RU2767606C1 - Method for processing and displaying signals of multispectral images - Google Patents

Abstract

FIELD: imaging technology.

SUBSTANCE: invention relates to the field of multispectral television, using the registration of reflected and radiated flux in several spectral regions (zones), including the ultraviolet, visible and infrared regions of the optical spectrum. It can be used to solve problems of selection, discrimination and recognition of objects by amplitude, spectral and spatial characteristics and find applications in systems of technical vision, video surveillance and object tracking. The result is achieved in that a method for forming and display color, multispectral and thermal images, enables the logging of input of radiant flux F(λ) within a wide spectral range, the formation of two separate radiant F₁(λ) and F₂ (λ) streams reflecting areas (zones) of the registration of the radiant flux Δλ₁ and Δλ₂, the conversion of radiant fluxes F₁(λ) and F₂(λ) and the formation on their basis in the first channel of image signals U_Δλ1(t), and in the second channel of image signals U_Δλ2(t), separate processing of the received signals of multispectral images U_Δλ1(t) and U_Δλ2(t), including amplification, conversion of analog signals into digital, digital aperture and gamma correction, when performing joint signal processing of multispectral images U_Δλ1(t) and U_Δλ2(t), including operations of summation and subtraction of two source signals, adding operations of their division and multiplication and in general form m-signals of multispectral images, which are fed to the inputs of a multichannel switch, further, any three signals are selected from all the signals and fed to the RGB inputs of a color video monitoring device for displaying and visual analysis of multispectral images in conditional colors, and also using these signals of multispectral images for automatic image analysis.

EFFECT: expanding the functionality and achieving an increase in the contrast of images for the selection and distinction of objects in the joint signal processing of multispectral images using two spectral areas (zones) of the registration of the radiant flux.

1 cl, 3 dwg, 6 tbl

Description

The present invention relates to the field of multispectral television, which uses the registration of the reflected or radiated flow in several spectral regions (zones), including the ultraviolet, visible and infrared regions of the electromagnetic wave spectrum (Let's define such a region of the electromagnetic wave spectrum as a wide spectral range of wavelengths).

The claimed method can be used to solve the problems of selection, discrimination and recognition of objects by the distribution of the amplitude and spectral characteristics of objects and find application in systems of technical vision, remote video surveillance and object tracking.

To observe objects of the surrounding space and the earth's surface from aircraft (LA), various types of optoelectronic, television (TV) and thermal imaging (TPV) systems are used, which record the radiant (light) flux in a selected spectral region.

For the formation of black-and-white and color TV images, registration of the light flux in the visible region of the spectrum is used. In the transmitting path of the TV system, image signals are formed and processed, and in the receiving path they are displayed on the screen of video monitoring devices (VCU) in the form of black-and-white or color images.

The principle of spectral selection of objects is based on the simultaneous or sequential registration of the reflected or emitted radiant flux in several wavelength zones - Δλ _i - within a wide spectral range of wavelengths. Here i=1, ..., m, where m is the number of wavelength zones, depending on the problem of observation and recognition of objects being solved. In this regard, spectral-zonal (multi-spectral) TV images can be formed by recording and converting a radiant (light) flux in the ultraviolet (UV), visible (VI) and infrared (IR) regions of the spectrum. The information content of such images can be much higher (tens of times) compared to color RGB images, especially when distinguishing objects of the earth's surface that have the same spatial features (in shape, size, etc.).

Methods and devices for separate formation of video signals using separate black-and-white, color, spectral-zonal TV cameras in different spectral regions (zones) for visual analysis of images of objects are reflected in a large number of sources of domestic literature [1], [2], [3] and others

The transparency of the environment between the aircraft and the objects of the earth's surface may deteriorate due to adverse weather conditions: rain, fog or smoky atmosphere, which leads to the observation of low-contrast objects or their "complete loss" when recognizing in the image. This requires, along with the use of VI and near-IR spectral regions, to carry out observations of objects in the thermal IR region of the spectrum.

Today, there is a large class of individual thermal imaging devices in which matrix photodetectors (MPDs) operating in broadcast TV format are used to register the radiant flux from objects, which makes such systems in some cases indispensable for observing objects at night, even in black and white. white form, which is carried out by recording and converting the radiant flux in the thermal part of the IR region of the spectrum in the spectral regions of 3–5 µm and 8–12 µm into image signals [4].

