KR100681609B1 - Method for over-sampling to compensate align error in a hdds system - Google Patents

Abstract

The present invention relates to an oversampling method for overcoming an alignment error that increases a signal-to-noise ratio in a HDDS system and reduces the probability of an error that may occur when decoding data. That is, according to the present invention, in the oversampling method in the HDDS system, when an alignment error between the playback image and the measuring device occurs, the edge of the CCD pixel corresponding to one pixel of the SLM is detected according to the pixel similarity ratio of the SLM and the CCD. By compensating for the alignment error by determining the number and position of pixels to be extracted by comparing the brightness of the pixel lines constituting the detected edge, the signal-to-noise ratio can be maintained above a certain level even when there is an alignment error in the horizontal or vertical direction. In addition, the combination with the existing oversampling method for correcting the error due to the size change enables more accurate alignment error compensation.

Description

Oversampling Method for Overcoming Alignment Error in HDDS System TECHNICAL METHODS FOR OVER-SAMPLING TO COMPENSATE ALIGN ERROR IN A HDDS SYSTEM

1 is an exemplary reproduction image illustrating an oversampling method in a conventional HDDS system.

2 is a diagram illustrating a horizontal error occurrence of a playback image in a conventional HDDS system;

3 is a diagram illustrating a signal to noise ratio change graph according to a horizontal alignment error in a conventional HDDS system;

4 is a diagram illustrating a playback image showing misalignment compensation in an HDDS system according to an embodiment of the present invention;

5 is a flowchart illustrating a misalignment compensation process in an HDDS system according to an embodiment of the present invention;

6 and 7 are graphs illustrating a change in a signal-to-noise ratio graph according to a horizontal alignment error in an HDDS system according to an exemplary embodiment of the present invention.

8 is a flowchart illustrating a misalignment compensation process in an HDDS system according to another exemplary embodiment of the present invention.

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a holographic digital data storage (HDDS) system, and more particularly to an oversampling method that overcomes misalignment by reducing the probability of errors that may occur during data decoding by increasing the signal-to-noise ratio in an HDDS system.

In general, HDDS system is a page-oriented memory that uses the principle of volume hologram on the principle of data recording / reproducing, and can use the parallel data processing method as the input / output method to make the input / output speed higher than 1Gbps. In addition, it is possible to configure the system without the mechanical driving part, so that the data access time can be implemented very quickly with less than 100 ms.

On the other hand, the HDDS system detects and decodes the page unit information image recorded on the storage medium by the CCD during reproduction. The image data of the unit unit reproduced on the storage medium, that is, the binary pixel data, is originally stored through the CCD. In order to restore the data, a 1: 1 matching signal processing is performed, and the pixel of the image reproduced on the storage medium and the pixel of the CCD array are matched 1: 1 to restore the original pixel data.

That is, in the conventional oversampling decoding method, when the SLM (Spatial Light Modulator) and CCD pixel similarity ratios are 3: 1, an image of one pixel of the SLM is formed on the 9 pixels of the CCD. By selecting one of the brightest pixels out of the nine pixels and using that value as the signal, we were able to increase the signal to noise ratio. At this time, select 1 pixel of the image to select the brightest line in the horizontal and vertical direction to find the border, and then move the value by 3 pixels horizontally and 3 pixels vertically based on the position. It was conventionally used.

In addition, in the conventional oversampling method, when the pixel similarity ratio between the SLM and the CCD is 2: 1, it is formed in four pixels of the CCD as shown in FIG. 1, and when one pixel is selected, the horizontal or vertical generated as shown in FIG. 2. There is a problem that the signal-to-noise ratio is greatly worsened by the direction misalignment as shown in FIG.

Accordingly, an object of the present invention is to provide an oversampling method for overcoming an alignment error that increases a signal-to-noise ratio in a HDDS system to reduce a probability of an error that may occur when decoding data.

The present invention for achieving the above object is an oversampling method for overcoming an alignment error in a HDDS system, and (a) SLM according to the pixel similarity ratio of the SLM and CCD when an alignment error occurs between the playback image from the CCD and the measuring device. Detecting an edge of a CCD pixel corresponding to one pixel of the pixel; and (b) compensating for an alignment error by determining the number and position of pixels to be extracted by comparing brightness of pixel lines constituting the detected edge. Characterized in that.

