JP3009805B2 - Recording and playback device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a large-capacity, high-density recording / reproducing apparatus and an image processing apparatus used for the recording / reproducing apparatus, and more particularly to a recording medium obtained by two-dimensionally scanning the surface of a recording medium with a probe. A recording / reproducing apparatus for extracting read information from image data by regarding a read signal as image data, and an image processing apparatus used in the recording / reproducing apparatus capable of removing impulse noise in the image data without affecting other signal components Related to the device.

[0002]

2. Description of the Related Art In recent years, the recording capacity of data in a recording apparatus has been increasing. Therefore, it is necessary to further reduce the size of a unit recording bit and further increase the recording bit density. ing. For example,
In a digital audio disk using optical recording, the unit recording bit is reduced to about 1 μm ² .

On the other hand, recently, a scanning tunneling microscope (hereinafter abbreviated as STM) capable of directly observing the electronic structure on or near the surface of a substance has been developed. The real space image can be measured with high resolution. STM is
It has the advantage that it can be observed at low power without damaging the medium (sample) due to electric current, and it can be measured not only in ultra-high vacuum but also in the air and in solution,
It can be used for various materials. Therefore, S
TM is expected to be applied to a wide range.

In STM, a voltage is applied between a metal probe having a sharp tip (probe electrode) and a conductive substance to cause both of them to become one.
Utilizing the fact that a tunnel current flows between them when approaching a distance of about nm. This current is very sensitive to distance changes between them. Therefore, surface information in the real space can be obtained by scanning the probe in the in-plane direction while performing control so as to keep the tunnel current or the average distance between the two constant. At this time, the resolution in the in-plane direction is 0.1 nm or more.

By applying the principle of the STM and using a material having a memory effect on voltage-current switching characteristics, for example, a thin film of a π-electron organic compound or chalcogenide as a recording medium, the unit is It is possible to record information having a recording bit size of 0.01 μm ² . The material constituting such a recording medium is disclosed in, for example,
Nos. 161552 and 161553. In addition, if a method of changing the surface shape or surface state of the medium using an electromagnetic wave such as an electron beam or light is used,
Although the unit recording bit becomes large due to the limit such as the degree of convergence of the beam, recording and reproduction of information can be performed at the same recording density as the current optical recording.

However, when such high-density recording / reproducing is performed over a certain area or more, noise due to irregularities on the surface of the recording medium due to irregularities on the substrate of the recording medium, etc., causes the S / N ratio of the reproduced signal. And the error rate are deteriorated, which is a major obstacle to increasing the recording density and capacity. In order to reduce physical irregularities on the surface of the recording medium, a gold (Au) thin film can be monocrystal-grown on a cleavage surface such as mica (mica), and a recording layer can be formed on this smooth surface. However, when such a substrate is used, a diameter of 10 n
m, that is, substrate irregularities of the same order of magnitude as the recording bit size usually hardly occur, but the shape of the recording bit changes due to steps (steps) and ridges (shape opposite to the groove) in the atomic order, This gives the signal S
/ N and an error rate are deteriorated.

Some of the conventional recording / reproducing apparatuses include an image processing apparatus which regards signal information recorded on a recording medium as image data having a two-dimensional spread and extracts only a recording bit area from the image data. In some cases, the S / N at the time of reading a recording bit is improved by providing such an arrangement. This image processing apparatus extracts image data only as a function defined on a two-dimensional plane and calculates only the recording bit area by calculating the total curvature and the average curvature of the function corresponding to each position on the two-dimensional plane. try to. The total curvature and the average curvature in the vicinity of each pixel can be obtained by calculating a first basic amount and a second basic amount described later in the vicinity of the pixel. Then, it is possible to determine from the signs of these curvatures whether or not the pixel is a point belonging to the recording bit area, and to extract the recording bit area. Here, the first and second basic amounts are local amounts, respectively, and therefore, it is possible to perform advanced parallel processing without depending on the number of pixels.

