RU1812544C - Method of adaptive imaging of object located behind irregular phase interface - Google Patents

Abstract

Usage: when searching for objects under the excited surface of the water. SUMMARY OF THE INVENTION: illumination of an object is produced by a pulsed laser located on one side of the interface between the media and the receiving device, the pulse duration being less than the time of light propagation from the media interface to the object being studied and vice versa. The reflected radiation is received by a receiving device located in such a way that the laser-illuminated section of the medium’s interface falls into the field of view of the receiving device, the deviation of the optical axes of the laser and the receiving device from the normal to the plane closest to the medium’s interface does not exceed the mean-square inclination of the medium’s interface to of this plane, and on a similar aperture of the receiving device, a system equivalent to two confocal lenses and an inertial optical neuron matrix located in the image plane of the first lens of the illuminated portion of the media interface. 3 ill. (L 00 ju e b

Description

The invention relates to adaptive optics and can be used when searching for objects under an agitated surface of the water.

Conventional methods for correcting the image of an object located behind a distorting medium consist in the fact that an adaptive device controlled by radiation from a reference Source located on one side of the distorting medium with the studied object corrects the image of the reference source and then directs it to the studied object. Since the wavefront of radiation from the studied object is deformed by the distorting medium in the same place as the wavefront of radiation

reference source, so far as the adaptive system corrects the image of the object.

Consider adaptive phase compensation techniques in imaging systems using the circuit of Fig. 1. The methods used here to compensate for wavefront distortion (or wavefront adaptation) and sharpen the image are similar to phase conjugation and aperture sounding methods. In the first case, a reference point source located in one isoplanatic region with the object is required. Using the sharpening method, which is a variation of the aperture sounding method, phase distortions are analyzed indirectly by measuring the gradients of the sharpening function in the image plane.

When correcting the image of an object located beyond an uneven media interface, it is necessary to compensate for the tremendous wavefront distortions compared to the wavelength. The aperture sounding method does not work at large phase distortions (the sharpness functionals have innumerable local maxima) and cannot solve the problem posed. Therefore, a correction method similar to the phase conjugation method has been selected as a prototype. The light from the object 1 passes through the medium 2, which distortes the wavefront, and then enters the lens 3. The controlled correction element 4 corrects the phase of the beam in accordance with the information about the phase distortions of the radiation from the reference source 5 and creates an adjusted image of the object in the image plane. In the case of searching for an object located beyond an uneven media interface, as a rule, it is not possible to establish a source of reference radiation next to the search object.

It is an object of the invention to provide the ability to correct an image of an object without creating a reference radiation source on one side of the object from the media interface.

The goal is achieved in that in a method for adaptively forming an image of an object located beyond an uneven media interface, including illuminating an object with collimated laser radiation located on the opposite side of the object from the media interface, registering, using a receiving device, the image of the object in the light reflected from it and image correction by reference radiation control

unlike the prototype, the illumination of the object is produced by pulsed radiation with a pulse duration shorter than the time it takes for the radiation to pass from the interface to the object and back, as the reference radiation use illuminating radiation reflected from the interface, reg

the stratum begins after the leading edge of the reference radiation pulse arrives at the receiving device after a time greater than or equal to the pulse duration and less than or equal to the time it takes to pass the distance from the interface between the media and vice versa, the correction system is located on the input aperture of the receiving device and is made in the form of two confocal lenses and inertial matrices

0 optical neurons with a relaxation time longer than the time of light propagation from the interface between the media and the object, located in the image plane of the first lens of the illuminated portion of the media interface, while the receiving device and the laser are positioned so that the laser-illuminated portion of the media interface is in the field of view of the receiving device, and the deviation of the optical axes of the laser and the receiving device from the normal to the plane closest to the interface does not exceed the rms tilt media interfaces to this plane.

