CN119850889A

CN119850889A - Ponding detection and early warning method based on video or sequence images

Info

Publication number: CN119850889A
Application number: CN202510062285.XA
Authority: CN
Inventors: 梁明; 王学志
Original assignee: Anhui University
Current assignee: Anhui University
Priority date: 2025-01-15
Filing date: 2025-01-15
Publication date: 2025-04-18

Abstract

The present invention discloses a waterlogging detection and early warning method based on video or sequence images. Aiming at the limitations of traditional waterlogging detection technology, the present invention uses popular devices such as smart phones and driving recorders to shoot images of waterlogged areas, convert formats, preprocess images, crop waterlogged areas, and crop elevation models. Under no-waterlogging conditions, relevant steps are repeated to obtain comparative data, and finally the waterlogging depth is calculated and an early warning system is established. The present method uses popular devices to reduce costs, is applicable to a variety of environmental terrains, can monitor in real time, has flexible data collection, convenient processing and analysis, high visualization, supports sustainable monitoring, is simple to operate, and has the potential to be integrated with other technologies. It can effectively improve the efficiency and accuracy of waterlogging detection and early warning, and provide strong support for disaster prevention and mitigation.

Description

Ponding detection and early warning method based on video or sequence images

Technical Field

The invention relates to the technical field of ponding detection and early warning, in particular to a ponding detection and early warning method based on video or sequence images.

Background

The traditional road ponding detection method, such as a synthetic aperture radar and the like, has certain limitations, and is mainly characterized in the following aspects:

weather dependence SAR performance may be affected by cloud cover and other atmospheric conditions, especially during rainfall.

Geometric distortion-SAR images may be affected by terrain, resulting in geometric distortion on the image, especially on complex terrain.

In some areas, SAR data may be acquired less frequently, limiting the ability to monitor dynamic phenomena.

Cost and complexity SAR systems are relatively expensive to deploy and maintain, and data processing is relatively complex.

In addition, the performance of the conventional visible light imaging technology in severe weather conditions such as rain clouds can be affected.

Illumination conditions-visible light imaging relies on natural or artificial illumination, performance is limited in low-light or night conditions, and different illumination conditions and camera settings may cause deviations in image color, especially in special areas under bridges and inside tunnels, and the like, are limited in application.

Weather effects such as rain, fog, smoke, etc., can reduce the sharpness and contrast of the image.

Real-time is lacking in that in some cases, the visible light imaging system may not be able to achieve real-time or near real-time data acquisition and processing.

Therefore, how to provide a method for detecting and pre-warning accumulated water based on video or serial images is a problem that needs to be solved by those skilled in the art.

Disclosure of Invention

The invention aims to provide a ponding detection and early warning method based on video or sequence images, which can utilize video or sequence images recorded by widely popularized equipment such as a smart phone, a vehicle recorder and the like to carry out ponding detection so as to realize more efficient and real-time monitoring and early warning.

According to the embodiment of the invention, the ponding detection and early warning method based on video or sequence images comprises the following steps:

S1, shooting a sequence image or video around a ponding area by using a mobile phone camera;

S2, converting the shot photos or videos into a universal JPG format;

S3, aligning the photo, adding control points, generating point cloud, constructing a digital elevation model and generating an orthographic image by utilizing image processing software;

S4, drawing out a water accumulation area on an orthographic image of the water accumulation area by a visual discrimination method, and constructing a vector image layer of the water accumulation range;

S5, cutting out a digital elevation model of the water accumulation area by using geographic information processing software to obtain elevation data of the water accumulation area;

S6, repeating the steps of collecting and processing under the condition of no ponding in sunny days of the same position and area, and obtaining relevant data without ponding;

And S7, importing elevation data under the conditions of water accumulation and no water accumulation into geographic information processing software, and performing comparative analysis to calculate the water accumulation depth so as to realize water accumulation detection and early warning.

Further, the specific method for shooting the image and the video of the water accumulation area in the step S1 is as follows:

Shooting around the water accumulation area by using a mobile phone camera for one week, capturing a detailed view of the water collection area, shooting the video around the water accumulation area for one week when recording the video, and extracting key frames from the video as photos by using professional software;

When shooting is carried out in a ponding area, the grid line function of a camera is started to serve as visual assistance to assist in maintaining the orderly arrangement and consistency of the photos;

after shooting is completed, the photos are arranged and stored in a folder, and preparation is made for subsequent image processing and analysis work.

The ponding area has specific geographic position, can be urban streets, low-lying zones, farmlands or other areas easy to ponding, ponding can be temporary, such as short ponding after heavy rain, can also be continuous, such as ponding caused by long-term high groundwater level, preliminary judgment is carried out by observing whether obvious water accumulation exists on the ground surface, different colors can be displayed on photos of the ponding area, the photos are compared with surrounding dry ground, a mobile phone camera is used for shooting around the ponding area for a week, a detailed scene of a water collecting area is captured, when shooting of the ponding area is carried out, the function of grid lines of the camera is started, the functions are used as visual assistance, the tidy arrangement and consistency of photos are maintained, meanwhile, the camera is ensured to be kept stable in the shooting process, the unified shooting view angle is adopted, severe shaking or rotation is avoided, errors possibly occurring in the image splicing process can be reduced, and the smoothness and naturalness of the final splicing effect can be improved.

Further, the specific steps of the picture processing in S3 are as follows:

S31, primarily aligning photos, starting image processing software, importing pictures of a water accumulation area from equipment, carrying coordinate data on each picture, performing photo alignment operation through an alignment photo function in the software after importing the pictures, and determining the accurate position and the direction of each picture in a three-dimensional space by the alignment photo, so as to provide a foundation for generating a seamless spliced high-quality orthophoto, a three-dimensional model and topographic data;

When the photos are initially aligned, the image processing software is started to import the pictures of the ponding area, each picture is provided with coordinate data, when the photos are imported and aligned, high precision is selected in the popup window, and finally, click determining software intelligently completes automatic sequencing and calibration of the photos according to the coordinates, the elevations and the similarity of the photos, so that accurate alignment and seamless splicing of all the images are ensured.

