KR20230143499A - Raindrop Size Distribution and Rainrate Estimation Method Using CCTV - Google Patents

Abstract

The method for estimating rainfall particle distribution and rainfall intensity using CCTV according to the present invention includes step (a) in which the data collection unit receives original image data captured from CCTV installed in a randomly selected area and stores it in a database, and the image processing unit Step (b) of preprocessing the original image data stored in the database and converting it into processed image data, wherein the image processing unit generates image data for analysis by removing the background area except for the area corresponding to the rain streak from the processed image data. Step c), step (d) in which the rainfall analysis unit calculates the diameter information of each rainfall particle included in the image data for analysis, and step (d) in which the rainfall analysis unit calculates the diameter information of each rainfall particle calculated in step (d). Step (e) of collecting and calculating the number concentration by diameter of rainfall particles, and step (f) of the rainfall analysis unit deriving the final rainfall intensity based on the number concentration by diameter of rainfall particles calculated in step (e). Includes.

Description

Raindrop Size Distribution and Rainrate Estimation Method Using CCTV}

The present invention relates to a method for estimating rainfall particle distribution and rainfall intensity. More specifically, real-time rainfall particle distribution is collected using rainfall particle observation images obtained through CCTV screens, and water concentration values for each rainfall particle diameter section are used. This relates to rainfall particle distribution and rainfall intensity estimation methods that can be used as rainfall correction data for remote sensing observation data when estimating rainfall intensity, and can provide city-scale rainfall information by eliminating rainfall observation blank areas.

Rainfall refers to information measured by the amount of rain using tools such as a rain gauge. Currently, rainfall is measured by the Korea Meteorological Administration and provided as meteorological information. And conventionally, the Korea Meteorological Administration uses equipment such as AWS (Disaster Prevention and Weather Observation Equipment) and ASOS (Synoptic Weather Observation Equipment).

Among these, AWS automatically measures, calculates, stores, and displays weather-related indicators (barometric pressure, temperature, wind direction/speed, rainfall, temperature, wind) observation data at 510 locations across the country, while ASOS automatically measures, calculates, stores, and displays observation data at 102 locations across the country. In addition to weather-related indicators that can be obtained through AWS, indicator values such as visibility, cloud height, and new snow cover are also measured.

However, in the case of certain terrains such as mountainous terrain, there are local differences in rainfall, but the rainfall information provided by the Korea Meteorological Administration using the above equipment does not take into account these differences in rainfall and rainfall data from specific observatories are distributed over a wide area. Because it represents meteorological characteristics, there is a problem that the information is vague and inaccurate. There is also a limitation in that rainfall particle distribution information cannot be obtained.

Meanwhile, a rain drop meter such as PARSIVEL (Parsivel Size and Velocity) measures the laser intensity of the transmitter and receiver when rain particles falling in the atmosphere pass through the two-dimensional optical laser area coming from the transmitter of the optical sensor. It is a device that calculates the reduced time to retrieve the diameter and falling speed of rainfall particles, and can obtain information on not only the amount of rainfall but also the distribution of rainfall particles.

However, such a rain gauge has the problem of being difficult to install in various areas due to expensive equipment (approximately 10 million won), and also causes differences in the distribution of rainfall particles in the final product due to differences in the observation channel sections set for each type of rain gauge. do.

Therefore, a method to solve these problems is required.

Korean Patent No. 10-2057989

The present invention is an invention made to solve the problems of the prior art described above. In addition to reducing operating costs, it is possible to obtain rainfall information even in mountainous and marine areas where it is difficult to install ground observation equipment, and local rainfall information is possible. The purpose is to provide a rainfall particle distribution and rainfall intensity estimation method that can provide.

The problems of the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below.