Regardless of whether the STM application technique is used or the shape and state of the medium surface are changed using an electron beam or light, the shape of the recording bit is considered to be unimodal. More specifically, the shape is such that there is only one maximum value for one recording bit, and monotonically decreases as the position moves away from that position. It is considered that such recording bits are constituted by elliptical points at least around the maximum value. On the other hand, steps and ridges that are irrelevant to the recording bit are considered to be composed of parabolic points if they are linear with respect to a certain in-plane direction. FIGS. 7A to 7C are schematic perspective views showing the shapes of the ridge 91, the step 92, and the recording bit 93, respectively.

Hereinafter, the relationship between the curvature and the local property of the space will be described in detail. Using the total curvature K at each point, the local properties of the space are classified as follows. That is, (1) If K> 0, an elliptical point (elipt
ic point), (2) If K = 0, parabolic point (parabolic
point), (3) If K <0, hyperbolic point (hyperbolic
point).

Here, the points belonging to the recording bit area are elliptical points, and the points belonging to steps and ridges are parabolic points. Therefore, whether the point is a recording bit area depends on whether the total curvature is positive or not. You can judge whether. Also, among the elliptical points, the upwardly convex curved surface has an average curvature H <0, and the downwardly convex curved surface has an average curvature H> 0. Therefore, not only the extraction of the recording bit area but also the orientation thereof is determined. be able to.

The total curvature K and the average curvature H are given by the following equations (1) and (2).

[0012]

(Equation 1) However, E, F, and G are first basic quantities, and P, Q, and R are quantities called second basic quantities, and are given by the following equations (3) to (8), respectively.

[0013]

(Equation 2) However, x _vv 2 order differential coefficient of x by v, x _u denotes the 1 order differential coefficient of x by u, the other is the same. By the way, as discussed here, when treating an image as a function defined on a two-dimensional plane, x = u, y =
v, z = f (x, y). Therefore, the expressions (3) to (8) representing the first and second basic quantities can be rewritten as follows.

[0014]

(Equation 3) By substituting Expressions (9) to (14) into Expressions (1) and (2), the total curvature K and the average curvature H are obtained by the following Expressions (15) and (16).

[0015]

(Equation 4) Since the total curvature K and the average curvature H are local amounts, they are easily affected by rapidly changing noise such as impulse noise. Therefore, it is necessary to remove impulse noise when performing the above processing. Is an indispensable pretreatment. At this time, image processing that changes the total curvature and the average curvature must not be performed as preprocessing.

It has been known that a median filter is effective as a method for removing impulse noise in preprocessing of image data. For example, a median filter with a mask size of 5 × 5 pixels has a pixel f _{i, j}
Of the pixels in the range of i−2 ≦ k ≦ i + 2, j−2 ≦ l ≦ j + 2, (ij−1) / 2 + 1
A process for updating the pixel fi _{, j} with the value of the second largest pixel is performed.

[0017]

However, when a median filter is used as pre-processing for image data, since the invariant set family by this filter is a step, all shapes are deformed so as to be close to the step. . Therefore, there is a problem that the feature amount of the recording bit is deformed by the median filter.

An object of the present invention is to provide a recording and reproducing apparatus having an image processing equipment capable of removing impulse noise from the signal without deforming the like total curvature and mean curvature.

[0019]

[0020]

According to the present invention, there is provided a recording / reproducing apparatus comprising: a probe electrode provided to face a recording medium; voltage applying means for applying a voltage between the probe electrode and the recording medium; Displacement means for relatively displacing the probe electrode and the recording medium; and reproducing at least information from the recording medium based on a physical interaction between the probe electrode and the recording medium according to the relative displacement. Wherein the image processing is performed by a recording / reproducing apparatus comprising: a reproducing unit for performing noise reduction; and an image processing unit for removing a noise from the reproduced signal by regarding a reproduced signal from the reproducing unit as two-dimensional image data based on the relative displacement. Means for calculating an entropy value of a value of a pixel near the pixel of interest in the image data; and updating the value of the pixel of interest based on the entropy value. Means, for each pixel of the image after performing the update, and a curvature calculating means for calculating a total curvature and average curvature in the pixel neighborhood.

