KR950003634B1 - Recording/playing method and apparatus of digital image - Google Patents

Abstract

The method includes steps for discrete cosine transforming the picture data composed of lumination signals and color signals into the predetermined pixel block units; seperating the discrete cosine transformed data into the direct current data and the alternating current data; compressing only the separated direct current data step through the quantization; frequency modulating the compressed direct current data; compressing the separated alternating current data through the quantization and variable encoding; frequency modulating the compressed alternating current data; and recording the direct and alternating current data frequency-modulated on the recording medium.

Description

Digital image recording / playback method and apparatus therefor

1 is a block diagram showing a schematic configuration of a video signal processing circuit of an analog VTR.

2 is a block diagram showing a schematic configuration of a typical image data processing circuit of a digital VTR.

3 is a block diagram schematically showing the configuration of a conventional data compression circuit for compressing data by a discrete cosine transformation process.

4 is a diagram showing the structure of an 8x8 block as an example.

5 (a) and 5 (b) show the structures of the video data recording and reproducing apparatus according to the present invention, respectively.

6 is a diagram showing recording frequencies of direct current (DC) components and alternating current (AC) components of an image data signal.

Fig. 7 is a diagram showing band distribution of recording frequencies.

* Explanation of symbols for main parts of the drawings

50: discrete cosine conversion unit 51: DC quantization unit

52,57: error correction coder 53,58: modulator

54: AC quantization unit 55: variable chapter encoding unit

59: multiplexer 60: low pass filter

61: DC detector 62,68: demodulator

63,69: error correction decoding unit 64,71: quantizer encoding unit

65: inverse discrete cosine conversion unit 66: band pass filter

67: AC detection unit 70: variable obstacle reading unit

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data processing method in an image processing apparatus such as a VTR, and more particularly, to a reproduced image due to loss or damage of recorded data in an image processing apparatus which records and reproduces image (or image) data in a digital manner. The present invention relates to a recording / reproducing method of a digital image and a device therefor for preventing the deterioration of the image.

In general, a VTR for recording / reproducing image data may be classified into an analog method and a digital method when divided into data processing methods. In FIG. 1, a simplified configuration of an analog VTR that is commonly used is shown. In FIG. 2, a simplified configuration of the digital VTR is shown.

First, referring to the analog VTR, as shown in FIG. 1, the Y / C separation unit 1 separates the received NTSC color image signal into luminance (Y) and color (C) signals.

Thereafter, the luminance (Y) signal separated by the Y / C separator 1 is input to the frequency multiplexer 7 after being subjected to FM conversion by the frequency modulation (FM) converter 3 and the Y / C. The other signal separated from the separator 1, that is, the color C signal, is input to the frequency multiplexer 7 through the low frequency converter 5.

The frequency multiplexer 7 converts the luminance (Y) signal FM-processed by the frequency modulator (F) converter 3 and the color (C) signal low-frequency transformed by the low-frequency converter 5. It receives the input, processes it with multiple frequencies, and records it on a recording medium, ie tape.

When the aging phenomenon of the recording medium or each component of the analog-type VTR as described above occurs, there is a problem that the sharpness of the playback screen is reduced.

Conventional digital VTR for solving the problems of the analog system is composed of a data compression recording unit and a data expansion and playback unit, as shown in Figure 2 attached, the process of converting NTSC analog video signal to digital data The amount of data for the pixel increases rapidly. Therefore, an effective data compression method is required due to the limitation of the recording medium in the storage of data converted into digital signals. In addition, when data is compressed and the compressed data is stretched and reproduced, a digital data compression algorithm for minimizing the difference between the reproduced image and the original image and a coding for correcting an error are provided. It must be accompanied by necessity.

In general, the digital data compression algorithm is a discrete cosine transformation (DCT) method. An initial screen of an image is transmitted or recorded as an original image, and a subsequent image is different from a previous image. This is a method of transmitting or recording only an image for.

In addition, as an error correction coding method, a two-dimensional Reel Solomon coding method is generally used.

Referring to the operation of the conventional digital VTR applying the above method is as follows.