There is a method [5], in which the integral registration method is used for the input radiant flux and its conversion into image signals with the transition to the differential method at the level of processing the zonal TV image signals themselves. In this case, zones of registration of a radiant flux with a variable width in a wide spectral region of wavelengths are used and zonal TV image signals are formed by subtracting signals from each other. The use of registration of a radiant flux with a variable width of wavelength zones makes it possible to form zonal signals with a narrow zone of registration of a radiant flux at a high signal-to-noise ratio of subtracted signals at the level of electrical signal processing.

The disadvantage of the considered method is the impossibility of simultaneous separate formation of video signals of color, spectral-zonal and TPV images and combining the advantages of methods for registering a radiant flux in different spectral regions of wavelengths.

Recently, systems incorporating TV and TPV cameras have become widespread. TV cameras are designed to receive images of the observed scene in the VI region of the spectrum (0.38-0.76 microns), and TPV cameras are designed to receive images on the monitor screen in the IR region of the spectrum (3-5 microns) or (8-12 microns ). Useful information necessary for the operator to make a decision can be distributed between images of different parts of the spectrum. In this case, the operator is forced to analyze several images and compare them with each other, which leads to delays in making a decision.

For this reason, it is advisable to output to the display device a combined image synthesized from the original TV and TPV images. This mode allows you to increase the information content of the system by combining in one resulting image the details of the observed scene in the VI and IR regions of the spectrum.

Currently, there is no unified algorithm for combining TV and TPV images. To a large extent, this is due to the fact that the assessment of the quality of the presented images is subjective and depends on the psychophysiological characteristics of the operator, the task being solved, and the conditions of observation. But, nevertheless, the algorithm for processing multi-spectral images, in order to combine them, can be based on the criterion for achieving maximum information content, which provides for obtaining more different information on spectral-energy features.

In this regard, from the point of view of the joint processing of signals of multispectral images, methods and devices are known that combine two or more images obtained in different sections and zones of the optical spectrum (for example, visible and thermal). They make it possible to form the resulting image, which has a greater distinctiveness for the selection of given objects. First of all, signal processing methods, including various arithmetic operations [5], can be used for this.

As the closest analogue of the claimed invention in terms of the totality of features and operations on signals, a method for generating and displaying signals of color, spectral-zonal and thermal imaging images was adopted (RF Patent No. 2546982, - publ. 04/10/2015) [7].

The essence of the method is as follows.

For the formation and display of signals of color, spectral-zonal and thermal imaging, registration of a radiant flux within a wide spectral range of wavelengths, its splitting into two identical fluxes is provided. After splitting the input radiant flux into two identical streams, they are passed through two optical filters, the first of which has a spectral characteristic covering the spectral region in the visible region of the spectrum from 0.38 to 0.76 μm and at the output of the first filter form a radiant flux F ₁ (λ), and the spectral characteristic of the second optical filter covers the spectral region in the visible and near IR regions of the spectrum from 0.38 to 2.5 μm and at the output of the second optical filter form a radiant flux F ₂ (λ).

Further, the first radiant flux F ₁ (λ) is projected onto the first multi-signal "radiant flux-signal" converter, which has mosaic color filters of the form RGB on the surface of its photodetector, corresponding to the light flux registration zones in red (R), green (G) and blue (B) the visible region of the spectrum, then the light fluxes F _R (λ), F _G (λ) and F _B (λ) are converted into color television video signals U _R (t), U _G (t) and U _B (t) .

The second radiant flux F ₂ (λ) is projected onto the second multi-signal "radiant flux-signal" converter, which has mosaic spectral-zonal optical filters on the surface of the photodetector with its own spectral characteristic, corresponding to the selected zones for recording wavelengths Δλ ₁ , Δλ ₂ and Δλ _{3 of} the radiant flux in visible and near-IR spectral regions, where, after spectral-zonal optical filters, radiant fluxes F(Δλ ₁ ), F(Δλ ₂ ) and F(Δλ ₃ ) are converted into video signals of spectral-zonal television U _Δλ1 (t), U _Δλ2 (t) and U _Δλ3 (t).