[0021]

According to the present invention, in order to preserve other signals existing in the image data and remove only the impulse noise, each pixel of the image defined on the two-dimensional plane is calculated from the pixel values in the vicinity thereof. The value of the pixel is determined based on the calculated entropy value. That is, a so-called entropy filter is applied to the image data. Hereinafter, the configuration method and operation of the entropy filter will be described.

For an image f _{i, j} as a function defined on a two-dimensional lattice space (i, j) ∈Z ² , the value of a pixel in f _{i, j} is determined by the neighborhood N (i, j) ) Of the pixel f _{k, l} [k, l∈N (i, j)], based on the entropy ε _{i, j} calculated by the following equations (17) to (19), It is updated by a simple procedure.

[0023]

(Equation 5) Since the entropy ε of a pixel having impulse noise is 0, when ε _{i, j} = 0,

[0024]

(Equation 6) And the pixel f _{i, j} are updated, and when ε _{i, j} ≠ 0, f _{i, j} = f _{i, j}
(That is, the pixel value is not changed), only the impulse noise is removed. FIG. 6 is a diagram illustrating a difference in filter characteristics between the median filter and the entropy filter. When there is an input signal as shown in FIG. 6A, a signal as shown in FIG. 6B is obtained by the median filter, and as shown in FIG. 6C by the entropy filter. Signal can be obtained. From this figure, it can be seen that the median filter changes the shape of signals other than impulse noise, while the entropy filter removes only impulse noise. Therefore, the entropy filter preserves the local shape required for calculating the curvature. If such an entropy filter is applied to a recording / reproducing apparatus, a recording bit area can be accurately extracted, and the S / N of a recording bit read signal can be improved.
And the error rate is improved.

[0025]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a recording / reproducing apparatus according to one embodiment of the present invention.

A structure 11 made of an Invar alloy for supporting each structural part is provided. The recording medium 2 is attached to the bottom surface of the structure 11 via the coarse movement control mechanism 5. In the recording medium 2, information is actually written, and sclarylium-bis-6-octylazulene having a memory effect on current-voltage switching characteristics is formed on a graphite substrate by the Langmuir-Blodgett method ( LB method). The probe electrode 1 is arranged so as to face the recording medium 2.

The probe electrode 1 is used for recording and reproduction. The tip of the probe electrode 1 is obtained by mechanically polishing or electrolytic polishing the tip of a tungsten needle for improving the resolution of recording and reproduction. Of course, the material of the probe electrode 1 is not limited to this, for example, Pt-I
An r alloy, Pt, or the like may be used, and the processing method is not limited to these. The probe electrode 1 is attached to the inside of the upper surface of the structure 11 via the fine movement control mechanism 4. The fine movement control mechanism 4 is composed of a cylindrical piezoelectric element, and finely moves the probe electrode 1 in an in-plane direction (X, Y directions) of the recording medium 2 and in a direction between the recording medium 2 and the probe electrode 1 (Z direction). It is. During normal operation, the distance between the tip of the probe electrode 1 and the surface of the recording medium 2 is such that a tunnel current flows.

The structure 11 is mounted on the anti-vibration table 9 via a large coarse movement mechanism 10. The anti-vibration table 9 is for reducing noise caused by vibration of the apparatus during recording / reproduction. The large-coarse movement mechanism 10 controls a size outside the control range of the fine-motion control mechanism 4 and the coarse-motion control mechanism 5. It is for doing.

A voltage application device 8 for applying a bias voltage of an arbitrary magnitude in the range of -10 V to +10 V between the probe electrode 1 and the recording medium 2 is provided. The voltage application device 8 records / reproduces / reproduces information on / from the recording medium 2.
Generates a voltage for erasing. A current amplifier 3 that detects and amplifies a current flowing through the probe electrode 1 with this bias voltage is connected to the probe electrode 1. Further, an XY controller 6 for controlling the coarse movement control mechanism 5 and the fine movement control mechanism 4 and displacing the probe electrode 1 arbitrarily relative to the recording medium 2 is provided.