First, referring to the recording process, the NTSC color image signal is converted into a digital signal by the A / D converter 9, and the data processed by the A / D converter 9 is transferred to the Y / C separator 11. Through this, the luminance (Y) data signal and the color (C) data signal are separated.

Thereafter, the luminance Y data signal separated by the Y / C separation unit 11 is input to the Y encoder 13, and the color (C) data signal is input to the C encoder 15. In this case, the Y encoder 13 and the C encoder 15 divide and output pixel data constituting the entire image into 8 × 8 block units.

The luminance data and the color data divided by the 8 encoder block unit in the Y encoder 13 and the C encoder 15 are data of one channel by a demultiplexer 17. Is incorporated.

The data integrated in the demux 17 is compressed through the DCT method described above by the data compression unit 19 and then recorded on the recording medium through the error correcting encoder 21 and the modulator 23. The image data stored by the recording method as described above is reproduced through the inverse process to the above-described process. Briefly looking at the reproduction process, the data modulated and stored in the recording medium is detected and detected by the demodulator 25. After the error is corrected by the error correction decoding unit 27, the data is restored to the image data before compression by the data restoring unit 29.

Thereafter, the output of the data restorer 29 is divided into luminance data and color data by the multiplexer 31 and then provided to the Y decoder 33 and the C decoder 35, respectively.

In the Y decoder 33 and the C decoder 35, the data divided into 8x8 blocks is restored to the data before separation, and the Y / C synthesizing unit 37 performs the Y decoder 33 and the C decoder 35. Luminance and color data respectively provided from

The data synthesized by the Y / C synthesizing unit 37 is converted into NTSC analog image data by the D / A converting unit 39.

The configuration of the data compression unit (reference numeral 19 in FIG. 2), which is the most essential part in the recording / reproducing operation of the conventional digital VTR as described above, is shown in FIG.

As shown in FIG. 3, the data subjected to discrete cosine transform processing by the DCT unit 41 is compressed while passing through the quantization unit 43 and the variable length coder 45.

The buffer 47 for storing data provided from the variable length coding unit 45 outputs a target information amount unit of input data, that is, a target bit rate. In this case, when the discrete cosine conversion process in the DCT unit 41 is described, data divided into 8x8 block units are sequentially processed to form a data table as shown in FIG. . One 8 × 8 block is composed of one DC component and 63 AC components. In the drawing, the portion represented by 'DC' is the DC component data of the low frequency band, and the portion represented by '1' to '63'. The meaning of numbers as AC component data of this high frequency band is the order input to the said quantization part 43 and the variable length coding part 45. As shown in FIG.

In the conventional digital VTR, the DC and AC components of the DCT processed data are provided to the buffer 47 through a quantization process and a variable length code process through one channel. The buffer 47 is provided to the error correction coder 21 shown in FIG.

In the conventional digital method, the error of each data is mixed because the data of the DC component, which is a major issue in image quality, and the AC component having a high error occurrence rate are mixed while the error frequency is low in the recording frequency provided to the error correcting and encoding unit 21. Since the incidence rate is equally high, frequent error occurs, and thus there is a problem in that the degradation of the reproduced image occurs.

SUMMARY OF THE INVENTION An object of the present invention for solving the above problems is to independently process the DC component and the AC component at the time of recording / reproducing data without significantly changing the configuration of the conventional digital image processing apparatus, thereby lowering the error occurrence rate and further reducing the AC component. The present invention provides a method and apparatus for recording / reproducing a digital image for reconstructing image data using a DC component when an error occurs.

A feature of the present invention for achieving the above object is to digitally convert an analog video signal, separate it into a luminance signal and a color signal, compress it to a recording medium, record it, and reproduce the compressed data from the recording medium to the original video signal. A digital image recording method for reconstructing a digital VTR, comprising: a first step of performing discrete cosine conversion of image data composed of the luminance signal and color signal by a predetermined pixel block; and the data subjected to discrete cosine conversion in the first step. A second process of separating the component and the AC component data, a third process of quantizing and compressing only the DC component data separated in the second process, and a fourth process of frequency modulating the DC component data compressed in the third process Only the AC component data separated in the second step and the quantized and variable length encoding A fifth process, and a sixth process of frequency modulating the AC component data compressed in the fifth fixing; and frequency-multiplexing the data of the DC component and the AC component frequency-modulated through the fourth fixing and sixth processes. It includes the seventh process of recording on the medium.

