RU2622899C1 - Radiometer hardware function determination method - Google Patents

Abstract

FIELD: radio engineering, communication.

SUBSTANCE: method is carried out by placing the control object radiometer antenna in the viewing area, the object scanning by the radiometer antenna along the azimuth and the elevation angle, forming the radiometric and optical images of the area, containing the control object with the adjacent background; Y and X matrix formation, respectively, the radiometric and optical imaging, X matrix segmentation by the amplitudes contrast, X matrix representation as the reference radiometric image of the control object, and the subsequent mathematical Y and X matrices processing to obtain the A matrix, which is the matrix representation of the radiometer hardware function.

EFFECT: definition of the radiometer hardware function correct value in conditions of its operation to ensure the possibility of obtaining the radio heat images of the observable objects.

2 cl, 6 dwg

Description

The invention relates to radiolocation, namely to passive systems for monitoring objects using a scanning radiometer, and can be used to obtain a thermal image of various objects.

At present, passive radio-vision systems operating in the field of millimeter and submillimeter wavelengths are increasingly used for microwave monitoring of various objects and the environment (precipitation, aerosol pollution of the atmosphere, properties of underlying surfaces, etc.). But in order to use the radiometer in radio-vision systems and to obtain (synthesize) the radio thermal image of the observed object, it is necessary to know its hardware function with a high degree of accuracy.

As the closest analogue of the proposed method, a method has been adopted for determining the hardware functions of the radiometer, which consists in placing the control object in the viewing area of the radiometer antenna, scanning the control object with the radiometer antenna in azimuth and elevation, and generating a radiometric image of the control object as a result of scanning matrix Y followed by mathematical processing of the matrix Y. The control object is a point (reference) source, characterized by contrast with respect to Rouge the background and allows you to obtain a reference image as a delta function. The hardware function is obtained by subtracting from the image of the observed scene the image of the background scene formed in the radiometer in the absence of a control object [1].

The disadvantage of this method is as follows. In real conditions of observing objects using a radiometer, it is difficult to obtain a reference image of the control object in the form of a delta function, therefore, in the known method, the measurement of the hardware function is carried out in laboratory conditions. However, under actual operating conditions of the radiometer, in particular, when determining the thermal thermal contrast of extended objects, the value (value) of the hardware function of the radiometer will differ from its value measured in laboratory conditions due to blurring of the shape of the hardware function. As a result, it is practically impossible to obtain a true thermal image of the observed objects.

The technical result of the claimed invention consists in determining the correct value of the hardware function of the radiometer in the conditions of its operation in order to ensure the possibility of obtaining a thermal thermal image of the observed objects.

, умножения вектора

справа на матрицу Х⁺, формирования вектора

и построчной записи элементов вектора

в матрицу А, являющуюся матричным представлением аппаратной функции радиометра.The specified technical result is achieved by the fact that in the method for determining the hardware function of the radiometer, which consists in placing a control object in the field of view of the antenna of the radiometer, characterized by contrast with respect to the surrounding background, scanning the control object with the antenna of the radiometer in azimuth and elevation, forming the matrix Y of the radiometric image of the control object and finding the hardware function of the radiometer, use an arbitrary object as a control object and form radiometers eskoe image region containing the object reference with attached background; at the same time, an optical image of the said region is additionally obtained, an optical image matrix X of the said region is formed with a control object, the matrix X is adjusted to the scale of the matrix Y, the matrix X is segmented by the contrast of the amplitudes of neighboring segments, the average normalized radiometric amplitude is calculated for each segment found and the segmented matrix X for the reference radiometric image of the control object, and the procedure for finding the hardware function of the radiometer is carried out appear by converting the matrix X into a pseudoinverse matrix X ⁺ , and rewriting the matrix Y into a vector

multiplying the vector

to the right of the matrix X ⁺ , the formation of the vector

and line recording of vector elements

into matrix A, which is a matrix representation of the hardware function of the radiometer.

The specified technical result is also achieved by the fact that obtaining an optical image of the region containing the control object is carried out using a camera.

The invention is illustrated by drawings. In FIG. 1-3 show the steps of determining the hardware function, in FIG. Figure 4-6 illustrates radiometric imaging using a hardware feature found.