S32, adding control points, using a manual selection tool in image processing software, selecting three remarkable and easily-identified characteristic points on the photos as the control points, wherein the points have uniqueness and identifiability in different photos, after the control points are added, if some control points in the photos are found to be not aligned correctly, the positions of the control points can be manually adjusted until a satisfactory effect is achieved, after the control points are added and the accuracy of the control points is checked, the functions of an optimized camera are operated, the image alignment is optimized, finally, the photo alignment operation is carried out again, the high-quality output of the final image is ensured, and the accuracy and the efficiency of image splicing are remarkably improved;

Characteristic points such as curbs, building edges or other remarkable natural or artificial structures so as to ensure accurate alignment and seamless splicing of images, control points are added uniformly in each photo as much as possible, and concentration in a certain area is avoided, so that uniform alignment of images in the splicing process can be ensured, and each control point is marked in at least two photos.

S33, generating a point cloud, namely generating the point cloud through the function of generating the point cloud on image processing software, carefully checking the quality of the point cloud after the generation is completed, ensuring no obvious defect errors, returning and adjusting settings to improve the result if required, and storing the point cloud into a common point cloud format after the point cloud accords with expectations;

through the operation flow, the three-dimensional point cloud data of the water collecting area can be successfully captured, the data not only enriches the geographic information of the area, but also provides a solid foundation for further analysis and application.

S34, generating a digital elevation model, starting a function option for constructing the digital elevation model in software, carefully selecting a coordinate system and source data suitable for the project, wherein a data source is point cloud data generated in the previous step, ensuring that the selected data is consistent with geospatial data, and constructing the digital elevation model after confirming that all settings are correct;

The selected data and the geospatial data are ensured to be consistent, so that the spatial accuracy of the generated digital elevation model is ensured, the generated digital elevation model is checked after the generation is finished, the elevation information of the terrain is ensured to be accurately reflected, the step is a key link for ensuring the data quality, if the generated digital elevation model meets the requirements, the generated digital elevation model can be exported into a TIFF format, and therefore, the data can be used in Geographic Information System (GIS) software or related analysis tools for subsequent operation.

S35, generating an orthographic image, selecting and constructing an orthographic image function option in software, configuring parameters such as resolution, quality and projection of the orthographic image according to requirements, starting a generation process of the orthographic image after all the parameters are configured, checking the quality of the orthographic image in image processing software after the generation is finished, ensuring no obvious distortion and error, manually adjusting the range of the orthographic image, optimizing the color and the contrast, improving the visual effect of a final image, and exporting the orthographic image into a common image format after confirming that the orthographic image meets the quality requirements.

Once the orthographic image is confirmed to meet the quality requirement, it can be exported into a commonly used image format, such as TIFF format, and the generated orthographic image can be further analyzed in the geographic information processing software.

Further, the specific step of drawing out the ponding area and constructing the vector layer of the ponding range in the step S4 is as follows:

S41, cutting a range of a water accumulation area, outlining a water collection area on an orthographic image of the water accumulation area by a visual discrimination method, specifically, adding the orthographic image of the water accumulation area into a map view by using an adding data function in geographic information processing software, selecting an orthographic image layer, and then viewing and recording coordinate system information of the image in a coordinate system tab;

S42, creating new vector elements, namely, in a pop-up dialog box, naming new shape as a 'water accumulation area range', selecting 'face elements' as element types, ensuring that a space reference coordinate system identical to an orthophoto is selected when the shape is created, selecting a newly created shape layer for editing in processing software, drawing face elements along the boundary of the water accumulation area, and determining the vertexes of a polygon by using a mouse click according to the characteristics displayed on the orthophoto in a map view to finish polygon drawing;

S43, after drawing and editing of the surface elements are completed, an editing file is saved.

The water accumulation can be temporary, such as short water accumulation after heavy rain, or continuous, such as water accumulation caused by long-term high groundwater level, and preliminary judgment is performed by observing whether obvious water accumulation exists on the ground surface, and the photos of the water accumulation areas can be in different colors, which are compared with the surrounding dry ground.

Further, the step of cutting out the digital elevation model of the ponding region and obtaining the elevation data of the ponding region in the step S5 comprises the following specific steps:

S51, importing a water accumulation area range and a digital elevation model of the water accumulation area into a map view, opening geographic information processing software, selecting a digital elevation model layer in a mask extraction tool as an input grid, selecting vector surface elements of the water accumulation area as mask data, then executing cutting operation, and automatically extracting corresponding elevation data from the digital elevation model according to a boundary defined by the water accumulation area by the geographic information processing software to generate a new cut digital elevation model layer;

The digital elevation model of the ponding area range and the ponding area is imported into a map view by using the data adding function of the geographic information processing software, and the coordinate system of the digital elevation model is confirmed to be completely consistent with the coordinate system of the vector surface element of the ponding area, so that the spatial alignment of the data is ensured.

S52, storing the cut digital elevation model layer as a new data set, and obtaining the elevation difference when water is accumulated in the new data set.

Basic data is provided for the measurement and analysis of the depth of the accumulated water.

Further, the specific step of obtaining the relevant data of no ponding at the same position and area in S6 includes:

S61, performing surrounding shooting in the same position and area under the weather condition of clear weather without ponding, and acquiring photos or video data;

s62, applying a mask-based extraction tool in geographic information processing software, cutting out a digital elevation model of an accurate area from the digital elevation model by taking the ponding range obtained in the step S5 as a mask, setting output parameters, executing cutting operation, and generating the digital elevation model under the ponding-free condition.

