JP2000111333A - Correction method of linearity measuring apparatus and master gauge for correction - Google Patents

Abstract

PROBLEM TO BE SOLVED: To permit check of image processing of an linearity measuring apparatus by using a master gauge of simple structure which corresponds to a various kinds of pattern forms. SOLUTION: In this correction method of a linearity measuring apparatus which picks up the image of a magnetic tape of a helical track and measures linearity of a magnetized pattern by the image processing, a master gauge constituted of a plurality of parallel trenches 21 at intervals corresponding to the intervals of a magnetized pattern when scanning is performed along the tape width direction while corresponding to a mangetized pattern 2 having a specified form is previously formed. The image of the master gauge 20 is picked up and processed, thereby checking whether the image processing of a linearity measuring apparatus is adequate.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for calibrating a linearity measuring device and a master gage for calibration. More specifically, the present invention relates to a calibration method and a calibration master gauge for a linearity measuring device that measures linearity by image analysis including Fourier transform.

[0002]

2. Description of the Related Art In a helical scan recording system used as a magnetic recording system, recording and reproducing tracks are formed on a magnetic tape 1 as shown in FIG. The magnetization pattern 2 is inclined by the angle α. In order for the magnetic tape 1 composed of such oblique magnetization patterns 2 to maintain reproduction compatibility with other devices, it is necessary to determine how accurate the inclination angle α is.
And how exactly it is linear is an important issue. Therefore, the accuracy of such linearity is strictly defined by format specifications and the like in various formats, and the helical scan recording device must measure the linearity strictly and guarantee the linearity.

In order to measure the linearity of such a helical scan recording apparatus, conventionally, a colloid solution containing magnetic particles is first applied to a magnetic tape to be measured to visualize a magnetization pattern, and the magnetic tape is mounted on a stage. Set on the stage of a microscope with a microscope, observe the edge of the recorded pattern with a high magnification microscope, move the stage when observation is completed, observe the edge of the next pattern, and use the stage scale to determine the interval Measurement method.

However, in the above-described conventional linearity measuring method, the edge of the magnetic recording pattern is very difficult to see, and it is easy to generate a large error in observation with human eyes. It is necessary to use a lens, but there is a limit to the optical microscope, so that a very large magnification cannot be used, and there is a technical limit in increasing the mechanical accuracy of the stage. There is a problem in that costs are incurred, and since all measurements are performed manually, a large amount of time is required for one measurement, resulting in poor work efficiency.

Therefore, in order to address such a problem, the present applicant has proposed a linearity measuring method in which a magnetization pattern is imaged by a camera, the image is processed, and the linearity is measured. This method visualizes the recording pattern of the magnetic tape, illuminates the visualized magnetic tape with light from a light source whose irradiation angle is adjustable, captures an image of the illuminated magnetic tape with a camera, and analyzes the captured image. Is used to measure the linearity of the recording pattern.

According to this method, the linearity can be measured without manual operation by performing image processing on a recording pattern picked up by a camera, and errors due to human eyes and errors due to the accuracy of the stage are eliminated, thereby achieving high precision. Measurement can be performed, and manual labor can be omitted, and highly accurate measurement can be performed in a short time without skill.

[0007]

However, in such a linearity measuring device equipped with a camera and an image processing device, image processing may not be performed properly due to a change over time or a set state during use. Therefore, before measuring the linearity, it is necessary to check and calibrate whether the linearity measuring device is always performing a certain image analysis process. In this case, a pattern of the same shape as the magnetic tape whose pattern shape is known in advance is stamped on a helical concave and convex shape to produce a master gauge, the master gauge is photographed by a camera, and image processing is performed. By comparing the image processing data to be obtained from the known pattern with the data obtained by actually performing image processing on the master gauge, it is possible to check the image processing by the linearity measuring device.

However, a master gauge on which such an oblique magnetization pattern having the same shape as the magnetic tape is formed cannot easily form various bending patterns by machining, and is very expensive.

The present invention has been made in view of the above points, and has a linear measuring apparatus capable of checking the image processing of the linear measuring apparatus using a master gauge having a simple structure and corresponding to various pattern shapes. The purpose is to provide a calibration method and its master gauge.