The method for estimating rainfall particle distribution and rainfall intensity using CCTV of the present invention to achieve the above purpose is to receive original image data captured from CCTV installed within a randomly selected area of the data collection area and store it in a database (a ) Step, step (b) in which the image processing unit pre-processes the original image data stored in the database and converts it into processed image data, for analysis in which the image processing unit removes the remaining background area excluding the area corresponding to the rain streak from the processed image data. Step (c) of generating image data, step (d) of the rainfall analysis unit calculating diameter information of each rainfall particle included in the image data for analysis, and step (d) of the rainfall analysis unit calculating the diameter information of each rainfall particle included in the image data for analysis. Step (e) of collecting the diameter information of rainfall particles and calculating the number concentration by diameter of rainfall particles, and the rainfall analysis unit calculates the final rainfall intensity based on the number concentration by diameter of rainfall particles calculated in step (e). Includes step (f) of deriving.

At this time, in step (b), the image processing unit may convert the original image data into a black and white image with only contrast ratio information remaining.

And the step (c) is a step (c-1) in which the image processing unit specifies a group of pixels corresponding to rain streaks from the processed image data, and the image processing unit excludes a group of pixels corresponding to rain streaks from the processed image data. Step (c-2) of selecting the remaining pixel group, and the image processing unit removes the brightness of the remaining pixel group selected in step (c-2) to generate processed image data in which the pixel group corresponding to the rain streak is separated from the background. It may include step (c-3).

In addition, the step (d) is a step (d-1) in which the rain analysis unit performs convolution processing on the processed image data to separately classify a group of pixels corresponding to rain streaks, and the rain analysis unit performs a convolution process on the processed image data. ) Step (d-2) of calculating the central axis of the individual rain pixel group divided by step, the rain analysis unit calculates the long axis and short axis of the individual rain pixel group based on the central axis calculated in step (d-2). Step (d-3) of deriving and the rainfall analysis unit calculates diameter information of each rainfall particle by converting the short axis of the individual rain streak pixel group derived in step (d-3) into the diameter of the rainfall particle (d-4) ) may include steps.

At this time, step (d-4) may reflect the settings of the CCTV that acquired the original image data in the process of calculating the diameter information of the rainfall particles by the rainfall analysis unit.

And in step (f), the rainfall analysis unit,

The final rainfall intensity can be derived through the formula.

Meanwhile, between the step (e) and the step (f), a step (ex1) in which the rainfall analysis unit calculates an effective volume within the depth of field of the image data for analysis, and the rain analysis unit calculates the effective volume within the depth of field of the image data for analysis. Step (ex2) of removing meteorological data is further performed, and the results of step (ex1) and step (ex2) can be reflected in step (f).

The method for estimating rainfall particle distribution and rainfall intensity using CCTV of the present invention to solve the above problems is limited by the limited observation area of the rain gauge used conventionally through rainfall data acquired through video-based equipment called CCTV. can be reduced, and has the advantage of being able to build a high spatial resolution observation network using low-cost equipment.

In addition, the present invention can partially solve the problem of spatial errors in remote sensing observation data and contribute to increasing the usability of the limited rainfall information of existing ground AWS and ASOS observation data.

In addition, the present invention enables the collection of rainfall information in the direction opposite to the wind (downwind side), such as behind high-rise buildings or mountains, enabling the creation of a high-precision rainfall map at a city scale. Through such high-resolution rainfall map data, more accurate regional information can be generated when heavy rain occurs. By providing rainfall information, it has the advantage of being able to be used as basic data for establishing countermeasures against flooding and landslides.

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the claims.

1 is a diagram showing each process of a method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention;
Figure 2 is a diagram showing each component of a system for performing a method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention;
Figure 3 is a diagram showing the detailed process of step (c) in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention;
Figure 4 is a diagram showing the process of separating and processing a group of pixels corresponding to rain streaks from the background in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention;
Figure 5 is a diagram showing the detailed process of step (d) in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention;
Figure 6 is a diagram showing the process of separately classifying and processing pixel groups corresponding to rain streaks in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention;
Figure 7 is a diagram showing the process of calculating diameter information of rainfall particles in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention; and
Figure 7 is a diagram showing a graph of number concentration by rainfall particle diameter calculated by the rainfall analysis unit in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention, in which the object of the present invention can be realized in detail, will be described with reference to the attached drawings. In describing this embodiment, the same names and the same symbols are used for the same components, and additional description accordingly will be omitted.