In the same position and area, under the weather condition of clear water, the surrounding shooting is carried out, the photo or video data is obtained, the photo is utilized to generate point cloud data, a digital elevation model and an orthophoto image without water in image processing software, then, in a space analysis tool box of geographic information processing software, a tool which is extracted according to a mask is selected, the water accumulation range obtained in the previous step is used as a mask, the digital elevation model of an accurate area is cut out from a DEM, output parameters are set, cutting operation is carried out, and the digital elevation model under the water accumulation condition is generated.

Further, in the step S7, the specific method for calculating the water accumulation depth by contrast analysis and realizing water accumulation detection and early warning is as follows:

The elevation data of the water accumulation area obtained by cutting and the elevation data under the condition of no water accumulation on a sunny day are imported into geographic information processing software for deep comparison analysis, so that the surface elevation difference between the water accumulation time and the water accumulation time can be quantized respectively, namely the water accumulation depth of the water accumulation area is reflected, and an early warning system of the water accumulation depth is built based on the data, so that preventive measures are taken before the occurrence of potential disasters.

In the analysis process, the difference of the earth surface elevation difference when water is accumulated and the earth surface elevation difference when water is not accumulated can be respectively quantified, namely the water accumulation depth of a water accumulation area is reflected, and the calculation method not only provides a means for measuring the water accumulation depth, but also enables a warning system of the water accumulation depth to be established based on the data, so that preventive measures are taken before potential disasters occur.

Further, specific parameters of the sequence image and video shot in the step S1 are as follows:

the resolution of the photo is 3024 multiplied by 4032, the specific camera parameters are ISO speed 50, wide-angle camera 26mm, aperture f1.8, and each photo has GPS coordinates and comprises longitude and latitude information;

The video resolution is 1080 x 1920, including the specific date and time of the video capture, and also including the geographic coordinates of the video capture, including latitude and longitude, and the number of frames per second is 240.37.

Further, the water accumulation area in S1 includes an area where surface water is accumulated temporarily or continuously due to rainfall, rising groundwater, unsmooth drainage or other reasons.

The beneficial effects of the invention are as follows:

1. According to the invention, the equipment such as the widely popularized smart phone and the automobile data recorder is utilized for water accumulation detection, expensive professional detection equipment is not required to be purchased, the equipment can be easily obtained in daily life by common people, the equipment threshold and the economic cost of water accumulation detection are reduced, and the water accumulation detection work can be carried out by utilizing the existing mobile phone without additionally inputting a large amount of special and expensive equipment such as a synthetic aperture radar and the like, so that the method has remarkable economic advantages for small communities, remote areas or individual users with limited resources.

2. The method is suitable for various complex environments and terrains, including areas under the bridge, inside the tunnel and the like where the traditional ponding detection method is difficult to reach or has poor effect, and under the bridge, the traditional detection means can be limited due to the special structure and shielding, but by using the method, a user can conveniently shoot at the peripheral position by using a mobile phone to acquire ponding data, and in the tunnel, image recording can be performed by using handheld equipment, so that the limitations of traditional detection equipment such as insufficient light, narrow space and the like are overcome, and the method provides possibility for comprehensively monitoring various ponding areas.

3. The invention can capture ponding situation in real time during rainfall, provide timely and accurate information for relevant departments and personnel, and in the process of heavy rain, citizens can rapidly shoot images of ponding areas by using mobile phones, and can obtain the information of ponding depth, range and the like in a short time through rapid data processing, so that urban management departments can timely grasp road ponding situation, rapidly make decisions such as traffic guiding, drainage facility scheduling and the like, improve the rapid response capability of cities to flood disasters, and reduce the influence of ponding on traffic and citizens.

4. According to the invention, different shooting angles and ranges can be freely selected according to actual needs to obtain more comprehensive and detailed ponding information, when the street ponding is shot, shooting can be carried out from different directions and different heights, the state of the ponding under different visual angles can be captured, the distribution and depth change condition of the ponding can be reflected more accurately, the flexibility enables the monitoring of complex ponding areas to be more accurate, and the diversified requirements of ponding detection under different scenes can be met.

5. The data such as the orthographic image, the DEM, the point cloud and the like generated by the method are easy to analyze in geographic information processing software, and the data formats are standard formats in the field of geographic information, can be conveniently integrated and compared with other geographic data for analysis, are convenient for deep mining of geographic features and change rules of a ponding area, and provide more scientific basis for ponding monitoring and early warning.

6. The invention can monitor the ponding area regularly, shoot and analyze the easy ponding road section and area regularly or irregularly in rainy season, track the change condition of the ponding area along with time, including expansion or shrinkage of ponding range, increase or decrease of depth, etc., and can deeply understand the dynamic change rule of the ponding area through long-term data accumulation and comparison, thereby providing data support for urban planning, drainage system optimization, etc.

Drawings

The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:

FIG. 1 is a block flow diagram of a method for detecting and pre-warning water accumulation based on video or serial images according to the present invention;

FIG. 2 is an orthographic image of a study area with water accumulation;

FIG. 3 is an orthographic image of a study area without water accumulation;

FIG. 4 is a digital elevation model diagram of a study area prior to cropping when there is water accumulation;

FIG. 5 is a digital elevation model of the study area without water accumulation prior to cropping;

FIG. 6 is a digital elevation model diagram of the range of water accumulation after clipping;

FIG. 7 is a graph of a digital elevation model without water accumulation within the range of water accumulation after clipping.

Detailed Description

The invention will now be described in further detail with reference to the accompanying drawings. The drawings are simplified schematic representations which merely illustrate the basic structure of the invention and therefore show only the structures which are relevant to the invention.