[0010]

In order to achieve the above object, the present invention provides a method for calibrating a linearity measuring apparatus for imaging a magnetic tape of a helical track and measuring the linearity of a magnetization pattern by image processing. In correspondence with the magnetized pattern of a predetermined shape, when scanning along the tape width direction, a master gauge consisting of a plurality of parallel grooves at intervals corresponding to the interval of the magnetized pattern is prepared in advance,
A method for calibrating a linearity measuring device, characterized in that it is checked whether image processing of the linearity measuring device is appropriate by taking an image of the master gauge and performing image processing.

According to this method, a master gauge composed of simple parallel grooves corresponding to the pattern arrangement in the width direction of the tape is used without forming the entire shape of the helical magnetization pattern. And the cost can be reduced.

[0012]

FIG. 1 is an overall configuration diagram of a tape linearity measuring apparatus according to an embodiment of the present invention. On the inspection surface 3a on the upper surface of the measurement stage 3 composed of an XY table,
A magnetic tape 4 to be measured is placed. The magnetic tape 4 has a magnetized pattern 2 visualized by previously applying a colloid solution containing magnetic particles. Above the measurement stage 3, a CCD camera 5 is provided. The CCD camera 5 is connected to an A / D board 8 of a personal computer 7 via a driver 6. The CCD camera 5 has a magnetic tape 4
And scans in the width direction thereof and sends a waveform signal corresponding to the magnetization pattern 2 to the image processing circuit 9 of the personal computer 7 via the A / D board 8. The image processing circuit 9 includes an FFT 10 that performs Fourier transform processing of the transmitted image waveform signal, an analysis circuit 10 that performs inverse Fourier transform and other signal processing to create data corresponding to the original image, It is composed of a pasting circuit 12 for synthesizing the data of each part and obtaining a waveform signal of the entire scanning line.

The image processing circuit 9 is connected to a driving circuit (not shown) of the measuring stage 3 via the GPIB board 13 and controls the driving of the measuring stage 3. The image processing circuit 9 is connected to the monitor 14 and displays the image data of the magnetic tape 4 on which an image has been analyzed on the screen. Note that CC
A digital signal can be directly sent from the D camera to the image processing circuit 9 without passing through the A / D board. G
The drive of the measurement stage 3 can be controlled via another communication cable without using the PIB board.

In such a linearity measuring device, the magnetic tape 4 on the measuring stage 3 is irradiated with light from, for example, obliquely above, and the magnetic tape 4 is scanned in the tape width direction by the CCD camera 5 to thereby scan the magnetic tape 4. An image is fetched, the image is analyzed by an image processing circuit 9 of the personal computer 7, and the linearity is determined from the image data. In this case, by selecting the irradiation angle of the light on the magnetic tape 4, it is possible to form the stripe pattern by dividing the track of the adjacent pattern into black and white based on the properties of the colloid particles.

Next, a description will be given of a method of recognizing a black and white boundary of each pattern of an image captured by the CCD camera 5 with high accuracy. FIG. 2 shows a black-and-white stripe pattern of the magnetic tape 4 captured by a high-resolution CCD camera 5 and displaying the digital data on a personal computer. FIG. 3 shows a graph of the pixel numbers and the brightness when only one line is scanned in FIG. 2 in the horizontal direction. Here, it is considered to measure the linearity of the recorded tape. The linearity refers to how the recorded pattern follows the ideal inclination straight line. From a different viewpoint, the interval between the rising edges in the graph of FIG. It also means that it is.
Therefore, it is required to find the rising edge of this graph as accurately as possible, with a finer precision than the resolution of one pixel. Therefore, the following algorithm was devised.

An average value of the brightness of 1.1 lines is obtained. 2. It searches from the young pixel number and finds a place that straddles the average value of 1. An average value is calculated from a position shifted by a default pixel at the center of 3.2 to a position shifted by a younger default value and a position shifted by a larger default value. Find a point that crosses the value of 3 near 4.2. The area around 5.4 is linearly complemented, and a pixel point corresponding to the average value of 3 is found.

In this method, since the average of the front and rear black and white, rather than the entire average, is used for the threshold, the influence of the change in the degree of lightness and darkness within one screen as described in the above-described illumination method is used. The edge point can be obtained with high accuracy without receiving it.

Actually, since the average of the black portion and the average of the white portion are required, the second edge is obtained instead of the first edge. Its accuracy is 10
It is highly accurate such that there is only an average shift of 0.06 pixels in scanning of 0 lines.