The method for estimating rainfall particle distribution and rainfall intensity using CCTV according to the present invention is performed through a management server installed with a program for estimating rainfall particle distribution and rainfall intensity using CCTV stored in a storage medium, and is installed and managed on the management server. It can be driven by the server's processor.

In addition, the program for estimating rainfall particle distribution and rainfall intensity using CCTV driven by this can be output through an image output device such as a display module, and can provide visible information to users through a visualized graphic user interface.

In particular, the storage medium storing the program for estimating rainfall particle distribution and rainfall intensity using CCTV can be installed on the management server using a removable disk or a communication network, and the program for estimating rainfall particle distribution and rainfall intensity using CCTV can be installed on the management server. It can be operated through various functional means.

In other words, in the present invention, information processing by software is concretely realized through hardware.

Below, the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to the present invention will be described in detail.

Figure 1 is a diagram showing each process of the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

Figure 2 is a diagram showing each component of a system for performing a method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention. In the following description, the symbols assigned to each component are It is based on this drawing.

As shown in Figure 1, the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention includes steps (a) to (f).

First, step (a) is a process in which the data collection unit 110 of the management server 100 receives original image data captured from the CCTV 10 installed within a randomly selected area and stores it in the database 111.

Here, the arbitrarily selected area refers to the target area for deriving rainfall intensity. In the case of the Korea Meteorological Administration, which manages the entire country, all original video data captured from CCTVs (10) existing throughout the country are collected and randomly collected. It can be stored separately by region.

At this time, the original image data can be stored in both non-precipitation and precipitation cases, so that it is possible to identify trends in temporal changes in rainfall intensity data.

For reference, according to Statistics Korea's e-Nara Index, the number of CCTVs installed in public institutions in Korea is rapidly increasing, from about 150,000 in 2008 to about 1.34 million in 2020, and the number of CCTVs for private use is over 13 million as of 2017. It is estimated that it is sufficient to derive nationwide rainfall intensity.

Next, step (b) is performed in which the image processing unit 120 of the management server 100 preprocesses the original image data stored in the database 111 and converts it into processed image data.

That is, in this process, the original image data acquired through CCTV is corrected and processed into a form that is easy to analyze by the image processing unit 120 and the rain analysis unit 130.

In this embodiment, in step (b), the image processing unit converts the original image data into a black-and-white image with only contrast ratio information remaining, making it easier to separate each pixel thereafter.

Next, step (c) is performed in which the image processing unit 120 generates image data for analysis by removing the background area excluding the area corresponding to the rain streak from the processed image data.

Figure 3 is a diagram showing the detailed process of step (c) in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

As shown in FIG. 3, step (c) in this embodiment may include steps (c-1) to (c-3) in detail.

In step (c-1), the image processing unit 120 specifies a group of pixels corresponding to rain streaks in the processed image data, and in step (c-2), the image processing unit 120 specifies a group of pixels corresponding to rain streaks in the processed image data. The process of selecting the remaining pixel groups excluding the pixel group is performed.

And in step (c-3), the image processing unit 120 removes the brightness of the remaining pixel group selected in step (c-2) to generate processed image data in which the pixel group corresponding to the rain streak is separated from the background.

Figure 4 is a diagram showing the process of separating and processing a group of pixels corresponding to rain streaks from the background in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

That is, as shown in FIG. 4, the background area that does not correspond to the rain streak is converted to black in step (c-3), and subsequent processing can be performed through the processed image data converted in this way.