According to this method, a third channel is additionally organized within a wide spectral range of wavelengths, for which the input radiant flux F(λ) is passed through an infrared lens, the spectral characteristic of which covers the thermal region of the IR region of the spectrum, and a radiant flux F ₃ (λ) is formed, which project onto the third two-signal transducer "radiant flow-signal" and convert it into thermal imaging video signals U _Δλ4 (t) and U _Δλ5 (t), corresponding to the zones of registration of wavelengths Δλ ₄ (3-5 μm) and Δλ ₅ (8-12 μm ) of the thermal sections of the spectral regions, respectively.

и

спектрозонального телевидения

и

тепловидения

и

используют для их совместной обработки путем использования операций вычитания или суммирования видеосигналов между собой и др.The resulting groups of image signals are amplified, analog signals are converted into digital ones, digital aperture and gamma correction and other types of image signal processing are performed to improve image quality, then digital color television image signals

And

spectrum television

And

thermal imaging

And

are used for their joint processing by using the operations of subtraction or summation of video signals among themselves, etc.

The disadvantage of the considered method, firstly, is that when performing the operations of joint processing of image signals of color and spectral-zonal television and thermal imaging, the resulting (combined) image signals are formed based on the use of only two operations - summation and subtraction of video signals between themselves, which imposes certain restrictions to ensure the possibility of achieving maximum discrimination of objects for any combination of their spectral characteristics.

The technical result of the claimed invention is the expansion of functionality and the achievement of an increase in image contrast for selection and discrimination of objects in the joint processing of signals from multispectral images using two spectral sections (zones) for registering a radiant flux.

The technical result is achieved by the fact that, in contrast to the known method of forming and displaying color, spectral-zonal and thermal imaging images, including registration of the input radiant flux F(λ) within a wide spectral wavelength range, the formation of two separate radiant fluxes F ₁ (λ) and F ₂ (λ), reflecting the areas (zones) of registration of the radiant flux for the corresponding intervals of wavelengths Δλ ₁ and Δλ ₂ , the conversion of radiant fluxes F ₁ (λ) and F ₂ (λ) and the formation on their basis in the first channel of image signals U _Δλ1 (t), and in the second channel of image signals U _Δλ2 (t), separate processing of the received signals of multispectral images U _Δλ1 (t) and U _Δλ2 (t), including amplification, conversion of analog signals to digital, digital aperture and gamma correction and other types of signal processing aimed at improving the quality of images, for which, when performing joint signal processing of multispectral images U _Δλ1 (t) and U _Δλ2 (t), which includes operations of summation and subtraction of two original signals, in the proposed method, the operations of their division and multiplication are added and, in general, the following signals of multispectral images are formed

after which, for the newly formed signals U ₁ (t), U ₂ (t), U ₃ (t) and U ₄ (t), the operation of dividing the two signals between themselves is applied and additional image signals of the form are formed

then two original signals of multispectral images U _Δλ1 (t) and U _Δλ2 (t), as well as m-shaped signals (for example, n=1,…, 10, see expressions (1) and (2), subscript U _m (t) on the right side) multispectral images are fed to the inputs of a multichannel switch, then any three signals are selected from all signals and fed to the RGB inputs of a color video control device for displaying and visually analyzing multispectral images in conditional colors, and also using (m + 2) - multispectral image signals for automatic image analysis.

The involvement of signal processing operations, including division and multiplication of signals among themselves, allows you to expand the potential for selecting objects with low contrast, by increasing the number of generated signals (1) and additionally generated signals, according to expression (2). In this case, the formation of additional signals can also be carried out in the form

If separate processing of multispectral image signals provides for improving the quality of a separate multispectral image, such as increasing sharpness, image contrast and other characteristics and parameters, then the proposed joint processing of multispectral image signals allows you to:

- formation of resulting images from two original signals of multi-spectral images, in which the best distinguishing features in terms of contrast and signal-to-noise ratio in individual images are preserved;

- identification of objects by amplitude features, by analyzing their distribution in m-formed image signals instead of the original two.

The addition of signal processing operations, including division and multiplication of signals, allows for the expansion of functionality and the achievement of an increase in image contrast for selection and discrimination of objects in the joint processing of signals from multi-spectral images using two spectral regions (zones) of registration of the radiant flux.

In FIG. 1 shows a block diagram of a device for processing and displaying multispectral image signals.

In FIG. Figure 2 shows the spectral-energy characteristics of an arbitrary class of objects.

In FIG. Figure 3 shows a view of multispectral black-and-white images of the observed objects for these two registration zones.

