JP3184392B2 - Data recording method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

FIELD OF THE INVENTION The present invention relates to a video system, tele
Data recording used for video broadcasting, videoconferencing, videophone, etc., for recording the compression results of data such as image data
About the method .

[0002]

As a compression method for BACKGROUND ART image data, a method of using a discrete cosine transform (DCT) have been known. In the method using DCT, first, image data is divided into blocks and converted as frequency coefficient values (DCT coefficients) of a two-dimensional matrix. Here, each element of the DCT coefficient obtained as a result of the conversion can be regarded as a spatial frequency component in one certain block of the image data.

For example, one block of image data is N × N
Divided into blocks each consisting of N pixels.
DCT transform is performed on the image data of N pixels, and N × N
When converting to a frequency coefficient value (DCT coefficient) of a two-dimensional matrix, the numerical value of the upper left element of the DCT coefficient of the N × N matrix represents the DC component of the image. The higher frequency component represents the higher frequency component of the shear wave as the numerical value of the element on the right side.

In this way, when an N × N matrix of DCT coefficients is obtained, each element of the N × N matrix of DCT coefficients is represented by N
Quantization is performed by dividing by each element of a quantization matrix consisting of × N thresholds, eliminating data redundancy, and then
Encode and compress by Huffman coding etc. (More specifically, for example, transform the quantized transform coefficients from DC components to high-frequency components into a one-dimensional sequence in a certain order, and convert. (The number of consecutive zeros in the sequence obtained by encoding is compressed) and recorded in a compressed form on a recording device such as an optical disk.

At this time, the above-mentioned quantization matrix is defined by using the quantization parameter Q as follows, for example. u
Is an appropriate constant.

[0006]

## EQU1 ## Quantization matrix (i, j) = (Q / u) × default visibility table (i, j)

From equation (1), each element of the quantization matrix (i, j) is used as a default luminosity factor when a function example presented by the International Standards Group (JPEG) is used as the default luminosity table (i, j). It can be seen that the scaling is adjusted by changing the level of the table (i, j) linearly with the quantization parameter Q. In other words, in Equation 1, the quantization parameter Q functions as a scaling factor (compression rate control factor) for obtaining a scaled (compression rate controlled) quantization matrix (i, j).

Therefore, the data compression processing as described above
Strictly speaking, DCT coefficients obtained as a result of performing discrete cosine transform on image data are quantized by a quantization matrix (i, j), but in effect, data (image data) is quantized. It can be considered that the data is quantized by the quantization parameter Q. In the following, it is assumed that the data is quantized by the quantization parameter Q for convenience of explanation.

In recent years, attention has been paid to the application of this type of image data compression to data compression of pseudo moving images. However, in such a case, the problem is that the amount of data resulting from the compression is reduced. This is a so-called overflow that exceeds a prescribed limit data amount. FIG. 12 is a diagram showing a general relationship between a quantization parameter Q and a data amount D resulting from quantization (compression) by using various image data α, β, γ, and δ. Referring to FIG. 12, the QD curve representing the relationship between the quantization parameter Q and the data amount D is exponential, and the data amount (code length) D increases as the quantization parameter Q increases. It becomes small (compression rate becomes large), and the image quality deteriorates. On the other hand, as the quantization parameter Q decreases, the data amount (code length) D increases.
As shown in FIG. 13, when the data amount D exceeds the limit data amount _Dmax during recording on the recording device, as shown in FIG. 13, the data cannot be accommodated in, for example, one record of the recording device, and an overflow occurs.

When an overflow occurs in the compression result of the data of a predetermined field (for example, even field) of a certain frame, conventionally, the data of the predetermined field (for example, even field) of the frame in which the overflow has occurred is discarded and replaced with this. In addition, measures such as recording data of a predetermined field (even field) of the frame at the immediately preceding time point can be considered.

[0011]

However, in this case, in a frame in which an overflow has occurred, an imbalance or the like occurs between the odd field and the even field, and the reproduced image becomes unnatural. There was a problem that a viewer might feel uncomfortable.

According to the present invention, there is provided a data recording method capable of recording a compression result of image data so as not to give an unnaturalness (unnaturalness) to a reproduced image even when an overflow occurs in the compression result of the image data. It is intended to provide a way .