5 An optical neuron is a device with a transmission coefficient k equal to zero if the intensity of the control radiation is less than the threshold, and equal to unity if the intensity

0 control radiation is greater than the threshold. By transmission coefficient we mean the ratio of the signal at the output of the device to the signal at the input. Illuminated by radiation reflected from the interface of the media, the sections of the matrix of optical neurons go into a state with a transmission coefficient equal to one, and then relax to a state with a transmission coefficient equal to zero. The relaxation time should be much longer than the propagation time light from the interface to the studied object and vice versa. Radiation reflected from the object under study passes through a self-tuning device5 and is registered by a recording device, which is turned on no earlier than after a time equal to the pulse duration after the arrival of the leading edge of the pulse reflected from the interface

media, and no later than after the time during which the light travels the distance from the interface between the media to the object under study and vice versa. In other words, the essence of the proposed method consists in the fact that the radiation of a pulsed laser reflected from the interface is in a non-nesting device and converts portions of the matrix of neurons to a state with a transmission coefficient equal to unity. The recording device is turned on immediately after the end of the pulse reflected from the interface. Since the matrix of optical neurons is located in the image plane of the illuminated region of the media interface, only radiation from the object under study passes through the recording device, passing through only those sections of the media interface from which the reflected radiation that opened the optical neurons came. The radiation reflected from the interface is incident on the aperture of the self-adjusting device only from portions of the boundary whose normals differ by less than d / (), where d is the diameter of the receiving aperture; Li is the distance from the receiver to the illuminated media interface. If the mean-square inclination of the interface is greater than d / (), then the self-tuning system for the radiation reflected from the interface improves the image quality.

Figure 2 shows a structural diagram of an adaptive system that corrects the image of an object located beyond an uneven media interface, the proposed method. 1 - laser; 2-3 - telescope; 4 - border section of the medium. 5 - first lens of a self-adjusting device; 6 - matrix of optical neurons. 9 - recording device; 10 is a photo detector controlling a recording device.

The laser 1 and the receiving device 5-10 are placed above the portion of the media interface under which it is supposed to be the object under study. The optical axis of the laser can scan, deviating from the normal to the plane closest to the media interface, by less than the mean square inclination of the media interface to this plane. The optical axis of the receiver may remain parallel to the normal if, when scanning a laser beam, the illuminated portion of the media interface will fall into the field of view of the receiver. Otherwise, the direction of the optical axis of the receiver should change in accordance with the orientation of the optical axis of the laser. It is advisable to broaden the laser beam 1 using a telescope 2-3. As laser 1, one can take a pulsed-periodic solid-state laser operating at a wavelength of 0.53 microns. The pulse duration of such lasers varies from picoseconds to nanoseconds, and

0 energy per pulse reaches tens of kilojoules. Let us dwell on a repetitively pulsed laser with a pulse duration of 30 ns, pulse energy J and pulse repetition rate of 5 Hz

5 (20 W). For 30 minutes, no light passes about 10 m, so the object under study should be located at a depth of more than 5 m. Methods for realizing optical neurons are described in the monograph. Currently in connection with

0 work on the creation of optical computers is the process of miniaturization of optical neurons and lowering the threshold value of power at which their switching occurs. The role of an optical neuron5 can be played by a nonlinear Fabry-Perot interferometer, which, when the incident radiation exceeds the threshold intensity value -Lnop, changes its state from reflecting to

0 transmissive. In this case, the neuron should have a low threshold intensity Lnop and the inertia sufficient to be able to miss the weak radiation coming from the studied object located outside the media interface. Nonlinear interferometers with a thermal nonlinearity mechanism satisfy these requirements. Since thermal conductivity plays an important role in the dynamics of the operation of thermal optical neurons, it is useful to give a formula for the characteristic time constant of this process. If the radiation instantly heats a cylinder of diameter r0, the solution of the thermal diffusion equation

5 leads to the following expression for the characteristic time:

Tc ((2),. Where Cv is the specific heat capacity; .riness; kr is the thermal diffusivity of the medium.