The positions in FIG. one:

1 - spectral-zonal (two-spectral) TV camera, which generates image signals obtained by registering a radiant flux in two different sections (zones) of the optical spectrum;

2 ₁ and 2 ₂ - blocks of separate processing of TV image signals;

3 ₁ , 3 ₂ , 3 ₃ and 3 ₄ - blocks of summation, subtraction, division and multiplication of two-spectral TV image signals;

4 ₁ , 4 ₂ , …, 4 ₆ - blocks for dividing TV image signals obtained from the output of blocks 3 ₁ , 3 ₂ , 3 ₃ and 3 ₄ ;

5 - multi-channel switch of TV image signals;

6 - color video control device (VKU);

7 - block for automatic analysis of TV image signals;

8 - control panel;

9 - operator.

The principle of operation of the device that implements the proposed method for processing and displaying multispectral image signals is as follows. Let objects of some classes be in the field of view of the spectral-zonal (two-spectral) TV camera 1, for example, with spectral-energy characteristics as shown in Fig. 2.

The registration of the radiant flux (in Fig. 2 its values are normalized by dividing by the maximum value ρ(λ)=F(λ)/max(F(λ) and plotted along the y-axis) is carried out in a wide spectral range of wavelengths in two zones (sections) registration of wavelengths Δλ ₁ and Δλ ₂ (The wavelength is conditionally shown along the abscissa, the boundaries of the registration zones in Fig. 2 are indicated by vertical dashed lines).Further, it is assumed that the spectral-energy characteristics of objects (curves in Fig. 2) are some set of object classes

where A, ..., D - a specific class of objects.

After registration of the radiant flux in the zones (sections) of registration Δλ ₁ and Δλ ₂ , its transformation using matrix photodetectors, signals of multispectral TV images are formed in the form

U _∆λ1 (t) and U _∆λ2 (t).

Conditional visual images of a set of N object classes for these registration zones (areas) are shown in FIG. 3. In FIG. 3, the darker color of the circle conventionally depicts a class of objects with a lower value of the registered radiant flux in the corresponding registration zone Δλ ₁ , or Δλ ₂ . For the first registration zone Δλ ₁ the amplitude of the image signal is minimal for an object of class A and maximum for an object of class D. For the registration zone Δλ ₂ the situation will be reversed. For objects of class B and C, the amplitude of the image signal will be approximately the same for the registration zones Δλ ₁ and Δλ ₂ .

It should be noted that the two multispectral image signals used at the output of the two-spectral TV camera 1, for their joint processing, can be obtained based on the registration of radiant fluxes F ₁ (λ) and F ₂ (λ) from separate registration zones Δλ ₁ and Δλ ₂ . At the same time, the registration zones Δλ ₁ and Δλ ₂ can be located within the same spectral region, including the UV, VI or IR regions of the spectrum, or be in two different spectral regions, for example, there may be options in which the registration zone Δλ ₁ is located in the VI region of the spectrum , and the registration zone Δλ ₂ is in the near IR region of the spectrum, in another case, the registration zone Δλ ₂ may be in the mid-IR region of the spectrum, etc.

As an example of joint processing and display of signals of multispectral TV images, consider the option when part of the operations is used (Fig. 1), for example, including the summation of two signals (block 3 ₁ ), their subtraction (block 3 ₂ ) and division (block 3 ₃ ) , and then over the signals, after performing the summation operation (block 3 ₁ ) and their subtraction (block 3 ₂ ), they are divided among themselves (block 4 ₁ ). It should be noted that other variants of signal processing among themselves can be selected, according to the block diagram of the device shown in Fig. 1. The value of the resulting signal U ₁ (t) after the operation of summing the two original signals (block 3 ₁ ) will be the value

and the value of the resulting signal U ₂ (t) after the operation of subtracting the two original signals (block 3 ₂ ) will be the value

The value of the resulting signal U ₃ (t) after performing the operation of dividing the two original signals (block 3 ₃ ) will be the value

The value of the resulting signal U ₅ (t) (see expression (2)), after performing the operation of summation (block 3 ₁ ), subtraction (block 3 ₂ ) and division of two signals (block 4 ₁ ) will be the value

Here, in expressions (4)-(7), in comparison with expressions (1) and (2), for generalization, normalizing coefficients are added before each signal. The values of the normalizing coefficients can be distributed in the region 0<α_i, β_i, v_i, ξ_i≤1, where in the case under consideration i=1, 2. Let us assume that their values are equal to α_i, β_i, v_i, ξ_i=1 (expressions (4)-(7) will take the form similar to those in expressions (1) and (2)) and the amplitude of the signals after converting the radiant flux of objects and performing the operations of summing, subtracting and dividing the signals will have the values, as shown in tab. 1, based on the distribution of their spectral-energy characteristics in the selected spectral regions (Fig. 2).

