JPH04503721A - Optical correlation system and method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Optical correlation system and method Technical field The present invention generally relates to robot vision and a matted field for detecting the presence of objects. Optical correlation that extracts appearance information such as the position and size of the target object using a router. systems, particularly for robot vision systems that primarily utilize parallel optical processing. and a preferred optical correlation system.

Conventional technology The research and use of robotic vision systems has implications for both commercial and military operations. This is a field that is rapidly increasing in interest. Current commercial use , confirmed for machine operations and production lines. In the prior art, The Bot Manini builder is a comparative model that performs simple tasks including painting and spot contact. was used for large-scale assembly lines. These machines are generally of high standard. It operates according to a set of program instructions, and the positioning and direction of the handled parts are determined. The markings were quite accurate. Normally, the programming of the Mayu 1 builder mentioned above is The program is relatively complex and allows the positioning of parts to be within relatively accurate tolerances. useful only if kept at .

In recent years, the increase in various sensory abilities has increased the flexibility of the manicure Attempts have been made to increase sex. Tactile and auditory abilities are In addition to the visual abilities involved in inventions, they are currently being developed. Rande (gang) Light and binocular vision technology discovered and created by (llight and b) The robot vision described above by inocular vision technology 14ues) It is used in the sensory system. However, these systems specifically Developed into applications that require confirmation of the existence of an object and determination of the location and orientation of that target. There is nothing that is. Furthermore, the well-known Lobov) visual system It takes a considerable amount of processing time to confirm the existence of a target object. many robots or machine vision systems in their ability to perform special purpose tasks. This is disclosed and analyzed in the prior art.

These systems typically consist of sensors consisting of analog devices and digital There is a mix of processing and joint control. A further branch of this technology is the digital This is a selection of target object recognition using analog and analog methods. Digital systems are often , video taikan and tools that solve targets and parts corresponding to their size and appearance. It relies on algorithm. The storage library depends on the capacity of the computer memory. limited only. Most of the several optical systems rely to some extent on electronic processing. ing.

U.S. Patent No. 4.462.046 by Speight (Splgbt) A video camera and an off-axis view are controlled by the combination unit. An optical vision system stored in a Losefsa is disclosed. fast fourier transformation FFT performs video analysis and processing within a computer. this Such systems are limited to at most 3o frames per second.

This invention enables optical processing of Speight's 7 stems at speeds close to the speed of light. and execute it. Many appearances of many interesting objects can be achieved with a single compound matt filter. and their filters are specially designed to handle multiple targets. It's been in. Additionally, an advantage of this invention is that porogram lenses can Optically parallel processing of objects and/or many scenes (VIIJS) be specifically designed to handle the desired decomposition and the degree of target appearance desired. It lies in the fact that it can be measured in terms of ability.

U.S. Patent No. 3.779.492 by Grumet This is important for the The system is disclosed.

In this, coherent optical signal processing equipment is used for well-known targets. It is being The above-mentioned 77-tide filter comprises a pair of matched filters. This matched filter separately detects high and low frequencies spread across the space. Process. The outputs are combined with a logical AND operation, and the target is It is examined for minute features as well as appearance. Matsutido Fi The optical memory bank of router sets is , that is, all the decomposable objects of the target form the visual detection m-comb filter bank. It includes a diffraction pattern of the plane (VIEW). All aspects of the above recognition bank are Once the detected object is recorded in some of the planes in the memory bank, the object is In order to determine whether the target is a target that are investigated optically and symmetrically.

Disclosure of invention This invention improves the system by Grumet in several important respects. It has been improved and made different. One embodiment of this invention is 71 days. An inverse Fourier transform lens array to receive and receive the optical correlation output of the array of filters. and each optical correlation output is directed to a separate detector. Gourmet is an alternative to this. A single lens 29 is shown which detects all of the outputs on a common detector. Also , this invention uses independent matched filters to generate a normalized signal. Provide normalization means for each of the data. Gourmet also requires signal normalization. I don't fully understand sex. The above split detector amplifies each output signal separately and that each amplifier is used for independent normalization of its processing channel. There are advantages. In addition, Gourmet allows each mattified filter to have its own independent corner. has a reaction curve, is formulated, and is used to measure appearance information about a subject. do not understand that it can be applied to

The overall principle of operation of this invention is that an object is recognized, its size, position and external The goal is to measure the visual field and generate signals for control. This system Identify the object, measure its size, measure the position of the object, and To be able to measure the irregularities of the contour and, if necessary, the velocity of the object. Apply the characteristics of optical matched filters.