[0013]

In order to achieve the above-mentioned object, according to the present invention, when one frame is composed of two fields of an odd field and an even field, a frame of a predetermined frame is formed. The compression result of the data of the odd field does not cause overflow,
When an overflow occurs in the compression result of the even field data of the predetermined frame, the compression result of the odd field data of the predetermined frame is substantially copied and recorded in the even field of the predetermined frame. It has become. As a result, the reproduced image of the even-numbered field of this frame becomes the same as the reproduced image of the odd-numbered field of this frame, and time reversal does not occur unlike the conventional method, so that the viewer does not feel unnatural (unnatural). I'm done.

Further, according to the second aspect of the present invention, when one frame is composed of two fields of an odd field and an even field, an overflow occurs in the compression result of the data of the odd field of a predetermined frame. Is essentially copied and recorded in the odd field of the predetermined frame, the compression result of the data of the odd field of the immediately preceding frame in which no overflow has occurred, and the overflow is recorded in the even field of the predetermined frame. The compression result of the data of the even field of the immediately preceding frame that has not occurred is substantially copied and recorded. As a result, with respect to the frame in which the overflow has occurred, the compression result of the data of the odd-numbered field and the even-numbered field in the immediately preceding frame in which the overflow has not occurred is copied and recorded. ).

According to the third aspect of the present invention, when one frame is composed of two fields, an odd field and an even field, when compressing data of a predetermined frame, It is determined whether or not the compression result of the data of the odd field has caused an overflow. As a result, when it is determined that the compression result of the data of the odd field of the predetermined frame has caused the overflow, the odd field of the predetermined frame is determined. The result of the compression of the data in the odd field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded, and the immediately preceding frame in which no overflow has occurred in the even field of the predetermined frame. The result of compressing the data of the even field of When it is determined that the compression result of the data of the odd field of the frame does not cause an overflow, it is further determined whether or not the compression result of the data of the even field of the predetermined frame has caused an overflow. At times, the result of compressing the data of the odd field of the predetermined frame is substantially copied and recorded in the even field of the predetermined frame. This prevents the reproduced image from being unbalanced between the odd field and the even field of this frame, regardless of what overflow occurs in a frame, which makes the viewer unnatural.
It is not necessary to give (uncomfortable feeling).

Further, according to the present invention, when one frame is composed of two fields of an odd field and an even field, when compressing the data of a predetermined frame, It is determined whether or not the compression result of the odd field data has caused an overflow. As a result, if it is determined that the compression result of the odd field data of the predetermined frame has not caused an overflow, the predetermined It is determined whether or not the compression result of the data of the even field of the frame has caused an overflow. When the overflow has occurred, the compression result of the data of the odd field of the predetermined frame is substantially copied to the even field of the predetermined frame. While recording the data of the odd field of the predetermined frame. When it is determined that the compression result has caused an overflow, the compression result of the data of the odd field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded in the odd field of the predetermined frame; It is determined whether or not the result of compressing the data of the even field of the predetermined frame has caused an overflow. When the result of the overflow has occurred, the result of compressing the data of the even field of the immediately preceding frame that has not caused the overflow is substantially determined. It is designed to be copied and recorded. This prevents the reproduced image from being unbalanced between the odd and even fields of this frame, regardless of what overflow occurs in a frame, and makes the viewer unnatural (unnatural). No need to give.

[0017]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration example of a data compression system to which a data recording method according to the present invention is applied. Referring to FIG. 1, the system includes a frame memory 1 for holding image data to be subjected to compression processing,
A compression processing unit 2 for performing a compression process on the image data stored in the memory, a buffer memory 22 for temporarily storing the compressed data, and a recording device 23 for recording the compressed data. I have.

Here, the frame memory 1 stores, for example, the result (D
(CT coefficient) are stored. Further, the compression processing unit 2 quantizes and compresses the DCT coefficients stored in the frame memory 1 by the quantization matrix (i, j) (quantization parameter Q) as shown in Equation 1 as described above. It has become.

By the way, according to the data recording method of the present embodiment, when one frame is composed of two fields of an odd field and an even field, basically,
The compression result D _n (o) of the data of the odd field of the n-th frame F _n has no overflow, and the compression result D _n (e) of the data of the even field of the n-th frame F _n has the overflow. In this case, the n-th frame F is stored in the buffer memory 22 and the recording device 23.
substantially copying the n-th frame F _n compression result of the data of the odd field D _n (o) to the even field _n adapted to (copy) and recorded.