Hereinafter, with reference to the accompanying drawings will be described in detail the operation of the preferred embodiment according to the present invention.

4 is a diagram illustrating an example of a pixel image detected by a CCD when a pixel similarity ratio between a spatial light modulator (SLM) and a charge coupled device (CCD) is 2: 1 according to an embodiment of the present invention, and FIG. 5 is an embodiment of the present invention. Shows an oversampling method for reducing alignment error in the HDDS system. Hereinafter, the oversampling method of the present invention will be described in detail with reference to FIGS. 4 and 5.

First, in the present invention, when the pixel similarity ratio of the SLM and the CCD is 2: 1, a predetermined region is selected to determine a place where the two neighboring lines are the brightest. In order to find the brightest place, there is a method of comparing two neighboring lines together in order or finding the brightest line and comparing the left and right values of the found line (S500). 4 shows an example of determining a border using two neighboring lines, and according to the comparison of the two lines added together, “a” and “me”, “da” and “d” are judged as borders ( S502). Then, two lines forming a border should be selected to correspond to the brightness difference, wherein the magnitude of the brightness difference is large and small by Equation 1 below, and the constant k in Equation 1 is 1/2. Use a close number and depends on the experimental environment (S504).

| I-I | > k (ga + na)

In this case, when the brightness difference between the two lines forming the border is large according to the result of [Equation 1], only one pixel line of “A” or “I” shown in FIG. 4 is selected (S506), and the brightness difference is If small, the value obtained by adding the two pixel lines of "A" and "I" is used as a signal (S508).

Hereinafter, operation in each condition will be described in more detail. First, if the value of | ga + a | is greater than the value of k (ga + b), the pixel lines with larger k and later values are the other two pixel lines of the border perpendicular to the pixel line. The signal value is extracted while skipping by 2 pixels in the horizontal and vertical directions based on the value of the pixel that meets each other. Also, when the value of | da + la | is greater than the value of k (da + la), A pixel line with a large value is the pixel line. The pixel value is extracted by adding two pixels while skipping by two pixels based on the position of two pixels that meet the D.

In addition, in the case of | da-la | <k (da + la) or | da-la | = k (da + la), the pixel line A and the later pixel line having a larger value meet the pixel line D and D. A signal value is extracted by adding two pixels while skipping by two pixels based on the position of two pixels, and the | Ga-na | <k (ga + na) or | ga-na | = k (ga + na) Is a value of 2 pixels, skipping by 2 pixels in the horizontal and vertical directions, based on the value of the pixel that meets the other two pixel lines of the border perpendicular to the pixel line. Extract the sum plus the signal value.

In addition, when the value of | da-la | is greater than the value of k (da + la), the k and b pixel lines are based on the position of two pixels where the pixel line meets the pixel line with the larger value. A value obtained by adding a value of 2 pixels while skipping by 2 pixels is extracted as a signal value, and in the case of | da-la | <k (da + la) or | da-la | = k (da + ra), Kawa B. Two pixel lines are skipped by 2 pixels on the basis of the position of 4 pixels where the pixel lines meet each other, and 4 pixel values are extracted as signal values.

In this way, when the brightness difference between the two pixel lines of the selected edge is determined according to FIG. 4, if the brightness difference in both the horizontal and vertical directions is small, the pixels of "a + b + c + d" are oversampled. Extract as a signal value.

That is, when the determination is made according to the processing flow of FIG. 5, the signal value is 1 pixel, 2 pixels, or 4 pixels.

If there is a big difference in both the horizontal and vertical directions, select "a" pixels and so on. In the case of a big difference in the horizontal direction and small differences in the vertical direction, select "a + bs" or "b + d". If the difference in both directions is small, the pixel of "a + b + c + d" is selected as a signal value.

FIG. 6 illustrates a signal-to-noise ratio graph when selecting a signal obtained by adding four pixels while giving a horizontal alignment error, and FIG. 7 illustrates a signal-to-noise ratio graph when the proposed method is applied. As shown in FIG. 7, it can be seen that an effect such as selecting a case in which a signal-to-noise ratio is better among the cases in which only a value of 1 pixel and a value of 4 pixels is added is shown.

8 illustrates an oversampling method for reducing alignment error in an HDDS system according to another embodiment of the present invention.