As described herein, the present invention provides identification of input video. and one or more mattified fibres, to provide information on its appearance. A system for optically comparing the optical information stored in the router with the input video. Provide the stem. This input image is incident on an optical modulator that spreads out in space, and the coherence is modulated into light. Multiple multiplexed holographic lenses having a spatially modulated radiation beam, and a plurality of sets of spatially modulated radiation beams; Multiple Fourier transforms are performed to obtain an array of Fourier transforms consisting of . Matsushi The array matching the defilter consists of a Fourier transformed narogram of the object's appearance. With each matched filter, we have an array of Fourier transforms incident on it. .

Each matted filter also has a Fourier value recorded by the matted filter. The degree of correlation of the Fourier transform of the spatially modulated radiation beam above, along with the E transform It passes an optical correlation signal that depends on the alignment. The inverse Fourier transform lens array is transmits and receives the optical correlation outputs of the array of optical filters, and generates an inverse filter for each optical Perform the Rie transform. The detector array then performs the inverse Fourier transform of each optical correlation output. and generate a detection output signal representing each optical correlation output.

According to the previous embodiment, the detection output signal is used to measure appearance information about the input video. The signals are then electrically processed in a processing circuit that compares their relative amplitudes.

The above processing circuit includes a normalization amplifier circuit for each detection output signal, a normalization amplifier circuit for each detection output signal, and a normalization amplifier circuit for each detection output signal. Analog/digital conversion to convert the output signal to a matching digital signal. and a comparator for comparing the amplitudes of the digital signals.

The output of the comparator is a logic circuit that generates a control signal for the robot to move right or left. processed in the road.

The present invention provides a split mated filter channel for use in an optical correlation system. Make it clear that it is necessary to normalize the output of the channel. Normalization is The divided normalization amplifier for the detector in the filter channel This can be accomplished electrically. Here, the gain/gain of each normalization amplifier is is set to compensate for differences in the outputs of a number of independent matched filters. substitute In other words, normalization is processed through each independent matched filter channel. By adjusting the amplitude of the transmitted light or by adjusting the output power of the laser, Alternatively, a deflection filter is rotatably installed in the optical path of each mated filter channel. This can be achieved optically by an attenuation filter consisting of a filter.

The normalization function is usually an independent angular inverse that is the difference for each matched filter. can be linked to the response curve. Therefore, the angular response curves This is empirically determined for matched filters. for all angular response curves The maximum amplitude signals are set equal to normalize the angular response curve. and , the angular region of the visual field for each normalized angular response curve is determined. This is it thus producing the desired overall angular sensing response for the /stem. The number of matched filters requested is determined.

Brief description of the drawing Figure 1 is used for teaching in this invention and shows the usual principle of operation. 1 is a transparent perspective view showing the configuration of an exemplary embodiment of the robot vision system shown in FIG.

Figure 2 shows the angular sensitivity of the matched filter to the width of the frequency band spread in space. A theory to show! I! It is a diagram.

Figures 3a and 3b show the matches for image rotation at a given scale size. FIG. 2 is an explanatory diagram showing the sensitivity of a filter.

FIG. 4 is a plan view of a simple target object showing the effect of object orientation.

Figure 5 shows a Fourier transform to explain the effect of rotation of the target object shown in Figure 4. It is an explanatory diagram.

FIG. 6 shows the angle Il! of the matched filter. i degrees and more special objects of interest Angular sensitivity of the correlation signals of the three matched filters for the three possible orientations FIG.

Figure 7 shows a system in which images from the same angles in +7 directions are recorded on a matt filter. It is a figure showing the correlation signal in .

Figure 8 shows a position control system that develops a signal processing circuit for generating position control signals. FIG. 2 is a block diagram showing the configuration of FIG.

FIG. 9 shows the configuration of another embodiment with a detector having a segmented array. Demonstrates optical techniques for normalizing the output of each matched filter channel FIG.

Figure 10 shows the MF memory for the plane viewed from the direction of focus on the target object. FIG. 3 is a diagram showing a typical example of a configuration for preparing banks.

Figure 11 shows that all angular sensitivity curves of the matched filter have a range of amplitude and angle. FIG. 4 is a diagram showing an angular sensitivity curve under an ideal situation in which the angles are equal.

