KR20210105487A - Method for estimation of river bed change rate using hyperspectral image - Google Patents

Abstract

The present invention relates to a method for estimating water depth using a hyperspectral image, and specifically to a method that can estimate the depth of water with high accuracy only with hyperspectral images by setting a regression equation using a band ratio after acquiring a hyperspectral image using an unmanned aerial vehicle, and can measure the rate of river bed fluctuation using a river bed elevation calculated with the estimated water depth. The present invention includes a step of acquiring hyperspectral information for each preset image unit.

Description

Method for estimation of river bed change rate using hyperspectral image

The present invention relates to a method of measuring the rate of change of bed phase using a hyperspectral image. Specifically, after acquiring a hyperspectral image using an unmanned aerial vehicle, a regression equation is set using a band ratio, so that only a hyperspectral image is used. It relates to a method for estimating water depth with high accuracy, measuring the two-dimensional spatial distribution of river bed elevation, and measuring the river bed fluctuation rate based on this.

Although 3D modeling of information outside the water is relatively easy using various image equipment, 3D modeling that reflects information in water, especially the depth of water in a wide range of water bodies, is a very difficult task.

The most common method of measuring water depth is to use ultrasound. However, this method has low precision due to various external factors, such as reflections from obstacles, and is impossible at low water depth. In the low water depth, a person dives directly or a measurement method using a tape measure is mainly used. Even in this case, it is practically impossible to measure a wide range without omission.

As an alternative to this, in order to measure the depth of the water, a method of obtaining a hyperspectral image of the water body and using the obtained hyperspectral image is being discussed. This is because, when a hyperspectral sensor is used, it is possible to secure hundreds of times more information than that of a general image sensor, including not only the visible region but also the ultraviolet region and the infrared region.

In order to check the depth of the water using hyperspectral information, a database matching what the actual bands for each wavelength means is needed. For example, if the values of 8 bands (412, 490, 512, 555, 660, 680, 765, 875 nm) are known by the Goestationary Ocean Color Imager (GOGI) launched in 2010 , a database on whether each value means what kind of sea color is already secured. Therefore, when hyperspectral information for each point is confirmed by taking a hyperspectral image of the ocean, the sea color is estimated without direct observation using only the hyperspectral image by checking the value at the corresponding wavelength and the sea color matching it from the database. The same method can be applied to water depth.

In addition, since it is not possible to verify and database the observation values for each wavelength at all points, a regression equation is used. That is, if a regression equation is constructed using a large number of data, the actual value can be estimated using the band for each wavelength identified in the hyperspectral image by using the regression equation even at a location that is not in a database.

By combining these methods, it is theoretically possible to estimate the depth of water only with hyperspectral images. When a value for a specific wavelength is extracted from the hyperspectral image, the water depth can be estimated by using a database stored in which the corresponding value and the water depth are matched in advance or by substituting it into a regression equation.

However, it is known that there is a certain level of error in the method of estimating the depth of water using the photographing value of a specific wavelength in this way. As the causes of such errors, errors occurring in the hyperspectral information capturing step (for example, a situation in which the captured value does not represent the actual location value due to the influence of reflected light), and errors occurring in the image correction step (e.g., Errors in geometric correction, errors in image registration, etc.), errors in the database (eg, when the photographing value is correct but identification information is incorrect), and other external factors are suggested.

Moreover, when estimating the water depth through a regression equation constructed using a low-accuracy value, this error gradually increases.

In addition, it is difficult to secure a hyperspectral image of a wide area.

In order to secure a wide area of hyperspectral images, it is common to take pictures from a manned vehicle or an artificial satellite. For example, when measuring water depth, precision is reduced because it is measured in units of ^{100 m 2 on the surface.}

In order to solve this problem, attempts have been made to mount a hyperspectral sensor on an unmanned aerial vehicle such as a drone that can fly at a short distance. In this case, although the cell size is reduced and the accuracy is increased, another problem occurred in that the accuracy was lowered due to problems such as shaking during flight of the unmanned aerial vehicle. To prevent this, it is possible to consider mounting a high-performance gimbal that can compensate for shake, but the weight of the high-performance gimbal is quite large, which weakens the flight performance of the drone.

Moreover, since the hyperspectral sensor basically performs line scanning to secure hyperspectral information, software for matching a plurality of hyperspectral information is essential. However, the existing hyperspectral information matching software lacks accuracy, and thus, an image obtained by matching a plurality of hyperspectral information may provide information on a point slightly deviated from the actual shooting point.

Accordingly, the demand for a new type of depth identification method capable of increasing accuracy by reducing errors is on the rise, but the change in speed below the depth of the water has not been known.

Review the relevant patent literature.

Korean Patent No. 10-1621354 discloses a method for measuring depth using hyperspectral images. All of the above problems are included, so the accuracy is low.

Korean Patent Registration No. 10-1463351 discloses a technique for increasing the speed of information calculation by detecting a singular region using a band ratio in a hyperspectral image and excluding the detected region in a hyperspectral image. Only the point of using the band ratio and the present invention to be described below is common, but the overall purpose, configuration, and effect of the invention are different.

Korean Patent No. 10-1619836 discloses a drone using a gimbal and a shock absorber as a camera for hyperspectral imaging. It has an advantage in that an image that compensates for the shake of the drone is secured, but cannot start an image matching technology by line scanning.

Korean Patent Registration No. 10-1806488 discloses a technology for identifying a dark body by mounting a hyperspectral image sensor on a drone. Korean Patent No. 10-1744662 discloses a technology for classifying plants using hyperspectral images. To this end, hyperspectral image standard data for various dark bodies or plants must be secured in advance, and the photographed object is identified by comparing the photographed image thereto. However, it does not solve the fundamental problems identified during the shooting stage, such as a problem of decreasing accuracy due to the shaking of the drone, and a problem of increasing the efficiency of image matching.

Korean Patent No. 10-1689197 discloses a gimbal assembly that can be attached to a drone. The assembly disclosed herein is composed of a first unit capable of rotation/reverse rotation, a second unit capable of lifting and lowering, and a third unit rotating them, thereby compensating for the shaking of the aircraft. However, when configured in this way, the weight of the gimbal is excessively increased, which will cause a problem that the flight performance of the drone is lowered.

Korean Patent Application Laid-Open No. 10-2014-0014819 discloses a method of identifying data in the form of an artificial intelligence algorithm by applying an indexing/querying technique to data identified in a hyperspectral image. The prerequisite for using this method is the accuracy of hyperspectral images, but it cannot even suggest a method to increase the accuracy of images captured in an environment such as an unmanned aerial vehicle.

US Patent No. 7,181,055 proposes an algorithm for registering hyperspectral images. Here, a reflectance image is used to increase the registration accuracy.

Korean Patent Registration No. 10-1463351 Korean Patent Registration No. 10-1619836 Korean Patent No. 10-1806488 Korean Patent No. 10-1744662 Korean Patent Registration No. 10-1689197 Korean Patent Publication No. 10-2014-0014819 US Patent No. 7,181,055 Korean Patent Registration No. 10-1963604 Korean Patent No. 10-1621354 Korean Patent Registration No. 10-1936586

The present invention has been devised to solve the above problems. In particular, in the case of estimating water depth using hyperspectral images, various problems that may occur are to be solved.