Next, step (d) is performed in which the rainfall analysis unit 130 of the management server 100 calculates diameter information of each rainfall particle included in the image data for analysis.

Figure 5 is a diagram showing the detailed process of step (d) in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

As shown in FIG. 5, step (d) in this embodiment may include steps (d-1) to (d-4) in detail.

Step (d-1) is a process in which the rain analysis unit 130 separates pixel groups corresponding to rain streaks by convolution processing the processed image data converted in step (c) described above.

Figure 6 is a diagram showing the process of separately processing pixel groups corresponding to rain streaks in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention, as shown in Figure 6. Likewise, it can be seen that the pixel group corresponding to the rain streak is clearly separated by step (d-1).

And in the process of performing step (d-1), the blur effect can be removed from the processed image data, and then CNN (Convolution Neural Network) techniques can be applied.

Afterwards, in step (d-2), the rainfall analysis unit 130 calculates the central axis of the individual rain pixel group divided by step (d-1), and in step (d-3), the rainfall analysis unit 130 calculates ( A process is performed to derive the long axis and short axis of individual rain streak pixel groups based on the central axis calculated in step d-2).

At this time, in steps (d-2) and (d-3), the process of calculating the fall direction can be performed together, and this is to clearly distinguish between the long axis and short axis of the rain pixel group.

Additionally, in step (d-4), the rainfall analysis unit 130 converts the short axis of the individual rain streak pixel group derived in step (d-3) into the diameter of the rainfall particle to calculate the diameter information of each rain particle. This takes into account the falling speed of rainfall particles and sets the diameter of the minor axis as the representative diameter of rainfall particles.

Figure 7 is a diagram showing the process of calculating diameter information of rainfall particles in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

In step (d), as shown in the upper left of FIG. 7, the pixel group corresponding to the rain streak is separated separately, and then the central axis of the individual rain streak pixel group is calculated as shown in the upper right of FIG. 7. As shown below, the short axis of each rain pixel group is converted to the diameter of the rain particle to calculate the diameter information of each rain particle.

At this time, in the case of step (d), the rainfall analysis unit 130 may reflect the settings of the CCTV 10 that acquired the original image data in the process of calculating the diameter information of the rainfall particles. This is because the number of pixels representing the actual rain streak area may vary depending on the settings of the CCTV 10.

Next, step (e) is performed in which the rainfall analysis unit 130 collects the diameter information of each rainfall particle calculated in step (d) and calculates the number concentration for each rainfall particle diameter.

Figure 7 is a diagram showing a graph of number concentration by rainfall particle diameter calculated by the rainfall analysis unit in the method for estimating rainfall particle distribution and rainfall intensity using CCTV according to an embodiment of the present invention.

As shown in Figure 7, in step (e), the number of rain particle diameters for a set time is calculated, and a graph of number concentration by particle diameter can be created through the accumulated number of particles. Through this, the water concentration, that is, the density value for each diameter of the raindrop, is calculated.

Next, step (f) is performed in which the rainfall analysis unit 130 derives the final rainfall intensity based on the number concentration by rainfall particle diameter calculated in step (e).

In step (f), the final rainfall intensity can be derived through Equation 1 below.

At this time, in calculating v(D) in this patent, Equation 2, which is the longitudinal velocity relational equation for each diameter of rainfall particles, can be used. Terminal velocity refers to the final speed that a falling particle can reach without being influenced by the wind or the surrounding environment.

Meanwhile, in this embodiment, between steps (e) and (f), a step (ex1) in which the rain analysis unit 130 calculates the effective volume within the depth of field of the image data for analysis, and the rain analysis unit 130 Step (ex2) of removing non-meteorological data from the image data for analysis can be further performed, and the results of steps (ex1) and (ex2) can be reflected in step (f).

And the final rainfall intensity derived according to the above process can be provided to an external terminal 200, such as a mobile terminal owned by an arbitrary user or a separately built external server, through the management server 100.