Referring to fig. 1-7, a method for detecting and pre-warning accumulated water based on video or sequence images is characterized by comprising the following steps:

s1, data acquisition, namely shooting sequence images or videos around a ponding area by using a mobile phone camera;

and proper smart phones or automobile data recorders are selected, so that sufficient electric quantity and enough storage space of equipment are ensured, and the camera functions normally.

The method comprises the steps of checking and setting camera parameters, setting resolution ratio of 3024×4032, ISO speed of 50, focal length of a wide-angle camera of 26mm, aperture f1.8 when shooting video, recording shooting date, time and geographic coordinates, and number of frames per second 240.37 when shooting video, and starting a GPS positioning function of the device, so that the shot image can be attached with longitude and latitude information.

It should be noted that, when the water-logging area is reached, the shooting route and range are planned according to the size, shape and surrounding environment of the water-logging area, so that the shooting range can cover the water-logging area and surrounding areas which may affect the water-logging condition, such as a water outlet, adjacent areas with higher or lower topography, and the like.

S2, converting the photographed photo or video into a universal JPG format;

And installing and starting professional photo format conversion software, such as FormatFactory, xnConvert, and introducing all photos obtained by shooting into the conversion software in batches, selecting to convert the photo format into a JPG format in the software, setting conversion parameters, such as maintaining the resolution of an original image, selecting medium quality compression and the like, so as to improve the subsequent processing efficiency.

It should be noted that, the conversion process is started, after the conversion is completed, the converted JPG format photos are sorted and stored in a designated folder, and preparation is made for subsequent image processing and analysis work.

S3, image processing, namely aligning a photo, adding control points, generating point cloud, constructing a digital elevation model and generating an orthographic image by utilizing image processing software;

Starting image processing software such as AgisoftMetashape, pix, dmapper and the like with functions of image alignment, point cloud generation and the like, selecting a photo adding option by clicking a workflow in a toolbar on a software operation interface, and fully importing JPG format photos of a water accumulation area into the software, wherein in the importing process, attention is paid to checking whether the photos are fully imported successfully or not and whether the sequence and arrangement of the photos accord with expectations or not;

Image alignment:

after the photo is imported, clicking a 'workflow' menu again, selecting a 'photo alignment' function, and selecting a high-precision alignment option in a pop-up alignment setting popup window to ensure that software can intelligently complete automatic sequencing and calibration of the photos according to the coordinates, the elevations and the similarity of each photo.

Starting alignment operation, automatically analyzing photo data by software, calculating the position and direction of each photo in a three-dimensional space, wherein the process may need a certain time, the specific time length depends on factors such as the number of the photos, the computer performance and the like, after the alignment operation is completed, checking an alignment result to ensure that all images are aligned accurately and spliced seamlessly, checking whether obvious dislocation, overlapping or gaps exist at the spliced position of the images through a preview function provided by the software, and if so, further checking the photo data or adjusting alignment parameters.

Control point addition

By using a manual selection tool in image processing software, three remarkable and easily-identified characteristic points are carefully selected on the photos as control points, the characteristic points should be selected to have unique and identifiable positions in different photos, such as specific marks of road edges, corners of buildings, bottoms of telegraph poles and other natural or artificial structures, the control points are added in each photo uniformly as much as possible, the concentration in a certain area is avoided, and enough control points are ensured in the whole shooting range to ensure the accuracy of image stitching.

In the process of adding control points, if the positions of some control points in different photos are difficult to accurately match or align, the positions of the control points can be manually adjusted until the corresponding positions of the control points in the photos are accurate, at the moment, the symbol color of the control points is changed from blue to green to indicate successful adjustment, and after the control points are added, a camera optimizing function in software is operated, and the function can perform optimization adjustment on camera parameters according to the position information of the control points, so that the accuracy of image alignment is further improved.

Finally, photo alignment operation is carried out again, the highest precision setting is selected, final optimization is carried out on an image splicing result, high-quality output of the generated image is guaranteed, accuracy and efficiency of image splicing are remarkably improved, and a solid foundation is laid for the follow-up generation of an accurate three-dimensional model, a Digital Elevation Model (DEM) and an orthophoto.

Point cloud generation

Clicking on the workflow in the toolbar on the operation interface of the image processing software selects the option of creating point cloud.

In the pop-up point cloud generation setting dialog box, an appropriate point cloud density is selected according to the complexity degree of the ponding area and the required precision requirement, and generally, for the complex ponding area or a scene needing high-precision analysis, a higher point cloud density can be selected, but attention is paid to the higher density, so that the processing time and the computing resource consumption can be correspondingly increased, and in the step, the point cloud precision is set to be high so as to ensure that the generated point cloud data can accurately reflect the topography and the feature of the ponding area.

After confirming that all the settings are correct, clicking a 'determination' button starts a point cloud generation process, in the point cloud generation process, software calculates and generates three-dimensional point cloud data through a complex algorithm according to image data and previous processing results, the process possibly needs a long time, and in a waiting process, progress information displayed by a software interface can be observed to know the progress condition of point cloud generation according to the number of photos, the density of the selected point cloud and the processing capacity of a computer.

After the point cloud is generated, carefully checking the quality of the point cloud, observing the point cloud data from different angles and different scaling ratios through a point cloud checking tool provided by software, checking whether obvious defects or errors exist, such as uneven point cloud sparseness, partial deletion, excessive noise and the like, if the problems are found, returning to the previous steps according to specific conditions for adjustment and improvement, and if the quality of the point cloud accords with expectations, storing the point cloud into a common point cloud format such as PLY or LAS format so as to facilitate subsequent further analysis and application, wherein the three-dimensional point cloud data of a successfully captured water collecting area provides rich basic data for water accumulation depth calculation and other geographic information analysis.

DEM generation

The "workflow" menu is found in the interface of the image processing software, and the "build, DEM" (BuildDEM) option is clicked on, which is a key step in starting the Digital Elevation Model (DEM) generation process.