Figure 12 shows the angular sensitivity curves of each matched filter in both amplitude and angle ranges. FIG. 3 is a diagram illustrating an angle W & degree curve under a practical situation in which

Figure 13 shows the oscillation of each curve to obtain an interesting angular range for each matched filter. A diagram showing a normalized angular sensitivity curve of the sensitivity curve shown in FIG. 12 with equal width peaks. It is.

detailed description Many of the requirements and concepts related to this invention are used in this description, and this invention is essential for understanding the functions and general principles of operation. moreover , some characteristics of these concepts are mentioned below at the beginning.

A holographic lens (ML) combines continuous radiation from a diffuse point source with a point By recording the interference pattern with a collimated beam of radiation that produces a hologram of the source. is created. After the hologram lens is recorded on film and processed ) When illuminated, that hologram lens reproduces a point source, i.e. it act like If the recording process is repeated initially, the point source hologram An array of hologram lenses, or multiple hologram lenses (MHL), is recorded on the film.

The present invention utilizes a spatially modulated input, i.e. a Fourier transform of a laser radiation beam. Offset of multiplex holographic lens array to produce an array of One of several distributions in angle, position and focal length is utilized. in general, The particular requirements of the array will be determined by the particular problem being addressed.

In short, a holographic lens is a laser beam that illuminates a landscape or target object. Take the Fourier transform (FT) of the system. In addition, the holographic lens Take multiple sets of transformations simultaneously. Multiple holographic lens arrays are typically , combined into a similar multiplex array of matched filters.

When the lens is illuminated by a spatially modulated parallel beam (i.e. (when the image is spatially modulated by passing the film through the scenery, objects, etc.) , the lens produces a Fourier transform of the target to a focal point on the film. this is a lens This is the basic characteristic of Parallel (or reference) views from the same light source whose Fourier transforms When interfered with by a system, an interference pattern is created. This is a Fourier transform hologram , or called a matched filter (MF). It is the input target light It is a scientific spatial filter. When any scene is played through the system, the pine The Tido filter picks out objects and passes them through to do the above. centre The signal passed through the filter is Fourier transformed again, creating a ``correlation'' surface. Ru.

If there is a matched filter target, a sharp correlated signal will result, whereas Therefore, if there is no target, a broad small correlation signal will be generated on the correlation surface.

This invention uses the degree of the matched filter for the rotation or size of the object. ing. their appearance (i.e. at the angle at which the object appears, or at the distance at which it appears), or a specific magnitude) changes, the correlation signal changes.

FIG. 1 shows a memory bank used in a matched filter according to the technique of the present invention. FIG. 2 is a transparent perspective view showing the configuration of an optical correlator. The target lO is relative to the optical correlator With a moving input, an image is projected onto a spatial light modulator (SLM) 14 by an input lens 12. be converted into This spatial modulator (SLM) 14 is arranged on the laser beam from the laser 16. spatially modulate the image. The image is uploaded by #118 and beam splitter 20. The signal is directly input to the spatial modulator 14. The spatially modulated laser beam is multiplexed Fourier transform is performed by the holographic lens 22, and the matched filter ( MFs) 24 to corresponding arrays. Inverse Fourier transform lens array 26 performs an inverse Fourier transform on the output of the MFs 24 and supplies the output to the detector array 28. do. The output signal is 30 (details will be described later) in order to generate an output control signal. ) is electrically processed.

71 Tido filter is sensitive to the size, angle and position of the input object. This is a Fourier transform (FT) program with the characteristics of degrees. these The parameters of are the set of angle and range lines that cover the expected target appearance. is set in advance to instruct. The detector is located at the desired location to analyze It is partitioned to the extent that In the mated filter configuration, holographic The optical fringes are visually optimized with special spatially spread frequencies. This will satisfy the required size and/or appearance sensitivity.

Since both requirements cannot be satisfied at the same time, multiple independent MFs are used in the present invention. It is often used. The nature of different given applications may require MF sensitivity to be an important factor. Ru.

Matched filters are characterized by their size, wavelength, recording medium thickness, and Fourier transform lens. Complexity with focal length, contrast ratio, overlap, alignment and spatial frequency It is a holographic structure. Each of its features is limited to the following: or described using an approach to quantitative 71-tide filters.

(a) First, in a matched filter for a complex target, the correlation matrix A typical matrix is shown below.

90 93 97 96---Once''-=1 top-1-new 96 93 9-ni -L shell-beam ratio R signal The y-axis is the amplitude transfer (the amount of light propagation from 0 to 1.0, which is the reciprocal of the density). optical dew Phototime is often used as a practical parameter and is shown on the right.