In this way, the present invention can reduce the limitations of the limited observation area of the rain gauge previously used through rainfall data acquired through video-based equipment called CCTV (10), and build a high spatial resolution observation network through low-cost equipment. Make this possible.

As described above, the preferred embodiments according to the present invention have been examined, and the fact that the present invention can be embodied in other specific forms in addition to the embodiments described above without departing from the spirit or scope thereof is recognized by those skilled in the art. It is self-evident to them. Therefore, the above-described embodiments are to be regarded as illustrative and not restrictive, and thus the present invention is not limited to the above description but may be modified within the scope of the appended claims and their equivalents.

10: CCTV
100: Management server
110: Data collection department
111: database
120: Image processing unit
130: Rainfall analysis department
200: external terminal

Claims (7)

Step (a) where the data collection unit receives original video data captured from CCTV installed within a randomly selected area and stores it in a database;
Step (b) in which an image processing unit preprocesses the original image data stored in the database and converts it into processed image data;
Step (c) wherein the image processing unit generates image data for analysis by removing background areas other than areas corresponding to rain streaks from the processed image data;
Step (d) where the rainfall analysis unit calculates diameter information of each rainfall particle included in the image data for analysis;
Step (e) where the rainfall analysis unit collects the diameter information of each rainfall particle calculated in step (d) and calculates the number concentration for each rainfall particle diameter; and
Step (f) in which the rainfall analysis unit derives the final rainfall intensity based on the number concentration for each diameter of the rainfall particles calculated in step (e);
Including,
Rainfall particle distribution and rainfall intensity estimation method using CCTV. According to paragraph 1,
In step (b),
The image processing unit converts the original image data into a black and white image with only contrast ratio information remaining,
Rainfall particle distribution and rainfall intensity estimation method using CCTV. According to paragraph 1,
In step (c),
Step (c-1) in which the image processing unit specifies a group of pixels corresponding to rain streaks in the processed image data;
Step (c-2) in which the image processing unit selects remaining pixel groups from the processed image data excluding pixel groups corresponding to rain streaks; and
Step (c-3) in which the image processing unit removes the brightness of the remaining pixel group selected in step (c-2) to generate processed image data in which the pixel group corresponding to the rain streak is separated from the background;
Including,
Rainfall particle distribution and rainfall intensity estimation method using CCTV. According to paragraph 1,
In step (d),
Step (d-1) in which the rain analysis unit performs convolution on the processed image data to separately classify a group of pixels corresponding to rain streaks;
Step (d-2) in which the rainfall analysis unit calculates the central axis of the individual rain pixel group divided by step (d-1);
Step (d-3) in which the rain analysis unit derives the long axis and short axis of each group of rain pixels based on the central axis calculated in step (d-2); and
Step (d-4) in which the rainfall analysis unit converts the short axis of the individual rain streak pixel group derived in step (d-3) into the diameter of the rainfall particle to calculate diameter information of each rainfall particle;
Including,
Rainfall particle distribution and rainfall intensity estimation method using CCTV. According to paragraph 4,
In step (d-4),
The rainfall analysis unit reflects the settings of the CCTV that acquired the original image data in the process of calculating the diameter information of the rainfall particles,
Rainfall particle distribution and rainfall intensity estimation method using CCTV. According to paragraph 1,
In step (f),
The rainfall analysis department,

Deriving the final rainfall intensity through the formula,
Rainfall particle distribution and rainfall intensity estimation method using CCTV. According to paragraph 1,
Between step (e) and step (f),
Step (ex1) of the rainfall analysis unit calculating an effective volume within the depth of field of the image data for analysis; and
Step (ex2) of the rainfall analysis unit removing non-meteorological data from the image data for analysis;
is further performed,
The results of step (ex1) and step (ex2) are reflected in step (f),
Rainfall particle distribution and rainfall intensity estimation method using CCTV.