Another feature of the present invention for achieving the above object is to digitally convert an analog image signal into a luminance signal and a color signal, and then record on a recording medium through a discrete cosine transform into a low frequency band of a DC component and a high frequency band of an AC component. A digital image reproducing method in a base VTR for frequency multiplexing and recording, comprising: a first step of separating data recorded on the recording medium into a low frequency band DC component and a high frequency band AC component; A second process of frequency demodulating the AC component data, a third process of inversely quantizing and variable-field decoding the AC component data demodulated in the second process, and expanding the DC component data separated in the first process. A fourth process of frequency demodulating and dequantizing and expanding the DC component data demodulated in the fourth process The fifth process and the sixth process of generating inverse discrete cosine-converted data of the AC component and DC component extended through the third process and the fifth process to generate image data composed of luminance signals and color signals in predetermined pixel blocks. It is to include.

Another feature of the present invention for achieving the above object is to digitally convert an analog video signal, separate it into a luminance signal and a color signal, compress and record it on a recording medium, and reproduce the data compressed from the recording medium. A digital image recording apparatus in a digital VTR for restoring a signal, comprising: discrete cosine transforming means for performing discrete cosine transforming of image data composed of the luminance signal and color signal by a predetermined pixel block unit; A first quantization means for quantizing only the data of the DC component in the data, the second quantization means for quantizing only the data of the AC component in the discrete cosine transformed data, and the DC quantized in the first quantization means First error correcting means for receiving data of the component and correcting an error, and the first error A first modulating means for receiving frequency-modulated data output from an integer stage, a variable length encoding means for receiving a variable length encoding process by receiving data of an AC component quantized by the second quantizing means, and the variable length encoding means A frequency modulation for receiving the data output from the data outputting a target bit rate, a second error correction means for correcting an error by receiving the data output from the buffer, and a data modulation outputted from the data output from the second error correction means. And second multiplexing means for receiving data output from the first and second modulating means, and multiplexing the frequency and multiplexing the frequency and storing the data in the recording medium.

Another feature of the present invention for achieving the above object is to digitally convert an analog image signal into a luminance signal and a color signal, and then compress and record it on a recording medium through a discrete cosine transform, but the low frequency band of the DC component and the AC component A digital image reproducing apparatus in a digital VTR for frequency multiplexing and recording in a high frequency band, comprising: a low pass filter for receiving data recorded on the recording medium and outputting only a frequency signal of a low frequency band, and recorded on the recording medium. A band pass filter for receiving data and outputting only a high frequency signal within a predetermined frequency band, a first detector detecting DC component data from the output of the low pass filter, and demodulating DC component data output from the first detector. Error correction among data provided from the first demodulator and the first demodulator A first error correcting and decoding unit for decoding an error and correcting an error, a first quantizer for restoring the quantized data provided from the first error correcting and decoding unit to data before quantization, and the band Decodes an error correcting code from a first detector for detecting AC component data from an output of a pass filter, a second demodulator for demodulating AC components output from the second detector, and data provided from the second demodulator. A second error correction decoding unit for correcting an error, a variable length code decoding unit for restoring the variable length coded data provided from the second error correction decoding unit to its original state, and a quantized output from the variable decoding unit Inversely quantizes the signal and receives the output of the second and second quantizers and the first and second quantizers to restore the data before the quantization. Do Lee and may consists of parts of a predetermined pixel block IDCT unit that generates image data composed of a luminance signal and a color signal as been.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings. First, the theoretical background applied to the present invention is to process the DC component data and the data for the AC component separately.