,

, где М и N - число строк и столбцов матрицы Y, при этом изображение контрольного объекта представляет собой односвязную и однородную по амплитуде подобласть G⊂D, контрастную по отношению к прилегающему к контрольному объекту однородному фону.The inventive method is implemented as follows. An arbitrary control object, which is characterized by a certain contrast with respect to the background adjacent to it, is placed in the field of view of the antenna of the radiometer. As a result of scanning the antenna of the radiometer in azimuth and elevation of the area D of the location of the control object, a matrix Y of the radiometric image with elements y (i, j) is formed,

,

, where M and N are the number of rows and columns of the matrix Y, while the image of the control object is a simply connected and uniform in amplitude subregion G⊂D, contrasting with the uniform background adjacent to the control object.

,

, имеющими смысл интенсивности оптического излучения в i-м, j-м угловом направлении, где M_x=k⋅М, N_x=k⋅N, k - масштабный множитель (целое число). Матрица X приводится в соответствие масштабу матрицы Y пересчетом значений ее элементов по формуле (1):Using a camera located in the immediate vicinity of the radiometer, an optical image of the region containing the control object with an adjacent background is recorded, and an optical image matrix X of the region D with elements x (i, j) is formed,

,

that have the meaning of the intensity of optical radiation in the ith, jth angular direction, where M _x = k xM, N _x = k⋅N, k is the scale factor (integer). Matrix X is brought into correspondence with the scale of matrix Y by recalculation of the values of its elements according to formula (1):

.and storing the received elements x ₁ (i, j) in the matrix X: x (i, j) = x ₁ (i, j),

.

, где S - номер сегмента, которому принадлежит i-й, j-й элемент матрицы X. Далее для каждого s-го сегмента матрицы X вычисляется средняя нормированная радиометрическая амплитуда

усреднением амплитуд соответствующих i-x, j-x элементов y(i,j) матрицы Y с теми же метками S(i,j) по формуле (2):The resulting matrix X is segmented in accordance with known segmentation algorithms, for example [2], by the contrast of the amplitudes of the elements of neighboring segments. The result of segmentation operations is the label matrix {S (i, j)},

, where S is the number of the segment to which the i-th, j-th element of the matrix X belongs. Next, for each s-th segment of the matrix X, the average normalized radiometric amplitude is calculated

by averaging the amplitudes of the corresponding ix, jx elements y (i, j) of the matrix Y with the same labels S (i, j) according to formula (2):

where n _s is the number of elements labeled s; μ is a normalizing factor that takes into account the integral nature of the observations.

присваивается элементам матрицы X с меткой

. Полученная матрица X принимается за эталонное радиометрическое изображение контрольного объекта с четко выраженными по контрасту границами сегментов.Amplitude

assigned to elements of matrix X with label

. The resulting matrix X is taken as the reference radiometric image of the control object with clearly defined contrast boundaries of the segments.

The procedure for finding the hardware function of the radiometer is as follows. The elements of the matrix X are rewritten in the matrix X ₁ corresponding to the radiometric image model according to the formula (3):

- вектор измерений y(i,j), считанных построчно из матрицы Y={y(i,j)}; Х₁ - матрица элементов x(i,j), расположенных в соответствии с (1);

- вектор искомых значений аппаратной функции;

- вектор помех.Where

- vector of measurements y (i, j), read line by line from the matrix Y = {y (i, j)}; X ₁ - matrix of elements x (i, j) located in accordance with (1);

is the vector of the desired values of the hardware function;

is the interference vector.

Then, for the matrix X _{1, a} pseudoinverse matrix X ^{+ is found} by the formula (4):

. Матрица Х⁺ также может быть найдена сингулярным разложением Х₁, например, в среде Matlab: Х⁺=pinv(X₁,δ).where T is the transpose symbol; E is the identity matrix; δ - regularization parameter - small positive number necessary for stable inversion of the matrix

. The matrix X ⁺ can also be found by the singular decomposition of X ₁ , for example, in the Matlab environment: X ⁺ = pinv (X ₁ , δ).

, который затем умножается справа на матрицу Х⁺ и в результате получается вектор

.In the next step, the matrix Y of the radiometric image is rewritten line by line into a vector

, which is then multiplied on the right by the matrix X ⁺ and as a result we get a vector

.