Next, a method for measuring the degree of bending of a recording pattern will be described. With respect to the graph of FIG. 3, if all the edges are obtained and the intervals are measured as described above, it is possible to obtain the linearity therefrom. However, in practice, this calculation takes a long time, and there are problems such as being easily affected by development unevenness, dirt, and dust, and an error is likely to occur, so that accurate measurement cannot be performed.

Therefore, a moire observation method using a Fourier transform is used to measure the degree of bending. This method is performed by the following algorithm as shown in FIG.

1. Fourier transform the input data. 2. Only the data near the known reference frequency is extracted. 3. Data is moved on the frequency axis so that the reference frequency becomes a component of 0 Hz. 4. Inverse Fourier transform. 5. Averaging is performed to remove noise.

Thus, the output data becomes the data of the degree of bending, and indicates the position VS of the pixel and the amount of bending. Actually, one pixel = X μm is obtained in advance, and by multiplying this number, the actual bending degree is obtained, and the position VS bending degree (μm) in the tape is indicated.

Next, a sampling method for measuring the degree of bending will be described. Although the bending degree can be obtained by complementation by the bending degree measuring method, the following points are problematic.

1. If the sampled data is not exactly the data of the reference frequency, the data is greatly affected by aliasing and becomes data with a large error. 2. If the sampled data is not composed of 2 ⁿ , it takes a long time to calculate because FFT cannot be used.

Therefore, as shown in FIG. 5, the following problem is solved by the following algorithm. 1. The first edge is obtained by the algorithm of the method for recognizing the black and white boundary line with high accuracy. 2. Similarly, the last edge is obtained by the above algorithm. 3.1 Edge—The data between the first edge and the second edge + 1 is complemented by a spline. 4. Resample at (2 edges-1 edge) / 2 ⁿ .

By re-acquisition of data by this algorithm, data of the reference frequency can be accurately acquired from edge to edge, so that the influence of aliasing is eliminated and no calculation error occurs. In addition, since the number of data is 2 ⁿ , FFT can be used, and the calculation time can be significantly reduced. As a result, calculation errors and the like are reduced, but conversely, in this method, since only data corresponding to the reference frequency is used accurately, the component of the slope of linearity becomes invisible. Therefore, the inclination is obtained by the following method.

1. Observe the master in advance,
The number of pixels corresponding to the length of the reference frequency is measured. (M pixels) Find the interval between the last edge and the first edge. (X
Pixel) 3. The slope is calculated by the following equation. (X pixel−M pixel) / M
Here pixels * ^{2 n, 2 n} is 2 ⁿ using the sampling. Thus, the inclination for the 2 ⁿ division can be obtained. In order to convert this value from the sampling number to μm, first, it is necessary to find the correspondence between the number of pixels and the length. That is, the length (L) of the master measured in 1 is measured in advance, and 1 pixel = L / M μm can be obtained from M pixels = L. Next, since the X pixels are re-sampled at 2 ⁿ this time, 1 sampling number = X / 2 ⁿ pixels from the X pixel = 2 ⁿ sampling numbers. Collectively, one sampling number = X * L / M / 2 ⁿ μ
Data obtained by the correction formula of m can be converted to μm. Actually, since an error occurs at the time of complementation, the degree of bending is obtained over a plurality of tracks, and an average value between the plurality of tracks is obtained as data of the image.

Next, a method of pasting a plurality of data will be described. The inclination and the degree of bending in one screen can be obtained by the evaluation method up to now, but it is not possible to put all the tracks to be obtained in one image for analysis for the following reasons.

1. It is difficult to select a lens with an appropriate magnification to exactly include the track to be analyzed in one image. 2. If you put too many tracks in one image, CC
Since the number of pixels of D cannot keep up with the track, an input image cannot be obtained with good image quality. 3. When a very large number of tracks are included in one image, the number of pixels forming one black and white is reduced, the length per pixel is increased, and the resolution per pixel is deteriorated. 4. When linearity is obtained by Fourier transform, the number of pixels in one black and white becomes small, so that an analysis error increases.

For this reason, it is not possible to include all tracks in one image. Therefore, one data is obtained from a plurality of images. For that purpose, how many tracks should be included in one image is determined. This is because it is not preferable to include too many tracks as described above. Conversely, having too few tracks is undesirable for the following reasons.