An arithmetic unit for controlling the XY controller 6 and the voltage application device 8 and for receiving the output of the current amplifier 3 is provided. The arithmetic unit 7 converts the output of the current amplifier 3 into a digital signal, a microprocessor for controlling each system, a position reference memory on the recording medium 2 described later, and a recording signal to recognize and encode the pattern. And an image processing unit for extracting only the recording bit area by regarding the measurement signal as an image signal.
The image processing unit includes an entropy filter as described later.

Next, the position reference memory on the recording medium 2 will be described. FIG. 2 is a schematic diagram showing a position reference pattern on a recording medium and a recorded signal area.

In this embodiment, a position reference pattern is recorded on the recording medium 2 before recording signal information. Before recording or reproducing signal information, the recording medium 2 is scanned by the probe electrode 1 to recognize the position reference pattern. As a result of this recognition, information regarding the position of at least one recording / reproducing area on the recording medium 2 and the scanning direction of the probe electrode 1 during recording / reproducing is obtained, and recording / reproducing is performed based on the obtained information. It has become.

In the example shown in FIG. 2, four position reference patterns 12 a to 12 d are provided on the recording medium 2.
The general recording area 20 is arranged adjacent to each of the position reference patterns 12a to 12d. Each of the position reference patterns 12a to 12d is recorded on the recording medium 2 before recording and reproduction.

Here, each position reference pattern 12a-12
The recording (setting) procedure of d will be described. During standby, due to the mechanical accuracy of the structure 11,
The probe electrode 1 is located at a position about 200 μm away from the home position hp in the Z-axis direction with an accuracy of ± 10 μm. Then, a voltage of 300 mV is applied between the probe electrode 1 and the recording medium 2 by the voltage application device 8, and in this state, the output of the current amplifier 3 is processed by the arithmetic device 7, and the large coarse movement mechanism is 10 is controlled so that the probe electrode 1 and the recording medium 2 are brought close to each other until a current of about 1 pA flows between them. Then, similarly, the fine movement control mechanism 4 is controlled so that the current between them becomes about 1 nA. Thereafter, the entire area surrounded by the dotted line in FIG. 2 is operated to write the position reference patterns 12a to 12d.

In the initial state, the entire area of the recording medium 2 is O
It is in the FF state. The probe electrode 1 is set so that the area where the position reference patterns 12a to 12d are to be recorded is turned on.
Is applied, a voltage of +10 V is applied between the probe electrode 1 and the recording medium 2. This voltage value is larger than the threshold value of the ON / OFF information of the recording medium 2, thereby forming a pattern as shown. Here, at the time of recording, the average current flowing between the probe electrode 1 and the recording medium 2 is set to 0.9 nA in order to prevent the recording operation from affecting the distance control between the probe electrode 1 and the recording medium 2.
The fine movement control mechanism 4 is controlled so that the average distance between the two becomes constant.

Next, recording / reproducing of data in the general recording area 20 will be described. Before recording and playback,
The probe electrode 1 is positioned on the position reference patterns 12a to 12d. First, a voltage of +5 V, which is smaller than the threshold value of the ON / OFF state of the recording medium 2, is applied between the probe electrode 1 and the recording medium 2. In this state, the fine movement control mechanism 4 and the coarse movement control mechanism 5 are controlled so that the probe electrode 1 scans the entire recording medium 2, and the ON / OFF state of the recording medium 2 is changed by the change in the tunnel current value at this time. Then, the position reference patterns 12a to 12d are detected. Since the position reference patterns 12a to 12d occupy an area sufficiently larger than the size of the unit recording bit in the general recording area 20, the ON / OFF state detected when the probe electrode 1 scans this pattern area. The power of the change frequency is concentrated in a lower frequency band than the power of the change frequency of the ON / OFF state detected in the general recording area 20. Therefore, the scanning detection signal from the position reference pattern and the scanning detection signal from the general recording area 20 are easily and approximately separated as low-frequency components and high-frequency components, respectively, using a simple band-pass filter. Based on the two-dimensional information of the plurality of position reference patterns, the probe electrode 1 is moved to around one of the four position reference patterns 12a to 12d.