In the pop-up setting window, the coordinate system suitable for the current ponding detection project is carefully selected, the coordinate system is selected to ensure consistency with the actual geographic space data, the coordinate system can be generally selected according to geographic information standards of the region or with reference to the existing geographic data coordinate system, and meanwhile, the point cloud data generated in the last step are specifically selected in the data source option to serve as basic data for constructing the DEM, so that the accuracy and the integrity of the selected data are ensured.

After confirming that all the settings are correct, starting the construction process of the DEM, wherein the process involves a large amount of data calculation, the required time may be long, the specific duration depends on the scale of the point cloud data and the processing capacity of the computer, and in the construction process, progress prompt information can be observed through a software interface to know the progress condition of DEM generation.

After the DEM is generated, the DEM must be strictly checked to ensure that the DEM accurately reflects the elevation information of the topography of the water accumulation area, an elevation checking tool provided by software is utilized to check whether the elevation value of each point in the DEM is reasonable or not, whether an obvious abnormal value or an error area exists or not, the step is a key link for ensuring the data quality, if the generated DEM is found to have a problem, the reasons should be analyzed in time and corresponding measures should be taken for correction, such as rechecking cloud data, adjusting a coordinate system to set or optimizing DEM construction parameters, and the like, if the DEM meets the requirements, the DEM is exported into a common geographic information data format so as to carry out subsequent operations in geographic information processing software or related analysis tools, such as water accumulation depth calculation, topography analysis, and the like.

S35, orthophoto image generation

Entering the "workflow" menu of image processing software, find and click on the "build orthophoto" (BuildOrthomosaic) option, which pops up the setup window.

For the resolution ratio, the size of the ponding area, the precision required by analysis and the equipment performance need to be comprehensively considered, if the ponding area is smaller and the details need to be accurately checked, the resolution ratio can be set to be higher, and if the area is larger and the whole ponding condition is focused, the resolution ratio can be properly reduced so as to balance the data processing efficiency and the image definition.

In terms of quality, if the method is mainly used for qualitatively analyzing the water accumulation range, medium quality can be selected, and if accurate water accumulation depth estimation or comparison with other high-precision data is required, high quality is selected, and the high quality can reduce image noise and improve color fidelity, but can increase processing time.

The projection parameters are set according to the geographical position of the ponding area and project requirements, and the normal projection mode is selected to be suitable for local map projection, so that the normal projection image is ensured to be accurately matched with the actual geographical space position, and meanwhile, the correct parameters such as central meridian and the like are required to be set, so that the direction and the proportion of the image are ensured to be correct.

If an orthographic image is found to be problematic, the image range is manually adjusted to optimize color and contrast to enhance visual effects, for example, if an area of the image is darker, the brightness and contrast may be increased, and if there is local distortion, the relevant data and processing steps are examined to attempt regeneration or local correction.

After confirming that the orthographic image meets the quality requirement, the orthographic image is exported into a common image format, such as a TIFF format, and the generated orthographic image can be used for further geographic information analysis.

S4, cutting the range of the water accumulation area, namely outlining the water accumulation area on an orthographic image of the water accumulation area by a visual discrimination method, and constructing a vector image layer of the water accumulation area;

And opening geographic information processing software, adding the orthographic image of the ponding region into the map view by using the data adding function, selecting an orthographic image layer, and viewing and recording the coordinate system information of the image in the coordinate system tab.

Newly created vector elements, named "water area range" for new Shapefile in pop-up dialog, select "face elements" as element types and ensure that the same spatial reference frame as the orthophoto is selected.

Selecting a newly built Shapefile layer, clicking a right button to select 'start editing', selecting a 'drawing' tool, clicking a mouse to determine polygon vertex drawing surface elements according to boundary characteristics of a ponding area on an orthographic image, clicking a closed graph or clicking a drawing completion button after polygon drawing is completed, and moving or adjusting vertices by using an editing vertex function if the drawing surface element boundary is required to be adjusted, so as to ensure that the actual range of the ponding area is accurately reflected.

After the drawing and editing of the face element are finished, clicking a save editing button in an editing toolbar to save, clicking in the editing toolbar to stop editing, selecting to save or discard the change, clicking after confirmation to finish the editing process, and if necessary, exporting the face element into an independent data file by using a data export function so as to further apply the range of the water accumulation area and other geographic elements in a superposition analysis and the like.

the method is characterized in that the 'data adding' function of geographic information processing software is utilized, a vector surface element of a 'water accumulation area range' and an elevation model DEM of the water accumulation area which are created before are imported into a map view, before the DEM is added, the coordinate system of the DEM is required to be confirmed to be completely consistent with that of the vector surface element, the spatial alignment of data is ensured, and the DEM and the vector surface element are displayed in a superimposed mode, so that subsequent analysis and operation are facilitated.

Opening a space analysis toolbox of geographic information processing software, finding and opening a mask-based extraction tool in an extraction analysis category, selecting a DEM layer as an input grid and vector surface elements of a water accumulation area as mask data in a mask-based extraction tool dialog box, designating a storage position and a file name of a cutting result, and preparing to output cut DEM data which only contains elevation information of the water accumulation area.

Clicking to determine and execute cutting operation, the geographic information processing software automatically extracts corresponding elevation data from the DEM according to the boundary defined by the ponding range, generates a new cut DEM layer, finally saves the cut DEM layer as a new data set, obtains the elevation difference of the ponding area when ponding exists, and provides basic data for subsequent ponding depth calculation.

S6, data acquisition under the condition of no ponding is carried out, namely, the acquisition and processing steps are repeated under the condition of no ponding on a sunny day in the same position and area, and relevant data without ponding are obtained;

And under the weather condition of clear water accumulation, repeating the data acquisition steps at the same position and area, taking a photo or video data, wherein the photo resolution, the camera parameters, the video parameters and the like are set to be the same as those of water accumulation, and the acquired image data without water accumulation is ensured to be comparable with the data in water accumulation.