The X wheel is the ratio of the reference beam divided by the signal beam ratio, denoted by the symbol R. .

In practice, the beam ratio is given as different exposure times. This is the correlator's This can be easily realized with a shutter, and this procedure is described below. This creates a column of values Occur. Then, the new beam ratio R is determined to be +3 (IB or 2 times or more, The procedure is repeated.

The values in the matrix are based on the photosensitive plate used to make the matt filter. Obtained after being exposed, developed, dried and **** in an optical correlator This is the peak value of the correlation value. Matrix development is a long and time consuming process. It depends on the material, target, etc. This figure shows the complexity of a matted filter. Due to its properties, matt filters made of a given material are complex. This shows that it is a detailed process.

(b) The developed standard for the bench 2 is called the system standard S.

Here, λ (mm) and F (mm) are the operating waves of the matched filter, respectively. is the focal length of the lens. A small value of S means that a variety of parameters It is easy to control the sensitivity and with small S the position problem is also easy.

The system reference S indicates the spread and the MF is scattered on the medium (therefore the buffer window width, required exposure time, etc.).

(C) Kan 3's criterion for capacity C is 71 chips placed on a photosensitive plate for recording. developed to show the number of filters.

Capacity C trout constant x3-P-G/((Δν)N-F・λ) IC is pine per 1 cm square This is the capacity of the filter.

S is a separate spectral element (since the Fourier transform is 180° symmetric, the map half or all of the filters can be used), take the value l or 2.

P is the number of duplicate filters at a given location. This accepts values up to 30. However, conservatively speaking, P takes a value of about 4.

G is a geometric parameter and depends on the mixture of objects to be recorded. , takes a value between 1 and 1.5.

λ and F have already been described. Practical values are 0.45 μm≦λ≦2 μm, respectively.

and 25≦F≦1500cm.

N·Δν is the bandwidth of the matched filter.

N is the multiple of Δν used to obtain 96% of the autocorrelation value. Δν is usually The period at which the matched filter takes the optimum value for the holographic contrast ratio. set to the wave number band.

This representation illustrates the complexity of a matted filter; all of the elements are , when they are complex and there is a requirement to perform accurate normalization. must be thought through.

Special equipment provides matched filters for angular direction, size and load designed to measure the sensitivity of In addition, the criteria mentioned above For example, Δν and angular sensitivity must be given to each group of routers. They are linked, and changes in one affect the other. In addition, the exposure time Δt of the photograph is Affects Δν and angular sensitivity. Next, let's consider the value of the system standard. λ or F, or both, S changes, like Δν and angular sensitivity. During this continuous process, the FT begins as shown in Figure 5a. Look out Ru. In general, the number of required MFs is determined by the number of (8) target After normalization, each of the channel signals R, C and L is transmitted to the converters 42A-42D and It's inappropriate. Therefore, the normalization is performed according to the quality of each mabutide filter. This is a well-known technique, such as by driving a running motor. In the previous example, Figure 1 0 indicates the placement of the lactate mabutide filter (MF) 68. Filter this image In addition, an averter is used to allow the functionality of one filter while the filter is blocked. The view of the target from the direction of the object is a tool that itself has angular scale sensitivity. produces a matched filter. In the robot vision system, the object is When viewing a target from the opposite direction, each matted filter acts as a wing. , when an interesting signal is generated, it is treated as a matched filter for a new target. . Therefore, normalization is based on the same standard, i.e. exactly placed matt required for the filter. Additionally, normalization is affected by angular sensitivity Therefore, corrections not described in the prior art are required.

For example, for the Robofton stem for moving L-C-H, mattified fill The data must be assembled for each of the left, center, and right views.

And the angular sensitivity curve is the working response curve which is the response curve to normalize must be generated for each matched filter to generate .

An angular response curve is generated for the view from the central position and for views from the left and right so that the angular response curves for the filter are different. It is used as

The number of MFs required depends on the response of each independent filter. For example, certain A matched filter for the target is indicated by one of the protrusions in FIG. It may show a similar angular response. If this is the view from the top, rotated If the target always looks the same, each response will look the same, so 3 A complete memory to cover 60° is one fence as shown in Figure 11. will have a reaction. All reactions are of equal extent and height.