Next, scanning is performed over a slightly wider area than the position reference pattern area, and the detection signal at that time is stored as two-dimensional information in a memory (not shown). As is clear from FIG. 2, since the direction pattern 13 arranged in the X direction and the Y direction is provided in the position reference pattern, the scanning direction at the time of recording / reproduction can be recognized by the direction pattern 13. Also, the position pattern 1 in the position reference pattern
The position of the pattern being detected can be known from the relative position of No. 4. That is, since the direction information and the position information are included in the position reference pattern, the information of the scanning direction at the time of actual recording / reproduction and the currently detected The position information of the pattern is obtained. Further, from the positional relationship between the recorded reference position patterns, information on the approximate position of the position reference pattern corresponding to the next general recording area can be obtained.

To record information in the general recording area, the probe electrode 1 is moved to a certain area of the general recording area 20 on the basis of the position information, and a wider area than the general recording area 20 is scanned. This is performed by detecting a synchronization signal of the synchronization pattern 15 and then writing information at a predetermined position based on the position information. At this time, the accuracy of the position is determined by the accuracy of the piezoelectric element of the fine movement control mechanism 4.

On the other hand, in reproducing information from the general recording area 20, first, the entire area after the synchronization signal is scanned to obtain two-dimensional information as surface shape data. In addition, the hysteresis characteristic of the piezoelectric element constituting the fine movement control mechanism 3 and the temperature / humidity expansion / contraction characteristic of the recording medium 2 are checked in advance. And
Considering these hysteresis characteristics and temperature and humidity expansion and contraction characteristics,
The acquired surface shape data is divided into unit information areas (areas corresponding to unit recording bits). Then, image processing described later is performed to reproduce the recording information ("1" or "0"). When recording / reproduction is completed, the probe electrode 1 is moved around the next pattern based on the obtained position information.

By the way, in the recording / reproducing apparatus of this embodiment,
The arithmetic unit 7 includes an entropy filter. The entropy filter is for removing impulse noise superimposed on the surface shape data. FIG.
FIG. 3 is a block diagram illustrating a configuration of an entropy filter. In this figure, an entropy filter for one-dimensional data is shown for simplicity of description, but an entropy filter for two-dimensional data is actually used.

Let the pixel of interest be f_i, The pixel f_iTo
The value f of the center three pixels_i-1, f_i, f _{i + 1}Store each
Registers 201 to 203 are provided. Each cash register
Each of the output sides of the star 201-203 has a subtractor 20
5₁~ 205_ThreeIs provided. In addition, each register 20
A minimum value circuit 204 to which outputs of 1 to 203 are input is provided.
Have been. The minimum value circuit 204 calculates the value f_i-1, f_i, f_{i + 1}of
Output the smallest of these. Each subtractor 205₁~ 205_Three
Output of each register 201-203 is input to + terminal of
The output of the minimum value circuit 204 is commonly input to the-terminal.
You.

The outputs of the subtracters 205 _{1-205 3} is connected to an input of each of the division circuit 207 ₁ to 207 ₃ of the dividend side is input to the addition circuit 206. The adder circuit 206 calculates the sum of the outputs of the subtracters 205 _{1 to} 205 ₃ . The output of the adder circuit 206 is connected in common to the input of the division circuit 207 ₁ to 207 ₃ of the divisor side. The output of each division circuit 207 ₁ to 207 _3, respectively inputted to a logarithmic circuit 208 _{1 to 208 3.} Each logarithmic circuit 208
_{1-208 3,} when the input and x, -x · log
(x). Each logarithmic circuit 208 _{1 to} 20
The output of the 8 ₃ is connected to the summing circuit 210 calculates and outputs the sum of these outputs.