For example, if an original image has 720 x 480 pixel data, 90 x 60 8 x 8 columns in the process from the A / D conversion section 9 to the demux 17 in FIG. DCT conversion of the divided 8 × 8 blocks of pixel data generates a data table such as the 8 × 8 blocks shown in FIG.

In this case, the DC component is one and the AC component constitutes 63 pieces of data. In the case of coding them, each data amount can be separated into low and high frequencies in the frequency band, respectively. This is because in the case of dividing an array of pixels forming an image into blocks of 8x8 matrix units, frequency-dividing 64 pixel signals forming an 8x8 matrix generates a common data value of each image data. The component is data of the DC component, and the signal has low frequency.

In addition, the frequency signal generated when the DC signal is removed from the frequency signal of each of the 63 pixel data except one pixel data closest to the DC component data among the 64 pixel signals forming the 8 × 8 matrix is described. It becomes data of AC component according to each inherent frequency.

5 (a) shows the configuration of the video data recording apparatus according to the present invention, and FIG. 5 (b) shows the configuration of the video signal reproducing circuit.

The image data recording apparatus shown in FIG. 5 (a) has a configuration after the demux 17 of FIG. 2 attached thereto, and inputs only DC component data of the data subjected to the abnormal cosine conversion process by the DCT unit 50. A DC quantization unit 51 for receiving and quantizing, a first error correcting coder 52 for receiving an output signal from the DC quantization unit 51, and adding an error correcting code, and the first error correcting coder 52 A first modulator (53) for modulating an output signal of < RTI ID = 0.0 >), < / RTI > an AC quantizer (54) for inputting and quantizing only AC component data among data subjected to abnormal cosine conversion processing by the DCT unit (50), and A variable length encoder 55 for receiving the output signal of the unit 54 and converting the variable signal into a variable length code, a buffer 56 for outputting data provided by the variable length coder 55 at a target bit rate; A second error correction for receiving an output data of the buffer 56 and adding an error correction code; A frequency is received by receiving the output signals of the encoder 37, the second modulator 58 and the first and second modulators 53, 58 for modulating the output signal of the second error correcting encoder 57. It consists of a multiplexer 59 for multiplexing the data and recording it on the tape.

Referring to a preferred operation example of the image data recording apparatus according to the present invention configured as described above, DC component data of the 8 × 8 block data subjected to the discrete cosine conversion process by the DCT unit 50 is DC quantizer 51 And are uniformly quantized.

Since the maximum number of bits of the quantization size is 8, the DC component data is 8-bit quantized. In addition, AC component data excluding the DC component of the 8 × 8 block data is quantized with fewer bits than 8 bits. In the current quantization technique, the quantization stop size is determined according to the nature of an image. Alternatively, adaptive quantization techniques are used.

The AC component data quantized according to the above scheme is reduced by the variable length encoder 55 to reduce redundancy of the same data, and is output at a predetermined bit rate by the buffer 56.

Subsequently, the quantized DC component data output from the DC quantization unit 51 and the data of the AC component output from the buffer 56 are respectively provided by the first and second error correction decoders 52 and 57, respectively. An error correction coder suitable for the characteristic is attached.

In this case, since the first and second error correction coders 52 and 57 add bit data for error correction, the total data amount is increased by about 30%.

As described above, the DC component data with the error correction code and the data for the AC component are changed by the first and second modulators 53 and 58 and the multi-coder 59 to change 8-10 and non-returning inversion (NRZI). After the modulation is modulated, the frequency distribution is recorded on tape. In this case, a brief configuration of the operation process of the first and second modulators 53 and 58 and the multiplexer 59 is shown in FIG. 6, which is the data after the AC component data is compressed. The batting average was 20 Mbps. In addition, when the DC component data quantizes only the DC component with 8 bits in a frame of one frame having 720 × 480 image data, 8 bits of data are generated per 90 × 60 blocks, so that the total amount of data Is (720 ÷ 8) × (480 ÷ 8) × 8 bits = 43.2 kilobits (Kbit).

Therefore, the data rate per second is 43.2 x 10 ³ (bit) x 30 (frame / bit) x 1.3 = 1.6846 Mpbs.