The amplitude value of the signals in Table 1 is indicated based on the fact that, for clarity, the values of the initial signals U _Δλ1 (t) and U _Δλ2 (t) were first normalized similarly to the normalization of the radiant flux (see above), and then averaged in the corresponding zones (sections) registration of wavelengths Δλ ₁ and Δλ ₂ . The values U _Δλ1 (t) and U _Δλ2 (t) in Table 1 are taken approximately from FIG. 2 - this is enough for a qualitative description of the proposed technical solution.

An important characteristic of the visual perception of objects in optical images is the contrast sensitivity of the eye - the ability to distinguish some objects against the background of others, due to their unequal brightness, which is determined by the reflective or radiative properties of the observed objects in a given spectral region [3]. Usually, the object of interest (selection object) is observed against the background of another object. The difference between the object of observation and the background will be recorded by the human visual analyzer, if the contrast 8 (the ratio of the brightness of the object of observation and the background) exceeds a certain threshold value, we assume that the excess should be in the range 0.01≤ε≤0.02, depending on the visual analyzer of a particular person.

Selection and discrimination of one object against the background of another is possible if the contrast between them is higher than the minimum allowable contrast. Thus, when observing objects in images, an important place is occupied by the minimum value of the brightness increment, which can be detected by a human visual analyzer. Consider the value of contrast when an object of one class is located against the background of another class, for example, an object of class D against the background of objects of classes A, B and C, etc. based on the amplitudes of the signal values from Table 1. We will calculate the contrast for the signals received in the detection zones Δλ ₁ and Δλ ₂ , as well as for the resulting signals U ₁ (t), U ₂ (t), U ₃ (t) and U ₅ (t).

In section 2 of paragraph 2.21 of GOST R 52870-2007 [8], the calculation formula for determining the contrast C is given:

where L _screen - the brightness of the image element of the working screen;

L _f is the brightness of the background radiation.

In our case, the analogue of brightness is obviously the amplitude values of the above-mentioned signals C, and when calculating the contrast (we denote this value by k) of an object of one class against the background of an object of another class, the formula will take the form:

In (9), the modulus of the difference in the amplitude values of the signals in the numerator is used due to the fact that in the considered example of joint processing and display of signals, the value of the difference in their amplitudes obtained in the registration zones Δλ ₁ and Δλ ₂ , as well as for the resulting signals U ₁ (t ), U ₂ (t), U ₃ (t) and U ₅ (t) when an object of one class is located against the background of another class, may be of a different sign, as a result of their spectral-energy characteristics in the selected spectral regions (Fig. 2) and values (Table 1). For the same reason, the maximum value of the signal amplitude is selected in the denominator from the amplitudes of the signal values of both classes of objects for which the contrast is calculated.

In turn, the amplitude values of the signals U _i (t) and U _j (t) are selected from Table 1. In particular, to calculate the contrast of class D objects against the background of class A objects for the wavelength registration zone Δλ ₁ (we denote the contrast value in this case k _Δλ1 ) one should take the value U _i (t) from column U _Δλ1 (t) and rows D, and the value U _j (t) from the same column but row A. Thus, in this case

Let's enter this value in the corresponding column of Table 2 in row D/A.

To calculate the contrast of objects of class D against the background of objects of class B for the resulting signal U ₁ (t) (we denote the contrast value in this case by k ₁ ), one should take the value U _i (t) from column U ₁ (t) and rows D, and the value U _j (t) from the same column but row B. In this case

Let's put this value in column U ₁ (t) of Table 2 in row D/B.

Similarly, let's fill in the entire Table 2. As a result, Table 2 shows the values of the contrast value when an object of class D is located against the background of objects of other classes. At the same time, signal processing operations are advantageous (the maximum contrast of an object of class D against the background of objects of other classes) when the resulting signal of the form U ₂ (t), U ₃ (t) is formed.

Using expression (9), we fill in Table 3.