Further, a judgment output circuit 209 is provided in the entropy filter. Judgment output circuit 20
9 has a third input terminal of the x ₁ ~x _3, x ₃ = 0 If x ₁
Is output, and if x ₃ ｘ0, x ₂ is output. The output of minimum value circuit 204 to the x ₁ terminal is connected, to the x ₂ terminal is connected the output of the register 202, x ₃
The output of the adding circuit 210 is connected to the terminal.

The entropy filter is configured as described above.
As a result, the output of the addition circuit 210 is_iNeighborhood of
Entropy value of ε_iBecomes Therefore, this entro
Pi value ε _iIs 0, the output value of the minimum value circuit 204 is
Entropy value ε_iIs not 0, the register 202
Output, pixel value f_iBut this entropy
Output from the filter. As a result,
As described in the “Action” section, only the impulse noise
Will be removed.

In the recording / reproducing apparatus of this embodiment, the value of each pixel is updated according to the entropy value for each pixel by the above-described entropy filter (however, a two-dimensional entropy filter). Then, the updated image data is sent to the recording bit recognition processing unit, and the recording bits are extracted. FIG. 4 is a block diagram illustrating the configuration of the recording bit recognition processing unit.

The recording bit recognition processing unit includes an input image storage unit 401 for storing an input original image, a total curvature calculation unit 402 for calculating a total curvature K of each point of the input image, and an average of each point of the input image. An average curvature calculator 403 for calculating the curvature H, a logical product calculator 404 for calculating the logical product of the output of the total curvature calculator 402 and the output of the average curvature calculator 403 for each point of the input image, AND operation unit 40 to remove
4 and a median filter 405 provided on the output side of the control unit 4. The total curvature calculation unit 402 calculates the total curvature K of each point of the input image. If the total curvature K is positive, the value of the point is 1;
Is performed. On the other hand, the average curvature calculation unit 403 calculates the average curvature H of each point of the input image. If the average curvature H is negative, the value of the point is 1, and if the average curvature K is non-negative, the value of the point is 0. Is performed. By configuring the recording bit recognition processing unit in this way, the output of the median filter 405 is image data in which only elliptical points are 1 and other points are 0. That is, only the recording bit area is correctly extracted.

Next, the operation of the recording bit recognition processing section will be described with reference to FIG. Here, a case will be described in which image data having a step or a ridge is directly input to the recording bit recognition processing unit. In FIG. 5A, the surface shape of the recording medium is shown as an original image 501 in contour display. On the surface of the recording medium, recording bits having a shape as shown in FIG. 7C are regularly arranged as an elliptical point area 502, and ridges as shown in FIGS. 7A and 7B. And steps are formed as parabolic point regions 503.
A hyperbolic point region 504 also exists due to the combination of the elliptical point region 502 and the parabolic point region 503, and the discretized spatial coordinates as the domain of the image.

The original image 5 is calculated by the total curvature calculating unit 402.
01 is the total curvature calculation unit output image 506, and the result of processing by the average curvature calculation unit 403 is the average curvature calculation unit output image 508. These images 50
6,508 and the images described below, the area of value 1 is shown in black and the area of value 0 is shown in white. From the total curvature calculation unit output image 506, it can be seen that the parabolic point region and the hyperbolic point region become isolated points by the processing of the total curvature calculation unit 402. Similarly, from the average curvature calculation unit output image 508, it can be seen that the processing by the average curvature calculation unit 403 results in a flat area that is convex downward and becomes an isolated point.

When the logical product operation unit 404 calculates the logical product of the total curvature calculation unit output image 506 and the average curvature calculation unit output image 508, an upwardly convex elliptical point region can be extracted. AND operation unit output image 510 output from AND operation unit 404
, A region belonging to a parabolic point region other than an elliptical point region is extracted, which is due to an error due to discretization of spatial coordinates or an influence of impulse noise. When the median filter 405 is operated to remove a parabolic area, an elliptical point extraction image 212 is obtained.