Specifically, it is intended to solve the problem that the accuracy of image registration is lowered due to the characteristics of the line-scanning hyperspectral sensor, and to solve the problem of shaking of the image due to shooting from an unmanned aerial vehicle.

In addition, to propose a novel hyperspectral information identification method to obtain accurate hyperspectral information and prevent a situation in which the database for the corresponding photographing value does not reflect accurate identification information even when the photographing value of a specific wavelength is accurate. do.

In particular, we intend to propose a method for calculating the riverbed elevation using the estimated water depth and measuring the rate of change of the riverbed using the calculated riverbed elevation.

,

하상 변동속도 측정 방법을 제공한다.One embodiment of the present invention for solving the above problems is (a) the unmanned aerial vehicle 100, while flying along a set flight path, using the hyperspectral sensor 140, line scanning (line scanning) through, acquiring hyperspectral information for each preset image unit; (b) performing, by the geometric correction performing module 250, geometric correction on the hyperspectral information obtained for each image unit by using the position information of each pixel; (c) acquiring, by the image matching performing module 260, the hyperspectral image of the collection area by matching the hyperspectral information obtained for each image unit on which the geometric correction is performed; (d) inputting a depth observation value (Y) of each of a plurality of predetermined points to the hyperspectral observation value input module 310; (e) the band ratio calculation module 320, at each of the plurality of predetermined points, from the hyperspectral image obtained in step (c), a hyperspectral imaging value (I) that is a pixel value for a plurality of wavelengths (x) _{(x)), select a pair (λ 1} , λ ₂ ) of any two wavelengths among a plurality of wavelengths (x) included in the hyperspectral imaging value, and select the hyperspectral imaging value (I( determining λ ₁ ), I(λ ₂ )); (f) the band ratio calculation module 320 _{divides any one of the two identified hyperspectral imaging values (I(λ 1} ), I(λ ₂ )) by the other to obtain a band ratio (X) ) to calculate; and (g) assuming that the regression analysis module 330 has a linear relationship between the observed depth value (Y) and the calculated band ratio (X), the depth observation value (Y) at the plurality of points and generating a regression equation (Y = aX + b) by calculating a and b of a regression equation (Y = aX + b) using the calculated band ratio (X), (h) estimating, by the observation value estimation module 350, the depth of the water from the hyperspectral image using the regression equation generated in step (g); and (m) calculating, by the river bed fluctuation rate calculation module 380, the river bed fluctuation rate according to a preset method using the water depth estimated in the step (h). m1) calculating, by the observation value estimation module 350, a normal vegetation index (NDWI) using the intensity of light with respect to a wavelength corresponding to near infrared in the hyperspectral image; (m2) determining, by the terrain determination module 360, a waterbody zone when the calculated regular vegetation index (NDWI) is greater than or equal to 0, and determined as a river sub-region when it is less than 0 to determine each terrain; (m3) The elevation calculation module 370 calculates the elevation corresponding to the location by using the location information of each pixel of a plurality of points constituting the outer line of the water body area among the topography determined by the topography determination module 360 estimating as the elevation of the water surface by utilizing; (m4) calculating, by the elevation calculation module 370, the river elevation E(x,y,t) by subtracting the estimated water depth from the calculated elevation of the water surface; (m5) assuming that the change in the river bed elevation calculated in step (m4) is 0 during the time change amount δt by the river bed fluctuation rate calculation module 380, calculating the river bed fluctuation rate through the following formula; containing

,

A method for measuring the rate of change of the river bed is provided.

여기서,

이며, q_n은 수체 구역 내 임의의 지점인 것이 바람직하다.In addition, the step (m5) comprises the step of calculating the rate of change of the lower phase through the following equation,

here,

, and q _n is preferably any point in the waterbody zone.

In addition, after the (m3) step, before the (m4) step, the elevation calculation module 370, the database unit 400, the elevation information of the area corresponding to the river area determined in the step (m2) Collecting; further comprising, the step (m5), it is preferable to further include the step of calculating the rate of change of the river bed for the river sub-region.

In addition, after step (g), (h) the regression analysis module 330 selects any other than the pair of the two wavelengths in step (e) among the plurality of wavelengths included in the hyperspectral imaging value. generating a regression equation for each of a plurality of pairs of two wavelengths by selecting a wavelength of , checking a hyperspectral imaging value thereof, and repeating steps (f) to (g); (i) selecting, by the regression equation determination module 340, a regression equation having the highest correlation coefficient among the generated plurality of regression equations; (j) determining, by the regression equation determination module 340, the two wavelengths applied to the band ratio (X) used in the regression equation selected in the step (g) as a pair of optimal band ratio wavelengths; And, it is preferable to estimate the water depth from the hyperspectral image using the selected regression equation.

In addition, after step (j), (k) the observation value estimation module 350 corresponds to the pair of optimal band ratio wavelengths determined in step (j) in hyperspectral imaging values of points other than the plurality of points. checking two hyperspectral imaging values, dividing one into the other, and calculating a band ratio of the corresponding point; and (l) estimating, by the observation value estimation module 350, the observation value of the depth of the corresponding point by inputting the calculated band ratio into the regression equation selected in the step (g).

In addition, the step (l) may include: calculating, by the observation value estimation module 350, a normal vegetation index (NDWI) using the intensity of light for a wavelength corresponding to near-infrared in the hyperspectral image; and outputting, by the observation value estimation module 350, the depth of water as 0 when the calculated NDWI is less than 0.

여기에서, NIR은 상기 초분광정보에 포함된 값 중 근적외선에 해당하는 파장에서의 빛의 세기이고, Green은 상기 초분광정보에 포함된 값 중 녹색에 해당하는 파장에서의 빛의 세기인 것이 바람직하다.In addition, the regular expression index (NDWI) is calculated by the following formula,

Here, NIR is the intensity of light at a wavelength corresponding to near infrared among the values included in the hyperspectral information, and Green is the intensity of light at a wavelength corresponding to green among the values included in the hyperspectral information. do.

In addition, in step (i), the regression equation determination module 340 calculates the correlation coefficients for the generated plurality of regression equations, and converts each wavelength in the pair of two wavelengths to the X-axis and the Y-axis. and generating a correlation coefficient map to Preferably, the method includes determining each wavelength constituting the corresponding point as a pair of the optimal band ratio wavelengths.

In addition, in step (c), (c11) the corresponding point checking module 262, in any two hyperspectral information (A, B) obtained for each image unit on which the geometric correction is performed, any one With respect to the XY coordinates (x, y) of any one pixel of the hyperspectral information A and the pixel value (I(x, y)) at the coordinates, any one of the other hyperspectral information B Calculate the normalized correlation of the XY coordinates (x', y') of the pixel and the pixel value (I(x', y')) at the corresponding coordinates, but any one of the hyperspectral information (A) By calculating all the normal correlation coefficients of each pixel of the other hyperspectral information (B) for each pixel of the two hyperspectral information (A, B) confirming the corresponding point; (c12) image registration of the two hyperspectral information (A, B) based on the identified corresponding point by the image matching performing module 260; and (c13) the image registration performing module 260 performs the steps (c11) to (c12) for each pair of hyperspectral information for each image unit obtained in the step (a), It is preferable to include the step of acquiring a hyperspectral image of the collected area.