When the compression result D _n (o) of the data of the odd field of the n-th frame F _n overflows, the image data of the n-th frame F _n is replaced by the immediately preceding (n−1) -th image data. For image data of frame F _n−1 ,
It is often due to a major change. Therefore, at this time, regardless of whether the compression result D _n (e) of the data of the even field of the n-th frame F _n overflows, the compression result D _n (e) of the data of the even field.
Is regarded as overflowing, and buffer memory 2
2, the recording device 23, the odd field of the n-th frame F _n, 1 previous where there is no overflow ((n
-1) The compression result D _n-1 (o) of the data of the odd field of the frame F _n-1 is substantially copied (copied) and recorded, and the compressed result D _n-1 (o) is recorded in the even field of the n-th frame F _n. ,
The compression result D of the data of the even field of the immediately preceding ((n-1) th) frame _Fn-1 in which no overflow has occurred.
_n-1 (e) is substantially copied.

More specifically, in the present embodiment, when performing compression processing on the image data of the n-th frame, n
Th compression result data of an odd field of a frame is provided with a judging means for judging whether an overflow occur, the result of determination by該判by means, n-th frame F _n compression result D of data of the odd field _{When n} (o) overflows, the n-th frame F _n
No overflow occurs in the odd field of 1
The compression result D _n-1 (o) of the data of the odd field of the immediately preceding ((n-1) th frame F _n-1 is substantially copied and recorded, and the even number of the n th frame F _n is recorded. In the field,
The compression result D of the data of the even field of the immediately preceding ((n-1) th) frame _Fn-1 in which no overflow has occurred.
_{When n-1} (e) is substantially copied and recorded, while it is determined that the compression result D _n (o) of the data of the odd field of the n-th frame F _n does not cause overflow,
Further, when the compression result of the n-th even field of the frame F _n data D _n (e) it is determined whether or not an overflow occur, resulting in overflow, n-th even field n-th frame F _n The result D _n (o) of the compression of the data in the odd field of the frame F _n is substantially copied and recorded.

It should be noted that as a result of the above determination as to whether or not the compression result D _n (o) of the data of the odd field of the n-th frame F _n has caused an overflow, the n-th frame F n
When the compression result of the data of the odd field of the _n D _n (o) occurs an overflow, with respect to the odd field of the n-th frame F _n, before one where there is no overflow ((n-1) th) frame F _n-1 data compression result of the odd field D _n-1 of the (o) recorded in substantially copying, also with respect to the even field of the n-th frame F _n, the n-th frame F _n the odd field independently to determine an overflow has occurred, when the result of this determination, the compression result of the n-th data of the even field of the frame F _n D _n (e) does not cause overflow, the The compression result D _n (e) is recorded as it is, and when the compression result D _n (e) has overflowed, the immediately preceding ((n-n)
The data compression result D _n-1 (e) of the even field of the (1) -th frame F _n-1 may be substantially copied and recorded.

FIG. 2 shows such a data recording method of this embodiment.
FIG. 3 is a diagram illustrating an example of a hardware configuration of a compression processing unit 2 for implementing the method . In the configuration example of FIG. 2, the compression processing unit 2
A first compression processing unit 2-1 for performing compression processing on image data of an odd field in correspondence with the fact that one frame is composed of two fields of an odd field and an even field, and compression for image data of an even field. A second compression processing section 2-2 for performing processing, a control section 3 for performing overall processing control, and a first compression processing section 2-
1, a selection output unit (for example, a switch) 4 for selectively outputting the processing result of the second compression processing unit 2-2.

The first compression processing section 2-1 includes a controller 5 for controlling the compression of the image data of the odd field.
And a compressor 6 for compressing image data of odd fields
And write the compression result of the image data in the odd field,
A switching selector (for example, a switch) for switching between two readable memories 7a and 7b and a memory for writing a compression result from the compressor 6 among the two memories 7a and 7b.
9.

Similarly, the second compression processing section 2-2 includes a controller 5 'for controlling the compression of the image data of the even field, a compressor 6' for compressing the image data of the even field, and a Selector for switching between the two memories 8a and 8b capable of writing and reading the compression result of the image data and the memory for writing the compression result from the compressor 6 'among the two memories 8a and 8b. (For example, a switch) 9 ′.

Here, as described above, the compressors 6 and 6 'compress the image data using the quantization matrix (i, j) (quantization parameter Q) as shown in Equation 1, for example. Has become.