Now, consider a moving robot part. The robot part is considered from different directions. If so, the set of reactions shown in FIG. 12 would be obtained. independent reaction is a 3db response (i.e., the baseline is located at 172 of the peak level). Note that the angular width varies in magnitude and that the peaks of the correlated signal are at different heights. There is a need to. These are different views, more energy in one view than in another. This occurs because it has a lot of energy.

For practical results, their amplitudes must be moved to a common level.

This can be accomplished electrically by a normalizing amplifier, as described above, or , a matt filter photographic plate is placed in a uniformly dark situation and then selectively exposed to light. This is accomplished optically by

The resulting reaction then occurs as shown in FIG. numbers are arabic numbers It is expressed in letters. On the other hand, the peak values are equal. required for a specific angular range. The number of matched filters was normalized so that the angular response was shown in Figure 13. It will be decided later.

Embodiments of the present invention and variations thereof for robot vision systems are described in detail. However, the disclosure and technology of this invention may be replaced by technology in this field. is a matter of course.

Engraving (no changes to the content) control FIG. 3b FIG.7 FIG, I.O. MF memory bank equally spaced angular responses for all appearances to the target object FIG, l l Angular response due to rotation when corrected FIG. 13 Procedural amendment (formality) March 3rd, 1992

Claims (10)