In addition, before step (c11), (c21) the integral image operation module 261, in the hyperspectral information obtained for each image unit on which the geometric correction is performed, for each pixel, any one reference coordinate The method may further include calculating an integral image pixel value based on , wherein the pixel value in step (c11) is the pixel value of the integral image calculated in step (c21).

In addition, before the step (c21), (c31) the normalization module 240 sets the XY coordinates of the pixels of the hyperspectral information of any one image unit as an ordered pair of integers (x, y), and the set XY Calculate the average coordinates of the coordinates, set the calculated average coordinates as the origin (0, 0) so that the unit distance between the ordered pairs (x, y) of each integer becomes √2, the XY coordinates (x, y) It is preferable to further include normalizing the hyperspectral information of any one image unit by moving the to the normalized XY coordinates (x', y').

In addition, after step (c31), (c32) the normalization module 240 XY coordinates (x, y) before normalization of any one of the hyperspectral information and XY coordinates after normalization (x', y') calculating a homography (H), which is a determinant including parameters, by using a value of ; and (c33) normalizing the other hyperspectral information using the calculated homography (H).

상기 (c33) 단계는, 다른 하나의 초분광정보를 행렬식 A에 대입하여, 정규화된 XY 좌표인 행렬식 A'가 연산되는 것이 바람직하다.In addition, in the step (c32), homogeneous coordinates are used, so that the XY coordinates before normalization are (x, y, 1) and are referred to as the determinant A, and the XY coordinates after normalization are (x', y', 1) is referred to as the determinant A', and the homography (H) is calculated using the following formula but using a number of XY coordinates,

In the step (c33), it is preferable that the determinant A', which is the normalized XY coordinates, is calculated by substituting another piece of hyperspectral information into the determinant A.

Another embodiment of the present invention for solving the above problems is stored in a computer-readable recording medium, and provides a program for executing the above-described method, and a computer-readable recording medium in which the code is recorded.

The method according to the present invention obtains an excellent hyperspectral image with both high accuracy and precision by matching hyperspectral information secured in a small image unit (line unit) with high accuracy, and based on this, it acquires a water depth depth with high accuracy and precision. can be estimated In particular, it is possible to effectively correct the shake of the unmanned aerial vehicle without using a high-performance gimbal with a high weight. Used together with the above-described image registration technology, it allows to obtain excellent hyperspectral images with a high synergistic effect.

In addition, since only one wavelength having a value identified as closest to the wavelength value of the real object is used in the prior art, the accuracy of the actual estimated water depth is low when the matching is inaccurate or there is an error in the measurement value However, according to the present invention, the photographed value of the wavelength identified as closest to the value of the wavelength of the actual water depth is used as the numerator, and the photographed value of the wavelength identified as the closest to the value of the wavelength of the actual water depth is used as the denominator. Unlike matching by wavelength, it is possible to increase the accuracy by estimating the value of the water depth through a double matching process.

In addition, the riverbed elevation can be calculated using the estimated water depth value, and the riverbed change rate can be estimated using the calculated riverbed elevation.

In addition, this method is computer-programmed in a non-stop concept as only one program, and when the water depth and the captured hyperspectral image, which are observation values of multiple points, are input, the hyperspectral image is appropriately pre-processed and It is post-processed, and a regression equation is automatically determined based on this, and the water depth can be estimated automatically, and the river bed fluctuation rate can be estimated using this.

In particular, in the past, it was necessary to perform complex calculations and data transmission by going back and forth between multiple programs. However, all these problems are solved, and the calculation time that used to take several days can be reduced by about 50% or more by using the integral image.

1 shows a system for carrying out a method according to the invention;
2 is a flowchart illustrating an image processing method among methods according to the present invention.
3 is a flowchart illustrating an image analysis method among methods according to the present invention.
4 is a flowchart for explaining a method for measuring a river bed fluctuation rate according to the present invention.
5 is a photograph for explaining an example of an unmanned aerial vehicle used in the method according to the present invention.
6 is a schematic diagram for explaining a method for an unmanned aerial vehicle to acquire hyperspectral information in the method according to the present invention.
7 is a schematic diagram for explaining geometric correction in an image processing method among methods according to the present invention.
8 is a schematic diagram for explaining image registration in an image processing method among methods according to the present invention.
9 is a schematic diagram for explaining a method of using a normal correlation coefficient for image registration in an image processing method among methods according to the present invention.
10 is a schematic diagram for explaining a method of utilizing an integral image during image registration in an image processing method among methods according to the present invention.
11 is a schematic diagram for explaining a normalization step in an image processing method among methods according to the present invention.
12 is a diagram for explaining a band ratio of an image analysis method among methods according to the present invention.
13 is a view for explaining a process in which a regression analysis is performed using a band ratio and a water depth as an observation value in an image analysis method among the methods according to the present invention.
14 illustrates a correlation map set by an image analysis method among methods according to the present invention.
15 illustrates a method of selecting a regression equation using a correlation map by an image analysis method among methods according to the present invention.
16 shows a screen of a program in which the method according to the present invention is implemented.
17 illustrates a screen on which a final result is output through post-correction after completing the estimation of the water body by applying the water body determination step in the method according to the present invention.
18 shows an example of a result of classifying a waterbody zone in which water is present using the method according to the present invention.
19 shows a method for calculating the river bed elevation in the method according to the present invention.
20 shows a flow phase change rate measurement procedure using two hyperspectral images with a time difference using the method according to the present invention.

Hereinafter, the present invention will be described in more detail with reference to the drawings.

Hereinafter, a method of estimating the depth of the observation value with high accuracy without actually measuring using the hyperspectral image will be described. However, if the concept of the method described below and described in the claims is applied, identification of objects other than the depth of water (eg, sea color, rocks, vegetation, etc.) It will be understood by those skilled in the art that, if made according to the concept within, the scope of the rights thereto may also be included within the scope of the present invention as claimed in the claims below.

Description of the system

The system in which the method according to the present invention is performed includes an unmanned aerial vehicle 100 equipped with a hyperspectral sensor 140 and an echo sounder 101 to photograph a hyperspectral image and sense ultrasonic waves, and the accuracy of the captured hyperspectral image. It can be divided into an image processing unit 200 that performs processing to improve, and a database unit 400 electrically connected to an image analysis unit 300 that performs a function of estimating water depth based on the processed hyperspectral image. have.

In one embodiment, only the hyperspectral sensor 140 may be mounted on the unmanned aerial vehicle 100 .

The unmanned aerial vehicle 100 may be, for example, a drone, but may be any type of vehicle that flies at a short distance from the ground surface.

The unmanned aerial vehicle 100 includes a gimbal 110 that has a gyro sensor and an acceleration sensor to move the camera 130 in the opposite direction to the movement, a shock absorber 120 that is provided in the gimbal 110 and absorbs shock; It includes a camera 130 on which the hyperspectral sensor 140 is mounted, and a communication unit 150 capable of transmitting hyperspectral information obtained from the hyperspectral sensor 140 to the image processing unit 200 .

The gimbal 110 may use a gimbal of any structure that can be secured, but in consideration of the weight, a 1, 2, or 3 axis gimbal is preferable rather than a 4 or 5 axis gimbal. However, the limitation of shake compensation of the 1, 2, and 3-axis gimbals can be overcome by the compensation of the shock absorber 120 .