When the pixel similarity ratio between SLM and CCD is 3: 1, a certain area is selected to determine the brightest part of the three lines of pixels as the border. When the brightest spot is found, as shown in the sampling pixel shown in FIG. 9, the values obtained by adding the pixel lines of three lines “a, b, and d” are sequentially compared (S800). At this time, according to the comparison of the sum of the three lines, "ga" and "da" is determined as the frame (S802). Then, the two pixel lines constituting the edge should be selected to correspond to the brightness difference, wherein the big and small of the brightness difference are determined by Equation 2 below (S804).

| A-da | > k (more)

In this case, when the brightness difference between the two pixel lines forming the edge is large according to the result of [Equation 2], the pixel line of "ga + na" or "na + da" is selected (S806). If small, the value obtained by adding the "I" pixel line located in the center of the sampling pixel 900 is used as the signal (S808).

Hereinafter, operation in each condition will be described in more detail. First, if the value of | a-da | is greater than a value of k (more), the pixel line located in the middle of the three pixel lines is the other three pixel lines of the border perpendicular to the pixel line. The signal value is extracted by skipping 3 pixels in the horizontal and vertical directions based on the pixel value that meets the bar. When the value of | la-bar | is greater than the value of k (la + bar), three pixel lines are extracted. A pixel value of 1 pixel is extracted as a signal value by skipping 3 pixels based on the position of 1 pixel where the pixel line located in the center of the pixel meets the pixel line Ma located in the center of the other three pixel lines perpendicular to the pixel line.

In the case of | la-bar | <k (la + bar) or | la-bar | = k (la + bar), the other three pixels perpendicular to the pixel line out of the pixel line located in the middle of the three pixel lines. The alignment error is compensated by extracting two pixels as a signal value while skipping by three pixels based on the position of two pixels that meet the pixel line with the large pixel line, which is the pixel line at both ends of the line.

As described above, in the present invention, in the oversampling method, not only a value of 2 pixels or 4 pixels may be added but also a value obtained by dividing the added value by a constant constant or a value obtained by multiplying a value of each pixel by a constant constant, and the like. It is also possible to use a combination of the methods of the present invention and other oversampling methods previously proposed. At this time, if the size of the image becomes larger or smaller as a whole, the typical method that can be combined is calculated by increasing the size of the image, and the position of pixels extracted from the points such as 1/2, 1/3, 2/3, etc. of the image is increased by one pixel. There is a way to move. In combination with this method, we apply the proposed method according to the values determined at the four different corners of the image, extracting the value of 1 pixel until reaching a certain position, and then adding the value of 2 pixels or 4 pixels after the specified position. You can use a method to extract the values.

Meanwhile, in the above description of the present invention, specific embodiments have been described, but various modifications may be made without departing from the scope of the present invention. Therefore, the scope of the invention should be determined by the claims rather than by the described embodiments.

As described above, according to the present invention, in the oversampling method in the HDDS system, when an alignment error between the reproduced image and the measuring device occurs, the edge of the CCD pixel corresponding to one pixel of the SLM according to the pixel similarity ratio of the SLM and the CCD After detecting, the number and position of pixels to be extracted are determined by comparing the brightness of the pixel lines that form the detected edge to compensate for the alignment error, so that even if there is an alignment error in the horizontal or vertical direction, the signal-to-noise ratio exceeds a certain level. There is an advantage that can be maintained. In addition, a more accurate alignment error compensation is possible through a combination with an existing oversampling method for correcting an error due to a size change and an error due to rotation.

Claims (13)