With the photos, point cloud data, a Digital Elevation Model (DEM) and an orthophoto image without water are generated in the image processing software according to the previous steps, and the operations of image alignment, control point addition, point cloud generation, DEM construction, orthophoto image generation and the like are included, so that the accuracy and consistency of data processing are ensured.

In the geographic information processing software, a mask extraction tool is applied, the water accumulation region range determined during water accumulation is used as a mask, an elevation model of an accurate region is cut out from a DEM generated during water accumulation, output parameters are set, cutting operation is performed, and an elevation model under the water accumulation-free condition is generated, and can accurately reflect the surface morphology of the water accumulation region during water accumulation-free condition, so that important contrast data is provided for water accumulation depth calculation.

And S7, analyzing the depth of the accumulated water, namely importing elevation data under the conditions of accumulated water and no accumulated water into geographic information processing software, and performing comparative analysis to calculate the depth of the accumulated water so as to realize accumulated water detection and early warning.

The elevation data obtained by cutting when the water accumulation area has water accumulation and the elevation data under the condition of no water accumulation on a sunny day are imported into geographic information processing software or a professional data analysis tool together, so that the two groups of data are ensured to accurately correspond to each other in space position, and can be verified by checking a coordinate system, a data range and the like.

And performing contrast analysis operation in software, and respectively quantifying the ground surface elevation difference of each corresponding position in the water accumulation area with water accumulation and without water accumulation by utilizing functions of grid calculation, space analysis and the like, for example, subtracting the elevation grid data without water accumulation from the elevation grid data with water accumulation by using a difference calculation tool in the software to obtain grid data of water accumulation depth, wherein the value of each grid unit represents the water accumulation depth of the position.

The calculated water accumulation depth data are arranged and statistically analyzed, statistical indexes such as the maximum value, the minimum value, the average value, the standard deviation and the like of the water accumulation depth are obtained, the distribution condition and the overall characteristic of the water accumulation depth in the water accumulation area are comprehensively known, and a basis is provided for water accumulation condition assessment and early warning.

And establishing a ponding depth early warning system according to the ponding depth calculation result and the actual demand. Different ponding depth thresholds are set, and are divided into a plurality of early warning grades, such as mild ponding (depth is smaller than 10 cm), moderate ponding (depth is 10-30 cm), severe ponding (depth is larger than 30 cm) and the like, and early warning rules and triggering conditions are configured in geographic information processing software or related monitoring platforms.

When the water accumulation depth reaches or exceeds a certain preset early warning level threshold, the system automatically sends out early warning signals of corresponding levels, the early warning signals comprise voice prompts, popup window display, short message notification and the like, and the early warning forms can be set in a self-defined mode according to user requirements, so that relevant personnel can be ensured to acquire water accumulation information in time.

The method comprises the steps of establishing an early warning information release channel, integrating with an urban emergency command system, setting an electronic display screen to release information in public places, pushing by using a social media platform and the like, ensuring that early warning information can be timely and accurately transmitted to groups possibly affected by ponding, such as related departments, surrounding residents, passers-by and the like, so that effective preventive measures, such as drainage facility scheduling, traffic dredging, personnel evacuation and the like, are taken, adverse effects of ponding on life and production are reduced, the capacity of the urban coping with ponding disasters is improved, meanwhile, the early warning system is tested and maintained regularly, early warning threshold rationality, signal transmission timeliness and accuracy are checked, system parameters and settings are timely adjusted according to actual ponding conditions and feedback comments, and the effective running of the system is ensured all the time.

Embodiment one urban street ponding detection

Data acquisition

Obvious ponding appears after rain in the low-lying road section of a city street. The inspector uses the smart phone to start the function of the grid lines of the camera, and keeps stable posture around the ponding area for shooting. The resolution of the photographed picture is 3024×4032, the camera parameters are ISO speed 50, wide-angle camera 26mm, aperture f1.8, and the picture is provided with GPS coordinates. A video is recorded at the same time, and the resolution is 1080 multiplied by 1920, and comprises shooting time, geographic coordinates and 240.37 frames per second. After the shooting is completed, the photos are converted from HEIC format to JPG format and stored in a designated folder.

Image preprocessing

And starting image processing software and importing the photo. The 'align photo' function is selected in the software, the high precision option is selected, and the software automatically completes the sorting and calibration. And then manually selecting characteristic points such as a curb, the bottom of a street lamp post, a street store entrance and the like as control points, so as to ensure that each control point is marked in at least two photos and uniformly distributed. After adding the control point, the "optimize camera" function is run, and then the photo is realigned with the "highest" precision.

Point cloud generation and processing

Clicking the 'create point cloud' option, selecting a higher point cloud density, setting the precision to be 'high', and starting the generation process. And after the generation, the quality of the check point cloud is ensured, and the point cloud can clearly reflect the accumulated water region and the contours of peripheral objects. The point cloud is then saved as PLY format.

DEM and orthophoto image generation

And selecting a 'DEM construction' in a 'workflow', selecting a coordinate system suitable for local, constructing the DEM for a data source by using the generated point cloud, checking the accuracy of the generated point cloud, and deriving the point cloud into a TIFF format. Then selecting 'construct orthophoto', configuring proper resolution, quality and projection parameters, checking whether distortion exists after generation, manually adjusting and deriving into a TIFF format.

Cutting out range of ponding area

And opening geographic information processing software, importing an orthophoto, and viewing and recording coordinate system information. The vector elements of the 'ponding region range' are newly established, the 'surface elements' are selected, and the consistency with an orthographic image coordinate system is ensured. And starting editing, drawing surface elements by using a polygonal tool according to the ponding boundary characteristics, and storing and editing after the editing is completed.