In addition, the unmanned aerial vehicle 100 may further include a GPS sensor for confirming location information.

In addition, the echo sounder 101 is a device for measuring water depth using an ultrasonic sensor, and may measure a high depth of water. When the echo sounder 101 is mounted on the unmanned aerial vehicle 100 , there is a limit to the depth of water that can be measured using hyperspectral information, and it is mounted to solve this problem. In the echo sounder 101, the topography determination module 360 measures a water depth (depth that cannot be sensed by the hyperspectral sensor 140) that exceeds a predetermined criterion and transmits it to the image processing unit 200 as ultrasonic information. . Preferably, the predetermined criterion in the terrain determination module 360 may be the maximum depth that can be calculated through the hyperspectral image, and specifically, it may be 3.5m to 4m. That is, for a water body region having a water depth exceeding 3.5 m to 4 m, the echo sounder 101 measures the water depth using ultrasound information and transmits it to the image processing unit 200 .

An example of such an unmanned aerial vehicle 100 is shown in FIG. 5 .

The image processing unit 200 includes a communication module 210 , a flight path setting module 220 , a captured image input module 230 , a normalization performing module 240 , a geometric correction performing module 250 , an image matching performing module 260 ) and a hyperspectral image output module 270 .

The communication module 210 communicates with the communication unit 150 of the unmanned aerial vehicle 100 by wire or wirelessly so that the captured image input module 230 receives the hyperspectral information obtained from the hyperspectral sensor 140 . In an embodiment, the captured image input module 230 may receive ultrasound information obtained from the echo sounder. The received information may include location information.

The flight path setting module 220 performs a function of setting the flight path of the unmanned aerial vehicle 100 for securing the image based on a collection area for collecting hyperspectral information and/or ultrasound information.

The normalization performing module 240 performs a normalization function of the secured hyperspectral information. In all geometric transformations, rotation, magnitude, parallelism preservation, etc. occur with respect to the origin (0, 0), so normalization of each ordered pair of XY coordinates present in each image is necessary.

The geometric correction performing module 250 performs geometric correction for compensating for distortion caused by geometric conditions such as curvature of the ground in the secured hyperspectral information.

The image matching performing module 260 performs a function of matching a plurality of secured hyperspectral information. To this end, an integral image operation module 261 and a corresponding point confirmation module 262 are provided.

The hyperspectral image output module 270 performs a function of outputting a hyperspectral image obtained as a normalized, geometrically corrected, and image-registered final result. In this case, according to an embodiment, an image obtained as a final result by ultrasound information and an image obtained as a result of river zoning or topography information provided by the database unit 400 may be matched and provided. Accordingly, it is possible to provide accurate images by providing technology according to the range and level that can measure the topography of the river.

A method of performing a specific function of each module will be described later.

The image analysis unit 300 includes a hyperspectral observation value input module 310 , a band ratio calculation module 320 , a regression analysis module 330 , a regression expression determination module 340 , an observation value estimation module 350 , a topography a determination module 360 , and an elevation estimation module 370 .

The hyperspectral observation value input module 310 performs a function of inputting a water depth (hereinafter, referred to as a 'depth observation value') that is an actual hyperspectral observation value for a plurality of predetermined points. That is, the actual depth of water is input to check the correlation with the captured values secured in the hyperspectral image.

The band ratio calculation module 320 calculates a band ratio based on a photographing value (eg, a pixel value) secured from the hyperspectral image from the image processing unit 200 .

The regression analysis module 330 calculates a regression equation by performing a regression analysis of the depth observation value and the band ratio.

The regression equation determination module 340 calculates the correlation of a plurality of regression equations, determines the most highly correlated regression equation, and determines the optimal band ratio through this.

The observation value estimation module 350 uses the regression equation determined by the regression equation determination module 340 to input a specific hyperspectral image (that is, when an arbitrary band ratio is input), based on the observation value It estimates the actual water depth and performs the function of identifying a real object based on this.

The terrain determination module 360 determines the terrain using the estimated actual water depth and the regular vegetation index (NDWI). It can be broadly divided into river exclusion zone (waterbody zone) and river drainage zone (vegetation zone). The region having ? can be further divided into a low-water depth region.

The topography determination module 360 may determine an area using the estimated actual water depth and the NDWI, and may output the determined area as a water body area and a river area.

Specifically, if the calculated regular vegetation index (NDWI) is greater than or equal to 0, it is determined as a waterbody area, and when it is less than 0, it is determined as a river sub-region, and each topography is determined and then output.

The elevation calculation module 370 receives information from the observation value estimation module 350 and the terrain determination module 360, and the actual water depth estimated by the observation value estimation module 350 and the water surface elevation determined by the topography determination module 360 It is a module that calculates and determines the elevation of an arbitrary point using information such as At this time, interpolation is performed to convert the data measured as a point into spatial data.

The elevation calculation module 370 calculates the river elevation E(x,y,t) by using a stored equation or by subtracting the estimated water depth from the calculated elevation of the water surface.

The river bed fluctuation rate calculation module 380 receives information from the observation value estimation module 350, the topography determination module 360 and the elevation calculation module 370, and receives the information from the river bed fluctuation rate in a preset method using the stored equation. calculate At this time, the river bed fluctuation rate calculation module 380 receives the information during the time change amount δt and calculates the river bed fluctuation rate of the waterbody zone and the river sub-region.

A method of performing a specific function of each module will be described later.

description of the method

The method according to the invention is described. An image processing method with reference to FIG. 2 and an image analysis method with reference to FIG. 3 will be described separately.

First, an image processing method will be described with reference to FIGS. 2 and 6 to 11 .

The flight path setting module 220 sets the flight path of the unmanned aerial vehicle 100 using the collection area (S110). The collection area will be the waterbody and the river exclusion area around which the water depth is to be estimated. As the echo sounder 101 and the hyperspectral sensor 140 perform line scanning, as shown in FIG. 6 , the unmanned aerial vehicle is a target for acquiring ultrasonic information and hyperspectral information. In consideration of the angle of view of (100), a flight path of the unmanned aerial vehicle 100 is set. The setting method may be any according to the prior art, and detailed description thereof will be omitted.

Next, while the unmanned aerial vehicle 100 flies along the set flight path, using the echo sounder 101 and the hyperspectral sensor 140, through line scanning, ultrasonic information and hyperspectral information for each preset image unit is obtained (S120). Here, the image unit is preset by the manufacturer according to the characteristics of the camera 130 or the hyperspectral sensor 140, and ultrasound information and hyperspectral information may be acquired in different ways according to each characteristic.

With respect to the hyperspectral information confirmed in this way, geometric correction is performed (S140) and image registration is performed (S150), so that a hyperspectral image to be analyzed is secured. The same may be performed for ultrasound information. On the other hand, since the ultrasound information is generated by the echo sounder 101, the depth value may be directly obtained by a preset method without such a process.

A specific method will be described.

The geometric correction is performed by the geometric correction performing module 250 (see FIG. 7 ). Specifically, geometric correction is performed on the hyperspectral information obtained for each image unit by using the position information of each pixel. Any conventionally known method may be applied. This process can be performed in the same way for ultrasound information.