The controller 5 has a function of controlling the compression timing and the like for the compressor 6 and also controls the switching selector 9 to select a memory in which the compression result from the compressor 6 is to be written. And a function of determining whether or not the compression result from the compressor 6 has overflowed. Similarly, the controller 5 ′ switches the switch selector 9 in order to select the memory in which the compression result from the compressor 6 ′ is to be written, together with the function of controlling the compression timing and the like for the compressor 6 ′. It has a function of controlling, and further has a function of determining whether or not the compression result from the compressor 6 'has overflowed.

FIG. 3 is a diagram showing a configuration example of the controller 5 (or the controller 5 '). In the example of FIG. 3, the controller 5, the controller 5 'is the image data compressor 6,6' data amount detection unit for detecting an amount of compressed data D _n of the result of the compressed (eg counter) 11, performs determination of whether or not an overflow has occurred detected compressed data D _n is the whether exceeds the limit data amount D _max, the determination control unit for controlling the switch selector 9, 9 'according to the result of determination 13 are provided.

The controller 5 'is dependent on the controller 5, and when the controller 5 determines that the data in the odd field has overflowed, the control mode is uniquely determined. It has become. For example, when the controller 5 does not determine the overflow for the data of the odd field of a certain frame, the controller 5 ′ controls the overflow of the data of the even field of this frame independently of the controller 5. However, when the controller 5 determines that the data of the odd field of a certain frame has overflowed, the controller 5 ′ determines whether or not the data of the even field of this frame does not overflow (ie, overflows). Irrespective of whether or not it has been performed), it is determined that an overflow has occurred, and the switching selector 9 'is controlled.

The control unit 3 includes the memories 7a, 7b, 8
a, 8b is selected and the compression result is read from the memory, so that the control state of the controllers 5, 5 'is monitored,
Based on this, it has a function of performing selection control of the selection output unit 4. The control unit 3 further includes a memory 7
It is assumed that a function for controlling the timing of writing and reading of a, 7b, 8a and 8b is provided.

Such a system configuration can be realized by a compact circuit. For example, each of the first compression processing section 2-1 and the second compression processing section 2-2 can be easily realized by a microcomputer. .

The data compression method according to the present embodiment is specifically realized as follows under the system configuration examples shown in FIGS.

When no overflow occurs in the data compression result of each frame, the controller 5 of the first compression processing unit 2-1 switches the switching selector 9 alternately for each frame, so that the memories 7a and 7b have The compression result of the data in the odd field of each frame is written alternately for each frame.

The controller 5 of the second compression processing section 2-1
Similarly, the switching selector 9 'is alternately switched for each frame. Therefore, the compression results of the data of the even-numbered fields of each frame are alternately written to the memories 8a and 8b for each frame.

In this case, the control section 3 sequentially switches the selection output section 4 cyclically, thereby utilizing the period during which writing to each of the memories 7a, 7b, 8a, 8b is not performed. Then, the compression result is sequentially read out and stored in the buffer memory 22.

FIG. 4 shows the memories 7a, 7b, 7b, 7c and 7d when no overflow occurs in the compression result of the data of each frame.
It is a figure which shows an example of the control form of writing and reading of 8a, 8b. Referring to FIG. 4, the compression result D ₁ (o) of the data of the odd field of the _first frame F ₁ and the compression result D of the even field of the _first frame F ₁ are shown.
₁ (e), the compression result of the second data of the odd field of frame F ₂ D ₂ (o), the compression result of the second data of the even field of the frame _{F 2 D 2 (e),} 3 -th frame F _Three
, The compression result D ₃ (o) of the data of the odd field of the third frame F ₃ , and the compression result D ₃ of the data of the even field of the _third frame F ₃
(e), ... are the memories 7a, 8a, 7b, 8 respectively.
, b, 7a, 8a,... in this order, and from the point in time when the writing of the compression result of the data of the second frame starts, the memories 7a, 8a, 7b, 8b, 7a, 8a,. The written compression results are D ₁ (o), D ₁ (e), D
₂ (o), D ₂ (e), D ₃ (o), D ₃ (e),... In this way, when no overflow has occurred, the result of the compression process is used as it is.
It can be stored in the buffer memory 22.