построчно записываются в матрицу A={α(i,j)}, являющуюся матричным представлением аппаратной функции радиометра. Найденная аппаратная функция радиометра используются в дальнейшем для получения радиотеплового изображения наблюдаемых объектов в зоне обзора радиометра.Vector elements

are written row-wise into the matrix A = {α (i, j)}, which is the matrix representation of the hardware function of the radiometer. The found hardware function of the radiometer is used in the future to obtain a thermal thermal image of the observed objects in the field of view of the radiometer.

The inventive method was tested on a model experiment. The image of the control object in matrix X was modeled in the form of a square with an amplitude U ₀ = 10 against an external background with an amplitude U _f = 0. The matrix Y of the radiometric image (Fig. 1) was obtained by scanning the matrix X with an antenna pattern (BOTTOM) of size 7 × 7 with the standard deviation of interference σ _p = 0.05. The hardware function was modeled by an exponential with a quadratic exponent. In the Y matrix, a subregion of size M × N = 23 × 23 with a blurred image of the control object was distinguished.

A simulated image X of a control object with an external background in a matrix of size M × N = 23 × 23 was considered as an optical image. As a result of segmentation by the threshold method, two segments — an object and a background — were allocated on the optical image X, and the amplitude was assigned to the elements of each segment. The obtained reference radiometric image (Fig. 2) was used to find the hardware function.

Next, the found hardware function of the radiometer (Fig. 3) was used to determine a more complex image of an X object in the form of two segments - a 9 × 9 frame with a width of 2 elements with an amplitude of U ₀ = 10 and the inside of the frame in the form of a 5 × 5 square with an amplitude of U ₁ = 5 and external background with amplitude U _f = 0 - FIG. 4. The matrix Y was modeled for a 7 × 7 DND with a standard deviation of interference σ _p = 0.01. When finding the image X, the matrix method was used [3].

In FIG. 5 shows a radiometric image of a simulated object; FIG. 6 - the found image of the simulated object obtained using the previously determined hardware function of the radiometer.

The proposed method allows to determine the hardware function of the radiometer in real conditions of its operation due to the formation of a reference image using an optical image of an arbitrary control object. This approach eliminates the ambiguity that occurs when using the hardware function of the radiometer, found using the reference object in conditions that differ from the actual conditions of its operation.

Using the proposed method due to the correct determination of the hardware function of the radiometer contributes to more efficient functioning of the existing radiometric systems for obtaining thermal images of various objects and the environment.

BIBLIOGRAPHY

1. Methodology for measuring the hardware function of a passive radio-vision system / www.backstage.narod.ru/education/diplom98.pdf.

2. Gonzalez R., Woods R., Eddins S. Digital image processing in MATLAB. M .: Technosphere, 2006.616 s.

3. Vasilenko G.I., Taratin A.M. Image recovery. M .: Radio and communication, 1986. 304 p.

Claims (2)

1. The method of determining the hardware function of the radiometer, which consists in placing in the field of view of the antenna of the radiometer a control object characterized by contrast with respect to the surrounding background, scanning the control object with the radiometer antenna in azimuth and elevation, forming the matrix Y of the radiometric image of the control object and finding the hardware function of the radiometer , characterized in that as a control object using an arbitrary object and form a radiometric image of the area, contains living control object with an adjacent background; at the same time, an optical image of the said region is additionally obtained, an optical image matrix X of the said region is formed with a control object, the matrix X is adjusted to the scale of the matrix Y, the matrix X is segmented by the contrast of the amplitudes of neighboring segments, the average normalized radiometric amplitude is calculated for each segment found and the segmented matrix X for the reference radiometric image of the control object, and the procedure for finding the hardware function of the radiometer is carried out appear by converting the matrix X into a pseudoinverse matrix X ⁺ , and rewriting the matrix Y into a vector

multiplying the vector

to the right of the matrix X ⁺ , the formation of the vector

and line recording of vector elements

into matrix A, which is a matrix representation of the hardware function of the radiometer. 2. The method according to p. 1, characterized in that the optical image of the area containing the control object is carried out using a camera.

Images