Image registration is a process of matching a plurality of hyperspectral information using a certain reference point (refer to FIG. 8). At least two pairs of hyperspectral information are photographed so as to partially overlap each other, a predetermined reference point is found, and two pairs of hyperspectral information are superimposed on the basis of this reference point, and this process is performed in each pair of all hyperspectral information. By repeating in , hyperspectral images are obtained. That is, the basis of image registration is the process of superimposing two pieces of hyperspectral information.

In the present invention, focusing on the "corresponding point extraction algorithm", image registration is performed accordingly. This is by extracting a template from any one hyperspectral information among the two hyperspectral information having an overlapping area, and finding the point with the highest correlation in the other hyperspectral information with the extracted template. It is an algorithm that sets α as a point that serves as a reference point for superposition, that is, a matching point. This is shown in FIG. 9 .

Specifically, image registration of two hyperspectral information A and B (indicated by image1 and image2 in Fig. 9) with overlapping regions will be described.

First, the corresponding point confirmation module 262, in any two hyperspectral information (A, B) overlapping, the XY coordinates (x, y) of any one pixel of any one hyperspectral information (A) and the corresponding With respect to the pixel value (I(x, y)) in the coordinates, the XY coordinates (x', y') of any one pixel of the other hyperspectral information (B) and the pixel value (I( x', y'))), the normalized correlation is calculated. In this process, as will be described later, the XY coordinates (x') of all the pixels of the template of the hyperspectral information (B) are different for each XY coordinate (x, y) of all the pixels of the template of the hyperspectral information (A). , y') is made multiple times by matching all of them.

Equation for calculating a general normal correlation coefficient is the same as Equation 1 below, but in the present invention, Equation 2 using more pixel values is used to secure higher accuracy.

The XY coordinates of the hyperspectral information (A) are expressed as (x, y), and the XY coordinates of the hyperspectral information (B) are expressed as (x', y'). The upper bar means the average value (this is the same for all equations below). n is the total number of pixels, I(x, y) is a pixel value corresponding to (x, y) on XY coordinates, the pixel value of hyperspectral information (A) _{is expressed as I 1} and hyperspectral information (B) The pixel value of is expressed as _{I 2 .} W is the horizontal pixel size of the template, and H is the vertical pixel size of the template.

In this way, if a normal correlation coefficient is calculated for every pixel coordinate on the template and expressed in XY coordinates, a correlation map as shown in the lower right of FIG. 9 can be calculated. The position having the maximum value in the correlation coefficient map, that is, the XY coordinates of the point having a large normal correlation coefficient value is identified as the corresponding point of the two hyperspectral information (A, B).

Now, the image matching performing module 260 performs image matching on the two hyperspectral information A and B based on the corresponding point.

The image matching performing module 260 performs image registration by performing this process for each pair of all hyperspectral information for each image unit, thereby obtaining a hyperspectral image in which all captured images are matched.

However, in this case, since quite a lot of calculations are performed, the calculation speed may be slow depending on the performance of the image processing unit 200 . Accordingly, in another embodiment of the present invention, in order to realize a faster operation speed, image registration may be performed in a state in which the amount of computation is greatly reduced by using an integral image. In the integral image, reference coordinates are set, and the pixel values of each pixel are summed up using the directionality based on the reference coordinates, and then used as the pixel values necessary for the normal correlation coefficient calculation as in Equation (2).

A concept of using an integral image will be described with reference to FIG. 10 . The left shows a typical image. The number recorded for each pixel is the pixel value. The right side shows an integral image using the left image. The integral image here is an integral image formed by setting “2” in the upper left corner as a reference coordinate and then summing the values in the left and lower directions. For example, "13" in the 2nd row and 3rd column is the value of 2+1+5+0+3+2.

If you want to find the sum of the pixel values of nine pixels in the middle of the image, you need to perform a total of nine operations on the general image on the left. However, in the integral image on the right, it is sufficient to perform four operations. That is, by generating an integral image, the amount of computation is reduced to 50% or less.

Specifically, the method of using the integral image in the hyperspectral image is as follows. First, a new pixel value II(x, y) based on the integral image is set for each pixel by using Equation 3.

Each symbol is as shown in Equation (2). Now, using this, the normal correlation coefficient described in Equation (2) can be calculated by Equation (4).

Here, the value of the standard deviation σ used in the denominator may be calculated through Equation 5. The fact that I means a pixel value is the same as described above.

In this way, once the integral image is created and placed, by using Equations 3 to 5, there is no need to use the large amount of calculation of Equation 2 for every operation, so that the normal correlation coefficient can be calculated quickly. It is preferable to be able to

On the other hand, as described above, in all geometric transformations, rotation, size, parallelism preservation, etc. occur based on the origin (0, 0), so in an embodiment of the present invention, the XY coordinates present in each image are After performing normalization of each ordered pair, geometric correction and image registration may be performed.

Specifically, the normalization module 240 sets the XY coordinates of the pixels of the hyperspectral information of any one image unit as an ordered pair of integers (x, y), calculates the average coordinates of the set XY coordinates, and the operation The XY coordinates (x, y) are normalized to the XY coordinates (x', y) so that the unit distance between the ordered pairs (x, y) of each integer is √2 by setting the average coordinates as the origin (0, 0). '), normalizes the hyperspectral information of any one image unit. This concept is illustrated in FIG. 11 .

More specifically, this operation may use Equation (6).

Here, A is the XY coordinates in hyperspectral information to be normalized (ie, not normalized), and A' is the normalized XY coordinates. For the coordinates here, it is preferable to use a homogeneous coordinate system for parameter calculation to be described later. That is, the XY coordinates are referred to as the determinant A as (x, y, 1), and the XY coordinates after normalization are referred to as the determinant A' as (x', y', 1). It is expressed as a kind of vector. In other words, the (x, y) coordinates in the homogeneous coordinate system can be understood as a line made up of countless points.

H is a homography, which is a matrix composed of _{parameters (h 11} to h _{33 ) for normalization.} That is, if the homography (H) is calculated, all hyperspectral information can be easily normalized.

For the calculation of the homography (H), Equation 6 is used, but any one normalized value is required. That is, any one hyperspectral information is separately normalized using the method described with reference to FIG. 11 , and then the homography H is calculated using the value. For example, if four ordered pairs of x and y of the XY coordinates before normalization are known, and the normalized values of the four ordered pairs are known, using Equation 6, Equations 7 and 8 can be derived and, accordingly, each parameter (h ₁₁ ~ h ₃₃ ) can be checked.

If the homography (H) is calculated according to Equation 8, now, by substituting the XY coordinates of all pixels of all hyperspectral images into Equation 6, normalization is possible only with a small amount of computation without separately calculating normalized values do.

As described above, the existence of the homography H means that not only coordinate changes between corresponding points are possible, but also all points existing in the coordinate system of two hyperspectral information can be transformed into each other. Accordingly, when coordinate transformation is performed by normalizing the other hyperspectral information based on one hyperspectral information among the two hyperspectral information, the two hyperspectral images are moved to the same coordinate system.

Next, an image analysis method will be described with reference to FIGS. 3 and 12 to 17 .