On the other hand, when data of a certain frame is compressed, no overflow occurs in the compression result of the data of the odd field, but overflow occurs in the compression result of the data of the even field of this frame. And the switching control mode of the switching selectors 9 and 9 'of the controllers 5 and 5', the switching selectors 9 and 9 'are alternately switched for each frame as in the case where no overflow occurs in the compression result of the data of each frame. Therefore, the compression results of the odd field data of each frame are alternately written to the memories 7a and 7b for each frame, and the compression results of the even field data of each frame are written to the memories 8a and 8b. Are written alternately for each frame, but the control unit 3 generates an overflow in the compression result of the data of the even field. As for the frame, instead of reading the memory holding the compression result of the odd field data of this frame, reading the compression result of the odd field data, reading the memory holding the even field data in which overflow has occurred. Then, the compression result of the odd field data is read out again from the memory holding the compression result of the odd field data, and stored in the buffer memory 22 as the compression result of the even field data.

FIG. 5 shows that no overflow occurs in the compression result of the data of the odd field of a certain frame.
The memories 7a, 7b, 8a, and 8a, when an overflow occurs in the compression result of the data of the even field of this frame.
8B is a diagram illustrating an example of a write / read control mode of FIG. Referring to FIG. 5, even if overflow occurs in the compression result of the data of the even field of the _second frame F2, each compression result D ₁ (o), D ₁ (e), D ₂ (o), D ₂ (e),
D ₃ (o), D ₃ (e),... Are stored in memories 7a, 8a, 7b, 8b, 7a, 8a,.
Are written in this order. Also, the memories 7a, 7b,
8a and 8b are also read from the memory 7 from the point in time when writing of the compression result of the F2 data of the _second frame is started.
The compression is performed in the order of a, 8a, and 7b. However, since the compression result written in the memory 8b causes an overflow, the memory 8b is not read out, and instead, the memory 7b is read out again. Thus, the second frame F
When an overflow occurs in the compression result of the data of the even field of ₂ , the memories 7a, 8a, 7b, 7
b, 7a, 8a,..., D ₁ (o), D ₁ (e), D ₂ (o),
_{D 2 (o), D 3} (o), D 3 (e), ... compression result is read as, for the second even field of the frame F _2, the data of the odd field of the frame F ₂ The compression result D ₂ (o) can be substantially copied (copied) and stored in the buffer memory 22.

[0039] In the prior art, when an overflow occurs in the second frame F compression result data of the even field _2, as shown in FIG. 6, for the second even field of the frame F _2, 1 th Frame F ₁
Compression result D ₁ of the data of the even field (e) is copied (copy), and was stored in the buffer memory 22. However, in this case, during playback, this frame F ₂
Time reproduction occurs between the reproduced image of the odd field and the reproduced image of the even field, and after the reproduced image of the odd field is output, the image at the immediately preceding point is reproduced and output again. It gives unnaturalness (uncomfortable feeling).

On the other hand, in the present embodiment, when an overflow occurs in the compression result of the data of the even field of the _second frame F ₂ , the even field of the second frame F ₂ is replaced by the same frame F ₂ . copying the compressed results of the data of the odd field D ₂ (o) and (copy), since stored in the buffer memory 22, the reproduced image of the even field of the frame F ₂ is a reproduction image of an odd field of the frame F ₂ This is the same, and time reversal does not occur unlike the conventional method, so that it is not necessary to give the viewer unnaturalness (uncomfortable feeling).

If an overflow occurs in the compression result of the data of the odd field of a certain frame, as one aspect of the present embodiment, the controller 5 switches the switching selector 9 when shifting to the next frame. Absent. Thereby, the compression result of the odd field data of the next frame is written in the same memory as the memory in which the compression result of the odd field data of the frame before the overflow occurs. Further, the controller 5 gives information to the effect that overflow has occurred in the compression result of the data of the odd field of a certain frame to the controller 5 ′. In this case, the controller 5 'also selects the switching selector 9' when shifting to the next frame.
Do not switch. As a result, the compression result of the even field data of the next frame is written in the same memory as the memory in which the compression result of the even field data of the previous frame is written. As a result, the contents of the memory in which the compression results of the odd field data and the even field data of the frame immediately before the frame in which the overflow of the data of the odd field has occurred are written in the frame in which the overflow has occurred. Until the data write cycle of the next frame is completed, the data is stored as it is.