Table 3 shows the values of the contrast value when an object of class B is placed against the background of objects of other classes. In this case, signal processing operations are advantageous, when the summation and subtraction of the original signals are first carried out, and then they are divided among themselves, that is, the formation of the resulting signal of the form U ₅ (t).

Table 4 shows the values of the contrast value when an object of class A is placed against the background of objects of other classes. In this case, the formation of the resulting signal of the form U ₃ (t) will be advantageous.

Table 5 shows the values of the contrast value when an object of class C is placed against the background of objects of other classes. Here, the formation of the resulting signal of the form U ₅ (t) is advantageous.

Table 6 presents the final data of the required type of joint signal processing when an object of one class is found against the background of others. The plus sign in Table 6 indicates the preferred signal for obtaining the resulting image with the maximum contrast between the object and the background.

According to the proposed method, the considered variant of joint signal processing of two-spectrum TV images shows that the use of operations for subtracting and dividing signals, as well as dividing signals among themselves, after their summation and subtraction, in general, makes it possible to ensure the formation of the resulting images, in which an increase in contrast between different classes of objects compared to the contrast determined by the original signals of multispectral images U _Δλ1 (t) and U _Δλ2 (t). This ultimately makes it possible to increase the reliability of selection and discrimination of objects in image analysis.

Analysis of Table 6 shows that in order to increase the contrast and, accordingly, the visibility of an object of one class A, B, C or D against the background of other classes of objects, for example, an object A against the background of objects B, C, D, depending on the distribution of their spectral characteristics in the selected zones registration, there is a certain variant of joint processing of signals of multispectral images, which provides an increase in contrast to their maximum possible values. So, for the considered case, the increase in image contrast during the operation of signal subtraction is typical for one case of object selection, signal division, and signal division after their summation and subtraction is characteristic of two other cases of selection of a given object against the background of others.

Sources of information

1. Television. Textbook for universities / V.E. Jaconia, A.A. Gogol, Ya.V. Druzin and others: Ed. V.E. Jaconia. - M .: Radio and communication, 2000. - 640 p.: ill.

2. Sagdullaev Yu.S., Kovin S.D., Sagdullaev T.Yu., Smirnov A.I. Information-measuring television systems M.: "Sputnik +", 2013. - 199 p.

3. Sagdullaev Yu.S., Kovin S.D. Perception and analysis of multispectral images. M.: "Sputnik +", 2016. - 251 p.

4. Tarasov V.V., Yakushenkov Yu.G. Dual- and multi-band optical-electronic systems with matrix radiation receivers. - M.: University book; Logos, 2007. - 192 p.

5. RF patent No. 2604898. The method of forming spectral-zonal video signals / Kovin S.D., Sagdullaev Yu.S. - publ. December 20, 2016 Bull. No. 35.

6. Sagdullaev Yu.S., Kovin S.D. Methods for combining multispectral images. Journal of communication technology. TV Tech Series. Scientific and technical collection. M: 2016, S. 29-41.

7. RF patent No. 2546982. The method of forming and displaying signals of color, spectral-zonal and thermal imaging / Kovin S.D., Sagdullaev Yu.S. - publ. 04/10/2015 Bull. No. 10.

8. GOST R 52870-2007. Means of displaying information for collective use. Requirements for the visual display of information and measurement methods. Approved and put into effect by the Order of the Federal Agency for Technical Regulation and Metrology dated December 27, 2007 No. 530-st.

Claims (2)

Method for processing and displaying signals of multi-spectral images, including registration of the input radiant flux F(λ) in several spectral regions - zones, including the ultraviolet, visible and infrared regions of the spectrum of electromagnetic waves, the formation of two separate radiant fluxes F ₁ (λ) and F ₂ from it (λ), reflecting the areas of registration of the radiant flux Δλ ₁ and Δλ ₂ , the transformation of the radiant fluxes F ₁ (λ) and F ₂ (λ) and the formation on their basis in the first channel of the image signals U _Δλ1 (t), and in the second channel of the signals images U _Δλ2 (t), separate processing of the received signals of multispectral images U _Δλ1 (t) and U _Δλ2 (t), including amplification, conversion of analog signals to digital, digital aperture and gamma correction, characterized in that when performing joint signal processing multispectral images U _Δλ1 (t) and U _Δλ2 (t), including operations of summation and subtraction of two original signals, add the operations of their division and multiplication i, and in general form the following signals of multispectral images

Images