The hyperspectral image on which the above-described image processing method has been performed is secured. In addition, for the estimation of the depth, the observation value Y, which is each hyperspectral observation value, for each of a plurality of predetermined points must be separately identified (S210). That is, the observation value Y of each of a plurality of predetermined points is input to the hyperspectral observation value input module 310 . Here, for the accuracy of the regression analysis, it is preferable that the number of points at which the observation value Y is confirmed is greater.

Next, a photographed image having hyperspectral information is input to the photographed image input module 210, and a hyperspectral photographing value ( I(x)) is confirmed (S220).

_{Next, the band ratio calculation module 320 selects a pair (λ 1} , λ ₂ ) of any two wavelengths among a plurality of wavelengths (x) included in the hyperspectral imaging value, and a hyperspectral imaging value for each (I(λ ₁ ), I(λ ₂ )) is checked (S230). At this time, the band ratio X is calculated by dividing one of the two identified hyperspectral imaging values I(λ ₁ ) and I(λ ₂ ) by the other ( S240 ) (see FIG. 12 ). If this is expressed as an equation, it is shown in Equation 9 below.

In other words, the band ratio is calculated for each point to check the depth observation value (Y) and the band ratio (X), which is performed individually for each of a plurality of points.

Next, the regression analysis module 330 assumes that the depth observation value (Y) and the calculated band ratio (X) are in a linear relationship, and the depth observation value (Y) confirmed in step S210 at each of a plurality of points and By using the band ratio (X) calculated in step S240, a and b of the regression equation (Y=aX+b) are calculated, and a regression equation (Y=aX+b) is generated (S250). Here, a specific formula for calculating a is the same as Equation 10 below, and when the value of a is calculated, the value of b is naturally calculated. n is the number of depth observations, X is the band ratio, and Y is the depth observation.

13 shows an example thereof. Here, the regression equation calculated using the band ratio (X) and the depth observation value (Y) at five points (P1, P2, P3, P4, P5) (ie, n=5) is shown in the lower right corner. .

Meanwhile, this process may be performed for each pair while changing the two wavelengths selected in step S230. Even if the predetermined point is the same and the observation value (Y) of the depth at each point is the same, if two wavelengths are selected differently, the final calculated regression equation is different.

Specifically, the regression analysis module 330 selects an arbitrary wavelength other than the first selected pair of two wavelengths, checks the hyperspectral imaging value for this, and repeats steps S240 to S250, a plurality of regression equations to generate (S260).

Next, the regression equation determination module 340 calculates the correlation coefficients of the generated multiple regression equations, and selects a regression equation having the highest correlation coefficient among them ( S270 ). At this time, two wavelengths applied to the band ratio (X) used in the selected regression equation are determined as a pair of optimal band ratio wavelengths.

For example, if the number of wavelengths included in the hyperspectral information is N, the number of pairs of two wavelengths is _N C ₂ by the combination. The optimal regression equation and the band ratio and its wavelength will be calculated.

On the other hand, this method may be achieved using a correlation coefficient map. An example of the correlation coefficient map is shown in FIG. 13 (note that a concept different from the numerical value applied to the correlation coefficient map of the image processing method described in FIG. 9 is applied). Here, the regression expression determination module 340 calculates the correlation coefficients for the generated plurality of regression equations, but in a pair of two wavelengths, one wavelength is the X-axis and the other wavelength is the Y-axis. The generated correlation coefficient map is shown.

Next, the regression expression determination module 340 determines a point having the highest correlation coefficient using the correlation coefficient map, and determines each wavelength constituting the corresponding point as a pair of the optimal band ratio wavelengths.

15 shows the concept of automating this scheme. By extracting the imaging value for each wavelength, which is a spectral characteristic, and calculating each band ratio through this, and regression analysis using the depth observation value, a correlation coefficient map is generated, and by calculating the correlation coefficient for each pair of wavelengths, An optimal band ratio wavelength and an optimal regression equation are determined.

16 shows an output screen of a program implementing this. The correlation coefficient map on the right is shown, and it can be seen that the correlation coefficient and regression equation for each wavelength are automatically calculated below it.

Through this process, when the optimal regression equation and the optimal band ratio are determined, the depth observation value can be estimated using only the wavelength and the photographing value of the hyperspectral information included in the hyperspectral image even at a position where there is no depth observation value.

For example, when the observation value estimation module 350 calculates the band ratio by checking two hyperspectral imaging values corresponding to the pair of optimal band ratio wavelengths among the imaging values included in the hyperspectral image, it is used for optimal regression. By entering into the equation, the observation value of the depth of the corresponding point can be estimated with high accuracy.

Meanwhile, in a preferred embodiment of the present invention, the method may further include a step of preventing a phenomenon in which the final result, the observation value of depth, is distorted due to vegetation or the like. To this end, the observation value estimation module 350 calculates the normal vegetation index (NDWI) by using the intensity of light with respect to the wavelength corresponding to the near-infrared (NIR) in the hyperspectral image.

The regular expression index (NDWI) is calculated by the following formula. Here, NIR is the intensity of light at a wavelength corresponding to near infrared among the values included in the hyperspectral information of the hyperspectral image, and Green is the intensity of light at the wavelength corresponding to green among the values included in the hyperspectral information of the hyperspectral image. is the intensity of light.

The regular expression index (NDWI), which is a value calculated by Equation 11 above, has a value in the range of -1 to 1. If the regular vegetation index (NDWI) is 0 or more, it is judged as a waterbody (streams or rivers) and the water depth observation value estimated in the previous step can be output as it is. ), and output the water depth as 0 or output as a river exclusion zone. At this time, if it is less than 0, it is determined that the water body is not, and the water depth is output as 0, thereby compensating for the partial calculation of the water depth due to distortion due to vegetation or the like.

The left side of FIG. 17 shows an image captured by a general camera, and the right side shows the result of performing the method according to the present invention. The water chain part (blue part) and the non-water part part were distinguished by the use of the regular vegetation index (NDWI). By comparing this with real data, it was possible to verify that the accuracy of the method according to the present invention is high. By performing these steps, the accuracy of estimating the depth observation value can be further improved.

A method of calculating the elevation of the river bed for measuring the rate of change of the river bed will be further described with reference to FIGS. 4 and 18 to 20 .

According to the present invention, the river topography can be surveyed by further using the elevation calculation module 370 in order to improve the accuracy and compensate for the limitations that occur in estimating the observation value of the water depth for the high depth with the hyperspectral image.

First, the observation value estimation module 350 calculates the normal vegetation index (NDWI) by using the intensity of light with respect to the wavelength corresponding to the near infrared in the hyperspectral image (S310).

Next, the topography determination module 360 determines a region in which the calculated regular vegetation index (NDWI) is equal to or greater than 0 as a waterbody region, and determines a region less than 0 as a river subregion (S320).

Next, the elevation calculation module 370 collects elevation information among the topographic information of the waterbody zone determined by the topography determination module in the database unit 400 and the zone corresponding to the river subregion (S330).

In the database unit 400, the elevation information for the paper site is stored in advance by using a conventional topographic survey method, and information of a geographic information system (GIS) may be retrieved. Since the topographic survey method uses a well-known method, a separate description is omitted.

Next, the elevation calculation module 370 uses the location information of each pixel of a plurality of points constituting the outline of the waterbody area determined by the topography determination module 360 to obtain the elevation information of the location corresponding to the location information. Using it, it is calculated as the elevation of the water surface (S340). Equation related to this is as Equation 12 below.

