wave and tidal level measuring method in the Oversea bridge marine site based on wave pressure | |
| CN109543356A (en) | Consider the ocean interior temperature-salinity structure remote sensing inversion method of Space atmosphere | |
| Vortmeyer-Kley et al. | Detecting and tracking eddies in oceanic flow fields: a Lagrangian descriptor based on the modulus of vorticity | |
| CN114692895A (en) | Method and device for extracting water flow form perception and response relation of fish | |
| CN116296501A (en) | Method for evaluating running of fish passing facility | |
| Chen et al. | Estimating the flow velocity and discharge of ADCP unmeasured area in tidal reach | |
| KR102561270B1 (en) | Method and apparatus for estimating the position of the sea drift object for search and rescue | |
| CN117288283B (en) | River flow rate monitoring method and system based on video | |
| CN112508358A (en) | Engineering cost management system based on big data analysis | |
| CN115761520A (en) | River runoff analysis method and system based on autonomous learning of river floaters | |
| Meng et al. | Wind-sea and swell separation of 1D wave spectrum by deep learning | |
| Habersack et al. | Spatio-temporal variability of bedload transport | |
| CN108536910B (en) | Research method for hydraulic conditions of large fishes | |
| CN115422782A (en) | Flood forecasting system based on hydrological model | |
| KR102238888B1 (en) | Rainfall measuring method using artificial intelligence | |
| CN114529815A (en) | Deep learning-based traffic detection method, device, medium and terminal | |
| Khanarmuei et al. | Calibration and assimilation in hydrodynamic model of a micro-tidal estuary and comparison with Lagrangian drifter data | |
| CN111101477A (en) | Method for determining low water flow during supplement testing of data-free design basin | |
| CN116296266B (en) | Artificial fish reef structure optimization design method | |
| CN118464131B (en) | Ice-stage flow measurement method, system, electronic equipment and storage medium | |
| CN118396384B (en) | Mountain collapse deposit calculation method and system based on machine learning | |
| Johnson et al. | Remote monitoring of volumetric discharge based on surface mean and turbulent metrics |
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 |