Elevation model cutting for ponding area

And (5) leading the range of the ponding region and the DEM, confirming the consistency of the coordinate system, and then superposing and displaying. And cutting the DEM by using a mask extraction tool and taking vector surface elements of the ponding region as masks to obtain a ponding region elevation model, and storing the model as a new data set.

Data acquisition and processing under ponding-free condition

And after a plurality of days, when no ponding exists at the same street position, repeating the data acquisition steps to obtain a photo and a video, and also processing to generate point clouds, DEM and orthographic images when no ponding exists, and cutting out a ponding-free elevation model of a ponding area.

Ponding depth calculation and early warning

And importing elevation model data with water accumulation and without water accumulation into geographic information processing software, and calculating the water accumulation depth. Setting a ponding depth early warning threshold, wherein if the ponding depth early warning threshold is 10 cm for mild ponding, 20 cm for medium ponding and 30 cm for heavy ponding. When the calculated ponding depth reaches the corresponding threshold value, the system informs related personnel of the municipal administration management department through popup window display and short message, and simultaneously gives out an acoustic alarm at the traffic command center to remind the timely taking of drainage and traffic guiding measures.

Embodiment two, farmland ponding monitoring

Data acquisition

Large-area ponding occurs in certain farmland due to storm. Farmers use mobile phones to shoot ponding areas according to requirements, and notice is taken in the shooting process to keep the stable and uniform visual angle of the camera. The photo and video parameters are the same as in embodiment one. The format is converted to post-sort the photos.

Image preprocessing and subsequent processing

The image processing software is utilized to perform operations such as alignment, control point addition and the like on the photos, and the control point selects obvious characteristic points such as ridges, telegraph poles, farmland boundary marks and the like. And then sequentially generating a point cloud, a DEM and an orthophoto, and adjusting parameters according to actual conditions in the process to ensure the data quality. And finishing cutting of the range and elevation model of the ponding region in geographic information processing software.

Operating without ponding

And when the accumulated water in the farmland is returned, shooting and processing the data again in the same area, and obtaining the related data without accumulated water.

Water accumulation depth analysis and application

After the water accumulation depth is calculated, the data is provided to the agricultural department. The agricultural department evaluates the disaster situation of the farmland according to the depth and the range of the accumulated water, and makes corresponding drainage and reseeding plans. Meanwhile, a farmland ponding early warning model is established by using ponding depth data, early warning is sent to farmers in advance according to weather forecast and real-time monitoring data before a future rainy season comes, farmland protection measures are reminded to be made, and loss is reduced.

Embodiment III, detecting accumulated water below bridge

Data acquisition

Under an urban bridge, water accumulation often occurs due to unsmooth drainage. After raining, a detector goes to the area, a vehicle recorder is used for shooting a video around the ponding area, and then a key frame is extracted as a photo, and parameters of the photo and the video meet the requirements. And converting the format and then carrying out subsequent processing.

Image preprocessing and model construction

And (3) performing image alignment in the image processing software, and selecting corners, piers and the like of the bridge structure as control points to ensure accurate image splicing. The point cloud, the DEM and the orthographic image are generated, and special attention is paid to selecting a proper coordinate system to accurately reflect complex terrain below the bridge when the DEM is constructed.

Determining ponding area and cutting elevation model

Carefully outlining the range of the ponding area in geographic information processing software, accurately identifying the ponding boundary on an orthographic image by adjusting contrast and other modes due to darker light below the bridge, creating a vector image layer and cutting out an elevation model of the ponding area.

No ponding contrast and depth calculation early warning

Repeating the operation when no ponding exists in a sunny day, and comparing the ponding with an elevation model when no ponding exists to calculate the ponding depth. The early warning system is integrated with the monitoring system of the bridge management department, when the accumulated water depth reaches an early warning value, accumulated water information is automatically displayed on a popup window on a bridge management monitoring screen, and meanwhile, a short message notification is sent to bridge maintenance personnel, so that a drainage pipeline is cleaned in time, and bridge safety is guaranteed.

According to the different embodiments, the ponding detection and early warning method can be effectively implemented in different scenes, and a practical and reliable solution is provided for ponding monitoring and disaster prevention.

The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.

Claims

1. A method for detecting and warning waterlogging based on video or sequence images, characterized in that it comprises the following steps:

S1. Data collection: Use a mobile phone camera to shoot sequential images or videos around the waterlogged area;

S2. Format conversion: convert the captured photos or videos into the common JPG format;

S3, Image processing: Use image processing software to align photos, add control points, generate point clouds, build digital elevation models, and generate orthophotos;

S4. Clipping of waterlogging area: Outline the waterlogging area on the orthophoto of the waterlogging area through visual discrimination method, and construct a vector layer of the waterlogging range;

S5. Digital elevation model clipping: Use geographic information processing software to clip the digital elevation model of the waterlogged area to obtain the elevation data of the waterlogged area;

S6. Data collection under no water accumulation conditions: repeat the above collection and processing steps under no water accumulation conditions on a sunny day at the same location and area to obtain relevant data when no water accumulation occurs;

S7. Analysis of waterlogging depth: Import elevation data with and without waterlogging into geographic information processing software, conduct comparative analysis and calculate waterlogging depth to achieve waterlogging detection and early warning.

2. The method for detecting and warning waterlogging based on video or sequence images according to claim 1, wherein the specific method of shooting images and videos of the waterlogged area in S1 is:

Use a mobile phone camera to shoot around the waterlogged area to capture detailed views of the watershed. When recording a video, shoot around the waterlogged area and use professional software to extract key frames as photos.

When photographing areas with accumulated water, enable the camera's grid function as a visual aid to help maintain neat arrangement and consistency in your photos;

After you finish shooting, organize your photos into folders to prepare for subsequent image processing and analysis.