Next, the elevation calculation module 370 calculates this as the first river bed elevation by subtracting the estimated water depth from the elevation of the water surface calculated in step (S340) (S350).

FIG. 18 is an example of a result of dividing a waterbody region in which water exists as a result of applying Equation 11, and the waterbody region may be classified according to the presence or absence of water. For areas other than the waterbody area, the elevation of a digital elevation model measured by drones or directly is used, and for the area that is a waterbody area, the elevation of the water surface is calculated by using the elevations corresponding to the locations of multiple points constituting the outline of the waterbody area. The elevation of the riverbed is calculated by using the calculated elevation of the water surface and the water depth data for any point. Equation related to this is as Equation 13 below. The present invention can provide a depth observation method with high accuracy according to each zone, and for classification of the river bed elevation calculated using the observed water depth value, the river bed calculated using the water depth observation value estimated using the hyperspectral image The elevation was determined as the first riverbed elevation.

At this time, it is preferable that the elevation calculation module 370 calculates the first riverbed elevation as the riverbed elevation E(x,y,t) for the calculation in the riverbed fluctuation rate calculation module 380 .

19 is a diagram illustrating a method of calculating the elevation of a river bed by using a method for measuring depth using a drone-based hyperspectral image according to the present invention. The elevation of any point on the riverbed in the waterbody zone is calculated by subtracting the water depth of the point from the water surface elevation.

In another embodiment, the elevation calculation module 370 calculates a second river bed elevation using ultrasound information for a waterbody zone having a water depth exceeding a predetermined water depth among the waterbody zones by a preset method.

Preferably, the water depth predetermined in the topography determination module 360 is the maximum water depth that can be calculated through the hyperspectral image. This is because the maximum depth that can be estimated through hyperspectral images is limited, and there is a limit in measuring relatively high depths. Here, the river bed elevation calculated using the observation value of the water depth estimated using the ultrasound information was determined as the second river bed elevation.

Next, the water depth observation value estimated through the above steps may be calculated as shown in FIG. 20 according to a preset method (S350). The flow rate measurement procedure is expressed using the river bed elevation with a time difference. The elevation of the river bed expressed by two hyperspectral images does not change for a short time, and since the movement is very small, the elevation for a short time is calculated by the equation 14 can be defined as

Here, x is the X-direction coordinate, y is the Y-direction coordinate, t is the time, E is the elevation, δx is the amount of change in the x direction, δy is the amount of change in the y direction, and δt is the amount of time change.

Taylor expansion of the right term of Equation 14,

In order to simultaneously satisfy Equation (14) and Equation (15), the sum of the differential expressions of the right term of Equation (15) must be 0 as in Equation (16). By dividing both sides by the non-zero δt time, it can be converted into an equation composed of the x-direction and y-direction fluctuation rates as in Equation 17 (s350).

To this end, the elevation calculation module 370 will calculate the river elevation E(x,y,t) by subtracting the estimated water depth from the calculated elevation of the water surface.

Here, Ex is the X-direction differential matrix of elevation, Ey is the Y-direction differential matrix of elevation, Vx is the X-direction velocity of elevation, Vy is the Y-direction velocity of elevation, and Et is the elevation time differential matrix.

In order to obtain the lower-phase change rate through Equation 17, it is converted into a determinant form such as Equations 18 to 19 and solved. Since matrix A is a non-square matrix with different row and column sizes, the inverse of the non-square matrix is calculated using a pseudo inverse matrix.

here,

q _n is any point in the waterbody zone, and n is the number of points in the waterbody zone.

It will be possible to measure by calculating the lower-phase change speed in step S350 through the process of calculation as in the above equation.

In another embodiment, the elevation calculation module 370 calculates the river bed elevation by using ultrasound information for a depth exceeding a water depth predetermined by the topography determination module 360 among the estimated water depths. It is calculated by subtracting the water depth measured through the ultrasonic sensor from the calculated surface elevation.

Next, the elevation calculation module 370 collects elevation information among the topographical information of the area corresponding to the river subregion determined in step S320 in the database unit 400 .

Elevation information measured by using the topographical survey method is stored in the database unit 400 in advance, and information of a geographic information system (GIS) may be called.

The water depth predetermined in the topography determination module 360 is preferably the maximum water depth that can be calculated through the hyperspectral image. This is because there is a limit to the maximum water depth that can be calculated through the hyperspectral image, so there is a limit in measuring a relatively high depth. It is possible to calculate and provide all the fluctuation rates.

In the above, the present specification has been described with reference to the embodiments shown in the drawings so that those skilled in the art can easily understand and reproduce the present invention, but these are merely exemplary, and those skilled in the art may make various modifications and equivalent other modifications from the embodiments of the present invention. It will be appreciated that embodiments are possible. Therefore, the protection scope of the present invention should be defined by the claims.

100: unmanned aerial vehicle
101: echo sounder
110: gimbal
120: buffer
130: camera
140: hyperspectral sensor
150: communication department
200: image processing unit
210: communication module
220: flight path setting module
230: captured image input module
240: regularization performing module
250: geometric correction performing module
260: image registration performing module
261: integral image operation module
262: correspondence point confirmation module
270: hyperspectral image output module
300: image analysis unit
310: hyperspectral observation value input module
320: band ratio calculation module
330: regression analysis module
340: regression expression determination module
350: observation value estimation module
360: Terrain Determination Module
370: Elevation Estimation Module
380: lower phase variable speed calculation module
400: database unit

Claims (15)

(a) the unmanned aerial vehicle 100, while flying along a set flight path, using the hyperspectral sensor 140 to obtain hyperspectral information for each preset image unit through line scanning;
(b) performing, by the geometric correction performing module 250, geometric correction on the hyperspectral information obtained for each image unit by using the position information of each pixel;
(c) acquiring, by the image matching performing module 260, the hyperspectral image of the collection area by matching the hyperspectral information obtained for each image unit on which the geometric correction is performed;
(d) inputting a depth observation value (Y) of each of a plurality of predetermined points to the hyperspectral observation value input module 310;
(e) the band ratio calculation module 320, at each of the plurality of predetermined points, from the hyperspectral image obtained in step (c), a hyperspectral imaging value (I) that is a pixel value for a plurality of wavelengths (x) _{(x)), select a pair (λ 1} , λ ₂ ) of any two wavelengths among a plurality of wavelengths (x) included in the hyperspectral imaging value, and select the hyperspectral imaging value (I( determining λ ₁ ), I(λ ₂ ));
(f) the band ratio calculation module 320 _{divides any one of the two identified hyperspectral imaging values (I(λ 1} ), I(λ ₂ )) by the other to obtain a band ratio (X) ) to calculate; and
(g) the regression analysis module 330, assuming that the depth observation value (Y) and the calculated band ratio (X) have a linear relationship, and the depth observation value (Y) at the plurality of points Using the calculated band ratio (X), by calculating a and b of the regression equation (Y = aX + b), comprising the step of generating a regression equation (Y = aX + b),
(h) estimating, by the observation value estimation module 350, the depth of the water from the hyperspectral image using the regression equation generated in the step (g); and
(m) calculating, by the river bed fluctuation rate calculation module 380, the river bed fluctuation rate according to a preset method using the water depth estimated in step (h);
The step (m) is,
(m1) calculating, by the observation value estimation module 350, a normal vegetation index (NDWI) using the intensity of light with respect to a wavelength corresponding to near infrared in the hyperspectral image;
(m2) determining, by the terrain determination module 360, a waterbody zone when the calculated regular vegetation index (NDWI) is greater than or equal to 0, and determined as a river sub-region when it is less than 0 to determine each terrain;
(m3) The elevation calculation module 370 calculates the elevation corresponding to the location by using the location information of each pixel of a plurality of points constituting the outer line of the water body area among the topography determined by the topography determination module 360 estimating as the elevation of the water surface by utilizing;
(m4) calculating, by the elevation calculation module 370, the river elevation E(x,y,t) by subtracting the estimated water depth from the calculated elevation of the water surface;
(m5) assuming that the change in the river bed elevation calculated in step (m4) is 0 during the time change amount δt by the river bed fluctuation rate calculation module 380, calculating the river bed fluctuation rate through the following formula; containing,