1. To analyze every track,
Many images need to be input, and analysis takes a very long time. 2. The fact that there are only a few tracks in one image means that the number of data on the frequency axis after passing through the bandpass filter after the Fourier transform decreases, and the error after analysis increases.

That is, there is an optimum number of tracks in one image. As a result of the study, it has been found that in the present invention, the optimum number of tracks is a state where there are about 15 to 100 tracks at the time of 1024 samplings. As described above, since the number of tracks in one image is limited, it is necessary to combine analysis data of a plurality of images to obtain an overall analysis result.

Therefore, when considering combining data obtained from each image (FIG. 6), first consider data obtained from one image. With this data, the degree of bending and inclination from the starting point can be obtained. This data is data obtained with the start point being 0, and is data from the first edge to the last edge.

Next, consider the case where the object to be measured is moved and the next image is analyzed. At this time, the object to be measured is moved so that the last edge analyzed last time moves near the first edge of the current image. When the analysis is performed in the same manner, it is possible to obtain the degree of bending and the inclination starting from the last edge of the previous analysis. The result is similarly data with the starting point set to 0. That is, this data is data with the last point of the data analyzed last time being 0. That is,
Just by setting the first value as the last value of the result of the previous analysis, continuous data is obtained as it is.

At this time, since the movement amount of the object to be measured is irrelevant, it may be rough, and may be of such a degree that the last track of the previous analysis can be determined to be the first track of this time.
This can be as high as the track width, which is much larger than the required precision (about 0.01 μm) (VHS
(Approximately 60 μm), which is sufficiently rough on a general-purpose stage or the like. That is, unlike the conventional measuring method, an extremely expensive stage is not required, and an inexpensive system can be constructed.

As described above, in the linearity measuring device for detecting the curve of the track by performing the image processing by the Fourier transform on the image captured by the CCD camera, is the image processing data by this measuring device correctly processed? In order to check whether or not this is the case, several types of master gauges corresponding to curved tracks are prepared in advance, and the master gauges are measured by this measuring device. The accuracy control of the master gauge is performed by a high-accuracy mother measuring machine.

Next, the master gauge will be further described. In a helical scan magnetic recording device, track deviation occurs on a magnetic tape due to a unique frequency component of each component of a rotating drum used for feeding a magnetic tape. In the present invention, in a linearity measuring device for a helical track on a magnetic tape using the Fourier transform method, a master gauge corresponding to the track deviation of each frequency component is prepared, and each master gauge is measured. Check and calibrate the instrument.

FIG. 7 shows an example of such a master gauge. (A) is a straight magnetization pattern 2 and has a constant angle α.
Indicates a track inclined straight and parallel. When the magnetic tape of the magnetization pattern 2 is scanned in the width direction of the tape as shown by the arrow S, each pattern is detected at equal intervals along the scan line.
The master gauge 20 is formed such that 0 has parallel grooves 21 at equal intervals.

(B) shows a shape in which the magnetization pattern 2 is curved in one direction, and the frequency of a track bent at sinΘ / 2 (Θ is an angle corresponding to the inclination of the drum axis) due to the inclination of the drum axis with respect to the reference plane. The components are shown. When the magnetic tape of the magnetized pattern 2 is scanned in the tape width direction as shown by the arrow S, each pattern is detected along the scan line so that the upper portion becomes denser. A master gauge 20 consisting of narrow parallel grooves 21 is formed.

(C) shows a frequency component of a track in which the magnetization pattern 2 is curved in two directions in a wave-like manner, and is bent by sin に 対 し with respect to the reference plane. When the magnetic tape of the magnetized pattern 2 is scanned in the tape width direction as shown by the arrow S, a master gauge 20 is formed comprising parallel grooves 21 having a width corresponding to the width of each pattern detected along the scan line. I do.

FIGS. 8 and 9 are views for explaining a method of manufacturing a master gauge corresponding to a straight track pattern. As shown in FIG. 8, consider a magnetic tape 4 on which a straight magnetization pattern 2 is formed. This magnetic tape 4 is indicated by an arrow S
The edge position of each track is projected on a scan line crossing at right angles as shown in (1), and a new edge is provided on the master gauge 20 based on the projected edge position. In this figure, the hatched portion has a concave shape, but the reverse is also possible, and the edge portion may be formed in a shape suitable for the processing method. However, the illumination and the like are arranged so that the track of azimuth on either one side of + or-may be raised on the principle of measurement. For example, the surface of the master gauge is provided with a surface roughness of a cutting process such that a pattern rises when light is irradiated from obliquely above the magnetic tape.