0 Both k-r and Cv are temperature dependent, as a result of which Tc can change by orders of magnitude. Tc is proportional to the area (r0) 2, therefore, with a decrease in the size of the resonator, it decreases. 5} data are given on the matrix of optical neurons in the form of Fabry-Perot microcavities of 1.5 microns in diameter with a distance between the centers of neighboring microcavities of 3.4 microns. Microresonators are obtained by ion-beam engraving of a GaAs / AiAs plate. The energy required to switch one neuron is 0.6 pJ and the recovery time is approximately 200 ps. The authors suggest, by further miniaturization, to reduce the switching energy to 17 fJ. A cell with a self-luminous heterogeneous medium can play the role of an optical neuron. Self-bleaching is observed in heterogeneous media in which the difference in the linear refractive indices of the Ap components is commensurate with the difference in the non-linear refractive indices of DpNl and the Dpl and Anp icons are opposite. Optical neurons based on hybrid optoelectronic circuits have lower switching energy because they use the energy of an electrical circuit. The Keldysh-Franz effect makes it possible to implement an optoelectronic scheme of a saturable absorber, and hence an optoelectronic neuron. Application of this effect makes it possible to make a neuron without a resonator for non-monochromatic radiation. In this case, the laser 1 can be multi-frequency. The delay time of the optoelectronic neuron is determined by the time constant of the electric circuit m, where C is the capacity of this target; RH.- load resistance. Descriptions of miniature optoelectronic neurons made of GaAs are available in the literature for working with radiation with a wavelength of 0.85-0.89 µm. (The energy of the quantum should slightly exceed the width of the forbidden zone). Descriptions of miniature optoelectronic neurons operating at a wavelength of 0.53 µm are currently not available in the literature. Therefore, as an example of a specific embodiment of an optical neuron, a thin-film semiconductor interferometer with an intermediate layer of ZnS or ZnSe operating in the reflection mode can be taken. The switching energies of such thin-film interferometers with a diameter of 4 microns were measured in Ref. 9 and are J. In order to have a sufficient energy reserve for the type of laser we chose and the reflection coefficient from the interface (w - naf / tnt - n2). , the focal length of the lens 5 should be such that the area of the matrix 6 does not exceed 200 cm. Microcenters with a diameter of 1 mm can be placed on this area. Thermal conditions should be maintained such that the turn-off time of neurons is at least 1 ms. An essential characteristic of an optical neuron in this application is the ratio of the transmission coefficient of the closed neuron k (l), l Pores to the transmission coefficient of the open neuron k (l +), I + pores. Indeed, the fraction of open neurons - D in matrix 6 is

J.

D () 2

ГОL.

where d is the diameter of the aperture of the receiving device; dt - transmitting; L is the distance from the illuminated surface area to the receiving aperture, o is the standard deviation of the interface between the media from the nearest plane; r0 is the average size of the irregularities of the interface, (st7go is the mean-square inclination of the interface). In order for the radiation passing through the closed neurons not to reduce the image quality, the following condition must be fulfilled:

K (l-)

Wit

"D,

25 V; - -. -

for example, at 7 / g0 0.001, L 100 m,

. m, D 0.01, i.e. the condition must be fulfilled:

K (U) 00 ..,;

 W

This imposes rather stringent requirements on the quality of thin-film interferometers. The energy of radiation reflected from the studied signal object. Having passed through the matrix of optical neurons 6, it can be estimated as follows: .;.: ..

(N

:,. jnfoQiSL + bf

:. L.

where Eo is the energy in the Laser Pulse; Kotr coefficient of scattering back of the studied object; Kpog - absorption in the second medium

on a path with a length, where 1 is the distance from the interface between the media and the object under study. Since the Knog factor depends exponentially on LS and the exponent

absorption of the second medium, the value of EOSig can be very small. Therefore, the recording device 10 may have a cascade image intensifier controlled by a photodetector composed of two or three chamber electron-optical converters (EOPs), the screen of one converter being connected to the photocathode of the other via a fiber-optic element. An advantage of image intensifiers is the ability to register a signal for a short period of time up to 10 12-10 s. For operation at a wavelength of 0.53 µm, it is advisable to use PDAs with a GaAs / GaAIAs photocathode having a quantum yield of up to 40% and a low level of thermionic emission. As an example of a specific embodiment, we choose a solid-state laser with a pulse duration of 30 nsec, a pulse repetition rate of 5 Hz and a power of 20 W. 3. The matrix of optical neurons is selected in the form of a matrix of thin-film interferometers with an intermediate layer of ZnS or ZnSe with a diameter of 1 mm operating in reflection mode 9. To register the signal from the object under study, one can use photodetector-controlled image intensifier tubes with a photocathode of the GaAS / GaAIAs type.