,

Method of measuring river bed fluctuation rate.
According to claim 1,
The step (m5) is,
Including the step of calculating the rate of change of the river bed through the formula below,

here,

is,
q _n is any point in the waterbody zone,
Method of measuring river bed fluctuation rate.
According to claim 1,
After step (m3), before step (m4),
The step of the elevation calculation module 370, the database unit 400 collecting the elevation information of the area corresponding to the river sub-region determined in the step (m2); further comprising, the step (m5), Further comprising the step of calculating the river bed fluctuation rate for the river sub-region,
Method of measuring river bed fluctuation rate.
The method of claim 1,
After step (g),
(h) the regression analysis module 330 selects an arbitrary wavelength other than the pair of two wavelengths in step (e) from among the plurality of wavelengths included in the hyperspectral imaging value, generating a regression equation for each of a plurality of pairs of two wavelengths by verifying a spectroscopic image and repeating steps (f) to (g);
(i) selecting, by the regression equation determination module 340, a regression equation having the highest correlation coefficient among the generated plurality of regression equations;
(j) determining, by the regression equation determination module 340, the two wavelengths applied to the band ratio (X) used in the regression equation selected in the step (g) as a pair of optimal band ratio wavelengths; and
Estimating the water depth from the hyperspectral image using the selected regression equation,
Method of measuring river bed fluctuation rate.
5. The method of claim 4,
After step (j),
(k) the observation value estimation module 350 identifies two hyperspectral imaging values corresponding to the pair of optimal band ratio wavelengths determined in step (j) from the hyperspectral imaging values of points other than the plurality of points and dividing one by another to calculate a band ratio of the corresponding point; and
(l) the observation value estimating module 350, inputting the calculated band ratio into the regression equation selected in the step (g), further comprising the step of estimating the observation value of the depth of the point,
Method of measuring river bed fluctuation rate.
6. The method of claim 5,
The step (l) is,
calculating, by the observation value estimation module 350, a normal expression index (NDWI) using the intensity of light for a wavelength corresponding to near infrared in the hyperspectral image; and
The method further comprising the step of outputting, by the observation value estimation module 350, the depth of water as 0 when the calculated regular vegetation index (NDWI) is less than 0,
Method of measuring river bed fluctuation rate.
7. The method of claim 6,
The regular expression index (NDWI) is calculated by the following formula,

Here, NIR is the intensity of light at a wavelength corresponding to near infrared among the values included in the hyperspectral information, and Green is the intensity of light at a wavelength corresponding to green among the values included in the hyperspectral information,
Method of measuring river bed fluctuation rate.
6. The method of claim 5,
Step (i) is,
generating, by the regression equation determination module 340, correlation coefficients for the generated plurality of regression equations, and generating a correlation coefficient map in which each wavelength is an X-axis and a Y-axis in a pair of two wavelengths; including,
Step (i) is,
The regression formula determination module 340 determines a point having the highest correlation coefficient using the correlation coefficient map generated in step (i), and sets each wavelength constituting the point to the pair of optimal band ratio wavelengths. comprising the step of determining
Method of measuring river bed fluctuation rate.
The method of claim 1,
The step (c) is,
(c11) In any two hyperspectral information (A, B) obtained by the corresponding point confirmation module 262 for each image unit on which the geometric correction is performed, any one of the hyperspectral information (A) With respect to the XY coordinates (x, y) of the pixel and the pixel value (I(x, y)) at the corresponding coordinates, the XY coordinates (x', y) of any one pixel of the other hyperspectral information (B) ') and a normalized correlation coefficient of the pixel value I(x', y') at the corresponding coordinates, but for each pixel of the one hyperspectral information A, the other checking the XY coordinates of a point having a large value of the normal correlation coefficient as a corresponding point of the two hyperspectral information (A, B) by calculating all the normal correlation coefficients of each pixel of the hyperspectral information (B);
(c12) image registration of the two hyperspectral information (A, B) based on the identified corresponding point by the image matching performing module 260; and
(c13) the image registration performing module 260 performs the steps (c11) to (c12) for each pair of hyperspectral information for each image unit obtained in the step (a), Comprising the step of obtaining a hyperspectral image of the collection area,
Method of measuring river bed fluctuation rate.
10. The method of claim 9,
Before step (c11),
(c21) the integral image operation module 261, in the hyperspectral information obtained for each image unit on which the geometric correction is performed, for each pixel, an integral image based on any one reference coordinate Further comprising the step of calculating the pixel value,
The pixel value in step (c11) is the pixel value of the integral image calculated in step (c21),
Method of measuring river bed fluctuation rate.
11. The method of claim 10,
Before step (c21),
(c31) the normalization module 240 sets the XY coordinates of the pixels of the hyperspectral information of any one image unit as an ordered pair of integers (x, y), calculates the average coordinates of the set XY coordinates, and the operation The XY coordinates (x, y) are normalized to the XY coordinates (x', y) so that the unit distance between the ordered pairs (x, y) of each integer is √2 by setting the average coordinates as the origin (0, 0). '), further comprising the step of normalizing the hyperspectral information of any one image unit by moving to,
Method of measuring river bed fluctuation rate.
12. The method of claim 11,
After step (c31),
(c32) the normalization module 240 uses the values of the XY coordinates (x, y) before the normalization of any one of the hyperspectral information and the XY coordinates (x', y') after the normalization, calculating a homography (H) which is a determinant including; and
(c33) using the calculated homography (H), comprising the step of normalizing another hyperspectral information,
Method of measuring river bed fluctuation rate.
13. The method of claim 12,
In step (c32), homogeneous coordinates are used so that the XY coordinates before normalization are (x, y, 1) and are referred to as determinant A, and the XY coordinates after normalization are (x', y', 1) It is referred to as the determinant A', and the homography (H) is calculated using the following formula but using a number of XY coordinates,

In the step (c33), the determinant A', which is the normalized XY coordinates, is calculated by substituting another hyperspectral information into the determinant A,
Method of measuring river bed fluctuation rate.
stored on a computer-readable recording medium.
14. A program for executing a method according to any one of claims 1 to 13.
14. A computer-readable recording medium on which a program code for executing the method according to any one of claims 1 to 13 is recorded.

Images