An oversampling method that overcomes misalignment in HDDS systems, (a) detecting an edge of a CCD pixel corresponding to one pixel of the SLM according to the pixel similarity ratio between the SLM and the CCD when an alignment error between the reproduced image from the CCD and the measuring device occurs; (b) compensating for an alignment error by determining the number and location of pixels to be extracted by comparing the brightness of the pixel lines forming the detected edges; Oversampling method for overcoming alignment error in the HDDS system comprising a. delete The method of claim 1, In the step (b), when the pixel similarity ratio between the SLM and the CCD is 2: 1, the value of | ga-i |, which is the difference between two pixel lines forming a border, and the sum of the two pixel lines is multiplied by a constant constant (k). An oversampling method for overcoming alignment errors in an HDDS system, characterized by determining the number and position of pixels by comparing a value of k (ga + b). The method of claim 3, wherein If the value of | ga + na |, which is the sum of the two pixel lines A and I, is greater than the value of k (ga + b), then the other two of the borders perpendicular to the pixel line are larger than the pixel line. Oversampling method for overcoming misalignment in HDDS system, characterized by compensating for misalignment by extracting signal value by skipping 2 pixels in horizontal and vertical direction based on the pixel value that meets the pixel line Dawara . The method of claim 4, wherein If the value of | da + ra |, which is the sum of both other pixel lines of the border and la, is greater than the value of k (da + ra) multiplied by a constant constant k, the sum of the two pixel lines is In the HDDS system, a pixel line having a larger value is the pixel line, and the pixel line is skipped by 2 pixels based on the position of 2 pixels that meets D, and the 2S is added as a signal value to compensate for the alignment error. Oversampling method to overcome errors. The method of claim 4, wherein The value of | da-la |, which is the difference between the other two pixel lines of the border and la, is compared to the value of | da-la | by multiplying the sum of the two pixel lines by a constant constant k. If <k (da + la) or | da-la | = k (da + la), pixelline is greater than 2 and pixel value is greater than 2 An oversampling method for overcoming alignment errors in a HDDS system, characterized by compensating for an alignment error by extracting a value obtained by adding two pixels while skipping pixels. The method of claim 3, wherein In the case of | ga-b | <k (ga + b) or | ga-b | = k (ga + b), two pixel lines of Ga and My are two other pixel lines of a border perpendicular to the pixel line. Overcoming alignment error in HDDS system, which compensates for misalignment by extracting the value plus 2 pixels as signal value while skipping 2 pixels horizontally and vertically based on the pixel value that meets Dawara Oversampling method. The method of claim 7, wherein If the value of | da-la |, which is the difference between the other two pixel lines of the border and la, is greater than the value of k (da + la) multiplied by a constant constant k, the sum of the two pixel lines is k and b. Two pixel lines are skipped by 2 pixels based on the position of 2 pixels where the pixel line meets the larger pixel line, and the alignment value is compensated by extracting the value of 2 pixels as the signal value. Oversampling Method for Overcoming Alignment Errors in HDDS Systems. The method of claim 7, wherein The value of | da-la |, which is the difference between the other two pixel lines of the border and la, is compared to the value of | da-la | by multiplying the sum of the two pixel lines by a constant constant k. If <k (da + la) or | da-la | = k (da + la), k and b are each pixel line, skipping by 2 pixels based on the position of 4 pixels where they meet An oversampling method for overcoming alignment errors in an HDDS system, characterized by compensating for an alignment error by extracting a pixel plus a signal value. The method of claim 1, In the step (b), when the pixel similarity ratio between the SLM and the CCD is 3: 1, a constant k equal to the sum of the value of | The oversampling method for overcoming misalignment in a HDDS system, characterized by determining the number and position of pixels by comparing the value of k (more + multiplied) multiplied by). The method of claim 10, If the value of | a-da | is greater than a value of k (more +), the pixel line located in the middle of the three pixel lines is the other three pixel lines of the border perpendicular to the pixel line. An oversampling method for overcoming alignment errors in a HDDS system, characterized by compensating for an alignment error by extracting a signal value while skipping by 3 pixels in a horizontal direction and a vertical direction based on a value of a matching pixel. The method of claim 11, The other three pixel lines of the border is the difference between the la, the village, the bar, the la and the bar | la-bar | If the value is greater than the value of k (la + bar) multiplied by a constant constant (k) multiplied by the sum of the two pixel lines, one of the other three pixel lines perpendicular to the pixel line is located in the center of the pixel line. An oversampling method for overcoming alignment errors in a HDDS system, characterized by compensating for an alignment error by extracting a value of one pixel as a signal value while skipping by three pixels based on a position of one pixel that meets a pixel line line located at the center. The method of claim 11, The other three pixel lines of the border is the difference between the la, the village, the bar, the la and the bar | la-bar | The value is compared with the value of k (la + bar) multiplied by a constant constant k, the sum of the two pixel lines, | la-bar | <k (la + bar) or | la-bar | = k (la (+ Bar), the pixel line centered at the three pixel lines is referred to the position of two pixels where the edges of the other three pixel lines perpendicular to the pixel line meet with the pixel line with the large middle line. An oversampling method for overcoming alignment errors in a HDDS system, characterized by compensating for an alignment error by extracting a value obtained by adding two pixels while skipping by three pixels.

Images