[Claims] 1. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation system that provides presence and appearance information of force images, (a) spatial Modulate and analyze the coherent beam to form a beam of radiation modulated into (b) a spatial light modulator into which an input image is incident for a spatially modulated image; A beam of radiation is injected into the person and an array of multiple sets of Fourier transforms of the beam of radiation is obtained. multiple holographic lenses for performing multiple Fourier transforms; (c) the Fourier transform hologram of the target object's appearance upon which the Fourier transform array is incident; The matched filter includes a Fourier Record the transformation and degree of correlation of the Fourier transform of the spatially modulated radiation beam The map that passes the optical correlation signal to the channel of the matched filter depends on the an array of filters; (d) sending and receiving the optical correlation outputs of the array of matched filters, and transmitting and receiving the optical correlation outputs of the array of matched filters; (e) an inverse Fourier transform lens for performing an inverse Fourier transform on the force; and (e) each light beam. detects the inverse Fourier transform of the optical correlation output and generates a detection signal representing each optical correlation output. detection means for detecting (f) A normalization check provided for each channel of the matched filter. generate output output signals, Generate an angular response curve for each independent matched filter, and Normalize the lines and set the maximum amplitude signal to make all of the angular response curves equal, Determine the angle of view for each normalized angular response curve and the desired overlapping angle. A normal that determines the number of matched filters required to produce a detection response. and (g) Compare each of the normalized detection output signals to obtain appearance information about the input video. and a comparison means for generating an output direction control signal as determined by the information. An optical correlation system characterized by: 2. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation system for providing presence and appearance information of a force image, (a) The array of matched filters has at least a also one center matched filter and a small at least one left matched filter and a small filter for dawn from the right side of the target; (b) the detection means comprises at least one right matched filter; at least one center detection means for a heart matched filter; and said left match at least one left detection means for the matched filter and said right matched filter; (c) the comparison means comprises at least one right detection means for the data; The output signal of the left detection means is compared with the amplitude of the output signal of the center detection means, and which at least one left comparison means for determining whether the amplitude is large; and said right detection hand; The output signal of the stage is compared with the amplitude of the output signal of the center detection means, and which amplitude is larger is determined. and at least one right comparison means for determining whether the (d) transmitting and receiving the outputs of the left comparing means and the right comparing means, and determining the determined angle of the target object; (e) directional means for generating an output directional control signal indicative of degree position; and (e) from said detecting means. to determine appearance information about the input video. processing means for comparing relative amplitudes of the detected output signals; Equipped with The processing means further includes a normalization circuit for each detection output signal, and a normalization circuit for each detection output signal. an analog/digital converter for exchanging the output signal to the corresponding digital signal; , and comparison circuit means for comparing the amplitudes of the corresponding digital signals. The optical correlation system of claim 1, characterized in that: 3. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation system for providing presence and appearance information of a force image, the The normalization means has a normalization amplifier for each detection output signal, and the normalization means has a normalization amplifier for each detection output signal. 2. The optical correlation system according to claim 1, wherein the optical correlation system generates . 4. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation system for providing presence and appearance information of a force image, the The normalization means comprises an optical normalizer for each matched filter channel. The optical correlation system of claim 1, characterized in that: 5. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation method for providing presence and appearance information of a force image, (a) spatial A coherent beam is modulated and separated to form a beam of radiation that is highly modulated. inputting an input image to be analyzed into a spatial light modulator; (b) Fourier analysis of the radiation beam; To obtain an array of multiple sets of transforms, we can perform multiple numbers of Fourier transforms. , injecting the spatially modulated radiation beam into multiple holographic lenses; (c) with a matched filter containing a Fourier transform hologram of the target appearance; , record the Fourier transform by the matched filter and the spatially modulated of the matched filter depends on the degree of correlation of the Fourier transform of the radiation beam the Fourier to an array of matched filters that pass the optically correlated signal into the channel. input an array of transforms, (d) determining the optical correlation outputs of the array of matched filters for each optical correlation output; (e) Each light beam enters the inverse Fourier transform lens for performing inverse Fourier transform. detects the inverse Fourier transform of the optical correlation output and generates a detection signal representing each optical correlation output. , (f) normalize the signal through each channel of said matched filter, and The normalization step produces a normalized detection output signal, wherein the normalization step is independent of the peak amplitude and With each independent angular response curve having a specific angular range, each independent matched fill to normalize the angular response curve. After setting the peak amplitude signal for all of the angular response curves, To equalize the peak amplitudes of all matched filter channels, normalizes the signal passing through the matched filter and applies it to each matched filter channel. generate different angle ranges for Determine the angle of view for each normalized angular response curve and the desired overlapping angle. determine the number of matched filters required to produce a degree detection response; (g) Compare the amplitudes of the normalized detection output signals and determine the appearance information about the input video. a light source characterized in that the light source is configured to generate an output direction control signal as determined by the information; Academic correlation method. 6. The input image is optically compared with the optical information recorded in the matched filter, and the input image is An optical correlation method for providing presence and appearance information of a force image, comprising: (a) said Injecting the array of Fourier transforms onto the array of matched filters includes: at least one central seven-point filter for the centerline view of the target; at least one left matched filter for a left view of the target; and at least one right matched filter for a right view of said target. (b) injecting said array of Fourier transforms into said center match; at least one center detection means for the left matched filter; at least one left detection means for the filter and said right matched filter. (c) detecting the inverse Fourier transform with at least one right detecting means; The amplitude of the output signal of the stage is compared with the amplitude of the output signal of the center detection means, and which amplitude is determined. at least one left comparison means for determining whether the width is greater; and said right detection means. The amplitude of the output signal of is compared with the amplitude of the output signal of the center detection means, and which amplitude said normalization by at least one right comparison means for determining whether Comparing the amplitudes of the detection output signals, (d) output method representing the determined angular position of the target; outputs of the left comparison means and the right comparison means are used to generate a direction control signal. , (e) output signals of the detector to determine appearance information about the input image. Compare the relative amplitudes of the signals, A normalization circuit for each detection output signal and a digital circuit for each normalization detection output signal. Analog-to-digital converter for converting to digital signal and corresponding digital signal. 6. The method according to claim 5, further comprising comparison circuit means for comparing amplitudes. Optical correlation method. 7. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation method for providing presence and appearance information of a force image, the normalization Steps are performed for each detector output signal to generate a normalized detection output signal. , performed by a normalization amplifier circuit, and all of the matched filter channels are For each normalization amplifier circuit, in order to achieve a normalized detection output signal, 6. The method according to claim 5, further comprising the step of storing in a memory an amplification factor of Optical correlation method. 8. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation method for providing presence and appearance information of a force image, the normalization The step is performed by optical normalization in each matched filter channel. 6. The optical correlation method according to claim 5, wherein the optical correlation method is carried out by: 9. The input image is optically compared with the optical information recorded in the matched filter, and the input image is In an optical correlation method for providing presence and appearance information of a force image, the normalization Steps include optical attenuation filters in each matched filter channel. The normalization step is performed by The rotatable deflection attenuation is carried out by a rotatable deflection attenuation filter in the rotatable deflection attenuation filter. The filter achieves normalization for all of the matched filter channels. Selectively rotate for each matched filter channel to The optical correlation method according to claim 8, characterized in that: 10. Optically compares the input video with the optical information recorded in the matched filter, In an optical correlation method for providing presence and appearance information of an input image, the regular The conditioning step controls the power to the laser beam in each matched filter channel. 9. The optical character correlation method according to claim 8, wherein the optical character correlation method is carried out by controlling the optical character.