Further, as shown in FIG. 9, the recessed groove 21 is carved deeply so that even if light is irradiated, there is no reflective object and the light becomes a shadow. The size of the master gauge 20 is determined according to the tape. For example, the width H is 3 mm and the height B is 5
mm and the thickness T are about 0.5 mm. As shown in FIG. 9, such accuracy control of the master gauge 20 is performed by measuring the distances X, X2, X from the lower edge to the edge of each groove 21.
3,. . . It is managed by measuring Xn with high accuracy. The central portion of the lower edge is formed straight as a reference, and the straight portion of the lower edge is used as a measurement reference.

Such a master gauge 20 can be formed by machining, etching, photoetching or other various methods. The material is made of, for example, stainless steel, which does not change with time, maintains high accuracy for a long period of time, and has sufficient strength and corrosion resistance.

The master gauge shown in FIG.
In addition to the types of shapes, various bending components are assumed, such as when the track has an inclination, when it vibrates based on the structural mechanical vibration component, and when the rotating system fluctuates with the frequency component of the structural component. More accurate calibration can be performed by preparing several types of master gauges.

[0045]

As described above, according to the present invention, a master gauge consisting of simple parallel grooves corresponding to the pattern arrangement in the width direction of the tape is formed without forming the entire shape of the helical magnetization pattern. Because of this, the degree of freedom in processing methods and material selection is increased, making it easy to manufacture master gauges.Calibration of linearity measurement devices using low-cost, simple structure master gauges without using expensive master gauges Is performed, and highly reliable measurement data can be obtained.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a linearity measuring device according to the present invention.

FIG. 2 is a photograph of an input image of a recording pattern.

FIG. 3 is a graph formed by performing image processing on the pattern of FIG. 2;

FIG. 4 is an explanatory diagram of a linearity measurement method using a Fourier transform.

FIG. 5 is an explanatory diagram of a data sampling method.

FIG. 6 is an explanatory diagram of a data bonding method.

FIG. 7 is a view showing a shape example of a master gauge according to the present invention.

FIG. 8 is an explanatory view of a method for manufacturing a master gauge of the present invention.

FIG. 9 is an explanatory view of the shape of the master gauge of the present invention.

FIG. 10 is an explanatory diagram of a helical magnetization pattern.

[Explanation of symbols]

1, 4: magnetic tape, 2: magnetization pattern, 3: measurement stage, 5: CCD camera, 7: personal computer, 9: image processing circuit, 10: FFT, 11: analysis circuit, 12: bonding circuit, 20: master gauge , 21: groove.

 ──────────────────────────────────────────────────続 き Continuing on the front page F term (reference) 2F065 AA45 BB13 BB24 BB28 CC03 DD03 DD06 FF04 FF61 JJ19 MM16 MM26 PP11 PP21 QQ16 QQ31 RR05 SS13 UU05 2F069 AA55 AA62 BB13 CC06 DD15 DD25 GG04 GG07 GG15 GG07 GG07 GG07 GG07 GG07 GG07 GG07 GG07 GG07 GG07 GG15 GG15 NN08 PP06 QQ05 QQ10 RR05 2G051 AA90 BA00 CA03 CA04 CB06 DA07 EA02 EA12 EC03 ED07 5B057 DA03 DA07 DB02 DC02 5D112 AA22 JJ06

Claims (2)

[Claims] 1. A method for calibrating a linearity measuring device for measuring the linearity of a magnetization pattern by imaging a magnetic tape of a helical track and performing image processing on the magnetic tape, comprising the steps of: By preparing in advance a master gauge consisting of a plurality of parallel grooves at intervals corresponding to the interval of the magnetization pattern when scanning along the line, imaging the master gauge and performing image processing, the image processing of the linearity measurement device can be performed. A method for calibrating a linearity measuring device, characterized by checking whether the linearity is appropriate. 2. A master gage for calibrating a linearity measuring device comprising a plurality of parallel grooves having an interval corresponding to an interval of a magnetization pattern when scanning along a tape width direction corresponding to a magnetization pattern of a predetermined shape. .