3. The method for detecting and warning waterlogging based on video or sequence images according to claim 1, wherein the specific steps of image processing in S3 are:

S31. Preliminary alignment of photos. Start the image processing software and import the pictures of the waterlogged area from the device. Each of these pictures has its own coordinate data. After importing these pictures, use the photo alignment function in the software to perform photo alignment. Photo alignment can determine the precise position and direction of each picture in three-dimensional space, providing a basis for generating seamless high-quality orthophotos, three-dimensional models and terrain data.

S32. Add control points. Use the manual selection tool in the image processing software to select three prominent and easily identifiable feature points on the photo as control points. These points should be unique and identifiable in different photos. After adding the control points, if you find that some control points in the photo are not correctly aligned, you can manually adjust their positions until a satisfactory effect is achieved. After adding the control points and checking their accuracy, run the optimize camera function to optimize the image alignment, and finally re-perform the photo alignment operation to ensure high-quality output of the final image and significantly improve the accuracy and efficiency of image stitching.

S33. Generate point cloud. Generate point cloud through the function of generating point cloud in image processing software. After the generation is completed, carefully check the quality of the point cloud to ensure that there are no obvious defects or errors. If necessary, you can go back and adjust the settings to improve the results. After the point cloud meets the expectations, you can save it in a common point cloud format.

S34, generate a digital elevation model. In the software, enable the option of building a digital elevation model, and carefully select the coordinate system and source data suitable for the project. The data source is the point cloud data generated in the previous step. Make sure that the selected data is consistent with the geospatial data. After confirming that all settings are correct, build the digital elevation model.

S35. Generate orthophotos. Select the Build Orthophotos function option in the software and configure the resolution, quality, projection and other parameters of the orthophotos as needed. After configuring all parameters, start the orthophoto generation process. After the generation is completed, check the quality of the orthophotos in the image processing software to ensure that there are no obvious distortions and errors. If necessary, you can manually adjust the range of the orthophotos and optimize the color and contrast to enhance the visual effect of the final image. After confirming that the orthophotos meet the quality requirements, you can export them to common image formats.

4. The method for waterlogging detection and early warning based on video or sequence images according to claim 1, characterized in that the specific steps of outlining the waterlogging area and constructing the vector layer of the waterlogging range in S4 are:

S41. Clip the waterlogged area and outline the water collection area on the orthophoto of the waterlogged area by visual discrimination. Specifically, you can use the add data function in the geographic information processing software to add the orthophoto of the waterlogged area to the map view, select the orthophoto layer, and then view and record the coordinate system information of the image in the coordinate system tab;

S42. Create a new vector feature. In the pop-up dialog box, name the new Shapefile "Waterlogged Area Range" and select "Surface Feature" as the feature type. When creating the Shapefile, make sure to select the same spatial reference coordinate system as the orthophoto. In the processing software, select the newly created Shapefile layer for editing. Draw the surface feature along the boundary of the waterlogged area. In the map view, use the mouse to click to determine the vertices of the polygon according to the features displayed on the orthophoto to complete the polygon drawing.

S43. After completing the drawing and editing of the surface elements, save the edited file.

5. The method for waterlogging detection and early warning based on video or sequence images according to claim 1 is characterized in that the specific steps of cutting out the digital elevation model of the waterlogged area and obtaining the elevation data of the waterlogged area in S5 are:

S51, importing the waterlogged area and the digital elevation model of the waterlogged area into a map view, opening geographic information processing software and selecting the digital elevation model layer as the input raster in the Extract by Mask tool, selecting the vector surface element of the waterlogged area as the mask data and then performing a clipping operation, the geographic information processing software will automatically extract the corresponding elevation data from the digital elevation model according to the boundary defined by the waterlogged area, and generate a new clipped digital elevation model layer;

S52. The cropped digital elevation model layer is saved as a new dataset, from which the elevation difference when water is accumulated is obtained.

6. The method for waterlogging detection and early warning based on video or sequence images according to claim 1, characterized in that the specific steps of obtaining data related to no waterlogging in the same position and area in S6 are:

S61. In the same location and area, in clear weather without water accumulation, a surround shooting was conducted to obtain photos or video materials;

S62. Applying a mask extraction tool in geographic information processing software, using the waterlogging range obtained in step S5 as a mask, clipping the digital elevation model of the precise area from the digital elevation model, setting output parameters, executing a clipping operation, and generating a digital elevation model without waterlogging.

7. The method for waterlogging detection and early warning based on video or sequence images according to claim 1, characterized in that the specific method for comparing, analyzing and calculating the waterlogging depth in S7 to achieve waterlogging detection and early warning is:

The clipped elevation data of the waterlogged areas and the elevation data under sunny conditions without waterlogging are imported into geographic information processing software for in-depth comparative analysis. The difference in surface elevation when there is waterlogging and when there is no waterlogging can be quantified. The difference between these two values reflects the depth of waterlogging in the waterlogged areas. Based on these data, an early warning system for waterlogging depth can be established to take preventive measures before potential disasters occur.

8. The method for detecting and warning waterlogging based on video or sequence images according to claim 2, wherein the specific parameters of the sequence images and videos taken in S1 are:

The resolution of the photos is 3024×4032, and the specific camera parameters are ISO speed 50, wide-angle camera 26mm, aperture f1.8, and each photo comes with GPS coordinates including longitude and latitude information;

The video resolution is 1080×1920, including the specific date and time when the video was shot, as well as the geographical coordinates of the video shooting, including latitude and longitude, and the frame rate per second is 240.37.

9. A waterlogging detection and early warning method based on video or sequence images according to claim 1, characterized in that the waterlogging area in S1 includes an area where temporary or continuous waterlogging occurs on the surface due to rainfall, rising groundwater, poor drainage or other reasons.