That is, the data rate of the DC component data after the error correction code is attached is 1.6846 Mbps. In this case, when the data rate (1.6846 Mbps) of the DC component data is modulated by 8-10 conversion, the modulated data is obtained from the following equation (1, 2).

1.6846 x (1 ÷ minimum inversion interval) = maximum modulation data... Formula (1)

1.6846 x (1 ÷ maximum inversion interval) = maximum modulation data... Formula (2)

In general, the minimum inversion interval is 0.8T and the maximum inversion interval is 3.2T in 8-10 conversion (T is the time interval of one data bit).

Therefore, the maximum modulated data is 2.106 Mbps by equation (1) and the minimum modulated data is 0.5265 Mbps by equation (2).

When non-returning inversion (NRZI) conversion of 8-10 converted DC data is performed as described above, since the recording frequency is 1/2 of the 8-10 converted frequency, the final frequency band is 1.053Mbp to 263.25Kbps. . In the same way, when 8-10 conversion and NRZI conversion are performed for an AC component (e.g. 26 Kbps), the recording frequency is 26 Kbps x (1 ÷ 0.8) x 1/2 to 26 Kbps x (1 ÷ 3.2) x 1/2. Has a band.

7 is a diagram showing recording frequency bands of the DC component data and the AC component data obtained by the above-mentioned FIG. 6, and as shown, the recording signals of the DC component and AC component data have different frequency bands. Since it is completely separated, it can be recorded by frequency multiplexing.

In addition, when the filter is used during regeneration, the DC component data and the AC component data can be easily separated. A reproducing apparatus corresponding to the recording apparatus as described above (see also the attached fifth (b)) includes a low pass filter 60 and a corresponding pass filter 66 for filtering the frequency multiplexed signal, respectively, and the low pass filter. A DC detector 61 for detecting the DC component data from the output of the 60, a first demodulator 62 for demodulating the DC component data output from the DC detector 61, and the first demodulator 62 The first error correcting and decoding unit 63, which corrects an error by decoding an error correcting code among the data provided from the data, and restores the quantized data provided from the first error correcting and decoding unit 63 to its original state. A first quantizer 64, an AC detector 67 for detecting AC component data from the output of the bandpass filter 66, and a second demodulated AC component data output from the AC detector 67 Demodulator 68 and data provided from the second demodulator 68; A second error correction decoding unit 69 for decoding the correction code and correcting the error; and a variable length decoding unit for restoring the variable length coded data provided from the second error correction decoding unit 69 to its original state. 70, a second quantizer 71 for restoring the quantized signal output from the variable length code decoder 70, and the first and second quantizers 64 and 71; And an IDCT unit 65 that accepts the output of and performs reverse processing of the discrete cosine transform.

Since the operation of the reproducing apparatus according to the present invention configured as described above is a reverse process of the above-described recording apparatus, a detailed description thereof will be omitted, but the output of the IDCT 65 is a mux 31 (attached) of the conventional digital VTR. (See FIG. 2 in the figure).

In addition, the functions of the low pass filter 60 and the band pass filter 66 are used for the data recorded on the tape by the multiplexing unit 59 in the recording device (see also fifth (a)). And the frequency band is completely separated by the AC component data (high frequency), so that the frequency is separated and accessed.

Provided is a digital image recording / reproducing method and apparatus for operating according to the present invention, and a DC component which is a major issue of image quality and a high error occurrence rate while having a low error rate in a recording frequency in a conventional digital method. Since the error occurrence rate of each data is equally high due to the mixed data, the frequent error occurs and the degradation of the reproduced image is caused.

In addition, since the data of the DC component and the AC component can be restored to separate channels, there is an advantage that a block of the corresponding image can be reproduced only by the data of the DC component when the AC component data is damaged.