The following numerical experiment was carried out to evaluate the effectiveness of a system using the proposed method for correcting the image of an object located beyond an uneven media interface. The refractive index of the first medium is chosen equal to m 1, the second l 2 1.33. The Z axis is directed normal to the plane closest to the interface. The media interface was described by

Otg

Z (X, Y) 2 Anmexp {i (+) + i

n, mU1 Fpt,.

where FPT are random numbers uniformly distributed over the interval 0, di is the size of the region illuminated by the probe beam.

The coefficients Anm are determined by the form of the autocorrelation function in

In (ДХ, AY) Z (X, Y) Z (X + А X, Y + Д Y), where the sign means averaging over realizations. In our numerical experiment, a Gaussian autocorrelation function was taken

B (X, Y) o2 exp {

X2 + Y2 -1

}

where si is the dispersion of deviations of the interface between the media from the plane Z 0; r0 is the average size of the irregularities.

To simplify the calculations, the transmitting aperture was aligned with the receiving aperture, and the receiving device was focused at infinity (La f). A beam of 216 beams parallel to the Z axis was sent from laser 1. The intersection points of the rays with the interface defined by formula (1). were Newton's method. The rays reflected from the boundary of the section continued until they intersected with a matrix 6 consisting of optical neurons. If a neuron pop5 gave at least one ray reflected from the interface (3 µJ), then this neuron was considered open. Then, a beam of 218 rays was sent parallel to the Z axis from the second medium, covering a section of the medium’s interface,

0 falling into the field of view of the receiving device. The points of intersection of the rays with the media interface given by formula (1) were found by the Newton method. Refracted rays that hit the receiving aperture

5 were conducted through the system to the image plane 9.

In FIG. Figure 3 shows the distribution of energy W in the image plane 9 in a circle of radius r as a function of g (function

0 scatter a point). For comparison, the same figure shows the function of scattering a point in a system without a self-adjusting device. It follows from Fig. 3 that in the absence of a self-adjusting device, the image plane is almost uniformly illuminated. If there is a self-adjusting device in the image plane, we have a small speck.

0 FORMULA AND PREPARATION

A method for adaptively forming an image of an object located beyond an uneven media interface, including lighting the object with collimated

5 by laser radiation located on the side opposite to the object from the media interface, registration by means of a receiving device of the image of the object in

the light reflected from it and image correction by controlling the correction system with reference radiation:. with a topic located in front of the receiving device, with the exception that, in order to eliminate the need to create a reference radiation source5 on the same side of the medium as the object’s interface, the object is illuminated by pulsed radiation with a pulse duration shorter travel time

0 radiation from the media interface to the object: and vice versa, the illumination radiation reflected from the media interface is used as reference radiation, registration starts after the leading edge of the reference radiation pulse arrives at the receiving device after a time greater than or equal to the pulse duration and shorter or equal to the time the light travels the distance from the interface between the media and the object and vice versa, the correction system is located on the input aperture of the receiving device and is made in the form of two confocal lines h and matrices of inertial optical neurons with relaxation times longer than the propagation of light from the interface between the media and the object and vice versa, located in the image plane by the first lens of the illuminated portion of the media interface,

the receiving device and the laser are positioned so that the illuminated portion of the media interface is in the field of view of the receiving device, and the deviation of the optical axes of the laser and the receiving device from the normal to the plane closest to the media interface does not exceed the rms tilt of the media interface to this plane.

FIG. Z

. 5