When observing the elliptical point extraction image 212 in detail, there is an image that is partially extracted from the elliptical point region 502 of the original image 501 and is extracted. This is because they were removed together in the process of removing the parabolic point region. In the recording / reproducing apparatus of this embodiment, since only the impulse noise is removed by the entropy filter before obtaining the total curvature or the average curvature image as described above,
There is no parabolic point region or hyperbolic point region in the original image to the recording bit recognition processing unit, and the elliptical point region can be extracted while maintaining its shape.

[0051]

As described above, according to the present invention, by applying an entropy filter to degraded image data on which impulse noise is superimposed, impulse noise can be removed without changing the characteristic amount inherent in the signal. There is an effect that the bit region can be accurately extracted, the S / N of the recording bit read signal is improved, and the error rate is improved.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a recording / reproducing apparatus according to an embodiment of the present invention.

FIG. 2 is a schematic plan view illustrating a position reference pattern.

FIG. 3 is a block diagram illustrating a configuration of an entropy filter.

FIG. 4 is a block diagram illustrating a configuration of a recording bit recognition processing unit.

FIG. 5 is a diagram illustrating the operation of a recording bit recognition processing unit.

6A and 6B are diagrams illustrating a difference in filter characteristics between an entropy filter and a median filter, wherein FIG. 6A illustrates an input signal, FIG. 6B illustrates a processing result by a median filter, and FIG. 6C illustrates an entropy filter. FIG. 9 is a diagram showing a processing result according to FIG.

FIGS. 7A to 7C are schematic perspective views showing the shapes of a ridge, a step, and a recording bit, respectively, formed on the surface of a recording medium.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Probe electrode 2 Recording medium 3 Current amplifier 4 Fine movement control mechanism 5 Coarse movement control mechanism 6 XY controller 7 Computing device 8 Voltage application device 9 Slow shake table 10 Large coarse movement mechanism 11 Structures 12a to 12d Position reference pattern 13 Direction pattern 14 Position Pattern 15 Synchronization pattern 20 General recording area 91 Ridge 92 Step 93 Recording bit 210 to 203 Register 204 Minimum value circuit 205 _{1 to} 205 ₃ Subtractor 206, 210 Addition circuit 207 _{1 to} 207 ₃ Division circuit 208 _{1 to} 208 ₃ Logarithmic circuit 209 Judgment output circuit 401 Input image storage unit 402 Total curvature calculation unit 403 Average curvature calculation unit 404 Logical product operation unit 405 Median filter 501 Original image 502 Elliptical point area 503 Parabolic point area 504 Hyperbolic point area

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Akihiko Yamano 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc. (72) Inventor Katsunori Hatana 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon (56) References JP-A-1-240984 (JP, A) JP-A-5-264232 (JP, A) JP-A-6-259546 (JP, A) JP-A-63-161552 (JP, A A) JP-A-63-161553 (JP, A) JP-B-63-63947 (JP, B2) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06T 5/20 G11B 9/00 H04N 5 / 213 H04N 5/93

Claims (2)

(57) [Claims] A probe electrode provided to face a recording medium; voltage applying means for applying a voltage between the probe electrode and the recording medium; Displacing means for displacing, reproducing means for reproducing at least information from the recording medium based on physical interaction between the probe electrode and the recording medium in accordance with the relative displacement, and reproducing from the reproducing means. A recording / reproducing apparatus comprising: a signal processing unit configured to remove a noise from the reproduction signal by regarding the signal as two-dimensional image data based on the relative displacement. Calculating means for calculating the entropy value of the value of the pixel; updating means for updating the value of the pixel of interest based on the entropy value; and updating the image after the updating. For pixels, recording and reproducing apparatus characterized by having a curvature calculating means for calculating a total curvature and average curvature in the pixel neighborhood. 2. The recording / reproducing apparatus according to claim 1 , wherein the physical interaction is a tunnel current.