Claims (4)

A digital image recording method in a digital VTR which digitally converts an analog image signal, separates the luminance signal and a color signal, compresses the recorded image onto a recording medium, and reproduces the compressed data from the recording medium to restore the original image signal. A first step of discrete cosine transforming the image data composed of the luminance signal color signal by a predetermined pixel block; A second step of separating the discrete cosine-converted data into data of a DC component and an AC component in the first process; A third step of quantizing and compressing only the DC component data separated in the second step; A fourth step of frequency modulating the DC component data compressed in the third step; A fifth step of compressing the AC component data separated in the second step by quantization and variable length encoding processing; A sixth step of frequency modulating the AC component data compressed in the fifth step; And a seventh process of frequency-multiplexing the data of the DC component and the AC component, which are frequency-modulated through the fourth process and the sixth process, to record the data on the recording medium. Digital conversion of analog video signal is separated into luminance signal and color signal, and then compressed and recorded on recording media through discrete cosine transform, but digitally recorded in digital VTR for frequency multiplexing in low frequency band of DC component and high frequency band of AC component. A reproduction method comprising: a first step of separating data recorded on the recording medium into a low frequency band DC component and a high frequency band AC component; A second step of frequency demodulating the AC component data separated in the first step; A third step of decompressing and demodulating the AC component data demodulated in the second step by performing quantization and variable length decoding processing; A fourth process of frequency demodulating the DC component data separated in the first process; A fifth process of inversely quantizing and extending the DC component and data demodulated in the fourth process; And a sixth process of generating inverse discrete cosine transformed AC and DC component data extended through the third and fifth processes to generate image data composed of luminance signals and color signals in a predetermined pixel block unit. Digital image reproduction method. A digital image recording apparatus in a digital VTR which digitally converts an analog image signal, separates the luminance signal and a color signal, compresses the recorded image onto a recording medium, and reproduces the compressed data from the recording medium to restore the original image signal. Discrete cosine transforming means for performing discrete cosine transforming of image data composed of the luminance signal and color signal in a predetermined pixel block unit; First quantization means for quantizing only DC component data among the discrete cosine transformed data; Second quantization means for quantizing only data of an alternating current component among data subjected to discrete cosine transforming by the discrete cosine transforming means; Second error correction means for correcting an error by receiving data of the DC component quantized by the first quantization means; First modulation means for frequency-modulating the data output from the first error correction means; Variable-length encoding means for receiving variable length encoding data quantized by the second quantization means and performing variable length encoding processing; A buffer for receiving data output from the variable length encoding means and outputting a target bit rate; Second error correction means for receiving data output from the buffer and correcting an error; Second modulating means for frequency-modulating the data output from the second error correcting means; And multiplexing processing means for receiving data output from the first and second modulation means and multiplexing the frequency to store the data on the recording medium. Digital conversion of analog video signal is separated into luminance signal and color signal, and then compressed and recorded on recording media through discrete cosine transform, but digitally recorded in digital VTR for frequency multiplexing in low frequency band of DC component and high frequency band of AC component. A reproduction method, comprising: a low pass filter for receiving data recorded on the recording medium and outputting only a low frequency band frequency signal; A band pass filter for receiving data recorded on the recording medium and outputting only a high frequency signal within a predetermined frequency band; A first detector for detecting direct current component data from an output of the low pass filter; A first demodulator for demodulating the DC component data output from the first detector; A first error correction decoding unit for correcting an error by decoding an error correcting code among data provided from the first demodulation unit; A first quantizer for dequantizing the quantized data provided from the first error correcting and decoding unit and restoring the quantized data to previous data; A first detector for detecting AC component data from an output of the band pass filter; A second demodulator for demodulating the AC component data outputted from the second detector; A second error correction decoding unit for correcting an error by decoding an error correcting code among data provided from the second demodulation unit; A variable length code decoding unit for restoring the variable length coded data provided from the second error correction decoding unit to its original state; A second quantizer configured to inversely quantize the quantized signal output from the variable length code decoder to restore data before quantization; And an inverse discrete cosine transform unit which receives the output of the first and second quantizers and performs inverse processing of the discrete cosine transform to generate image data composed of luminance signals and color signals in a predetermined pixel block unit. Digital image reproducing apparatus.