For a frame in which an overflow has occurred in the compression result of the data of the odd field, the control unit 3 reads from the memory in which the compression result of the data of the odd field and the even field of this frame is written. Instead of this, the compression result of the data of the odd field and the even field of the immediately preceding frame is stored in the memory holding the compression result of the data of the odd field and the even field of the immediately preceding frame. Is read out again and stored in the buffer memory 22 as a result of compressing the data of the odd field and the even field of the frame in which the overflow has occurred.

FIG. 7 is a diagram showing an example of a control mode of writing and reading of the memories 7a, 7b, 8a and 8b when an overflow occurs in the compression result of the data of the odd field of a certain frame. Referring to FIG. 7, the compression result D ₁ (o) of the data of the odd field of the _first frame F ₁ ,
The compression result D ₁ (e) of the even field data of the _first frame F ₁ is written to the memories 7a and 8a in this order, and then the compression of the odd field data of the second frame F ₂ is performed. The result D ₂ (o) and the compression result D ₂ (e) of the data of the even field of the _second frame F ₂ are written to the memories 7b and 8b in this order. By the way, when the compression result D ₂ (o) of the data of the odd field of the _second frame F ₂ overflows,
The third data of the odd field of frame F ₃ compression result D ₃ (o), the third frame F compression result data of the even field of the ₃ D ₃ (e), the memory 7a, 8
rather than a, 2-th frame F ₂ of the odd field, data of the even field is written memory 7b, written to 8b.

The reading of the memories 7a, 7b, 8a and 8b is performed in the order of the memories 7a and 8a from the time when the writing of the compression result of the data of the second frame is started. Has overflowed in the compression result written in the memory 8b.
Is regarded as having caused an overflow in the compression result written therein, so that the memories 7a and 8
After the data D ₁ (o) and D ₁ (e) of the odd field and the even field of the first frame are sequentially read from a, the contents of the memories 7 b and 8 b are not read, but instead the memories 7 a and 8 a Is read again. When the memories 7a and 8a are read again, the memories 7a and 8a still hold the compression results D ₁ (o) and D ₁ (e) of the data of the odd field and the even field of the _first frame F1. Accordingly, the compression results D ₁ (o) and D ₁ (e) of the data of the odd field and the even field of the _first frame F ₁ are read out again. Thus, when an overflow occurs in the compression result of the data of the odd field of the _second frame F2, the memories 7a, 8a, 7a, 8a, 7
b, 8b,..., D ₁ (o), D ₁ (e), D ₁ (o), D
_{1 (e), D 3 (} o), D 3 (e), ... compression result as is read, the second odd field of frame F _2, with respect to the even field, the first odd frames F ₁ The compression result of the data of the field and the even field can be substantially copied (copied) and stored in the buffer memory 22.

As described above, in one aspect of the present embodiment, 2
When an overflow occurs in the compression result of the data of the odd field of the _second frame F2, it is considered that an overflow has occurred even if the even field has not caused an overflow (that is, it is considered that an overflow has occurred in the entire second frame). ), with respect to the second frame F _2, odd field in the first frame F _1, the compression result D data of the even field
₁ (o) and D ₁ (e) are copied.
Since it is stored in 2, the viewer does not have to feel unnatural (unnatural).

As another mode for the case where an overflow has occurred in the compression result of the data of the odd field of the _second frame F _2, a countermeasure against the overflow is performed only for the odd field in which the overflow has occurred, It is also possible to perform independent processing on the compression result of the field data without interlocking with the odd field. That is, when an overflow occurs in the compression result of the data of the odd field of the _second frame F2, and when no overflow occurs in the compression result of the data of the even field of the second frame, as shown in FIG. for the odd field of frame F _2, 1-th compressed data resulting odd field of frames F ₁ copies the D ₁ (o), with respect to the second even-numbered field of the frame F _2, the second frame F data ₂ of the even field compression result D ₂
(e) can be used as it is. In this case, 2
In the reproduced image of the _second frame F2, the scene is switched between the odd field and the even field, and the viewer does not feel unnatural (unnatural). That is,
When an overflow occurs in an odd field, the scene often changes greatly here, and the switching of the scene in the reproduced image is considered to reflect this faithfully and is very natural.
In the other aspect, when an overflow also occurs in the compression result of the data of the even field of the second frame, the even field of the first frame F ₁ is related to the even field of the _second frame F _2. In this case, the data compression result D ₁ (e) of the second frame F ₂ is copied as in the first embodiment.
The results in the first frames F ₁ of the compressed result data of the previous is recorded.

As described above, the present invention relates to control when overflow occurs in the compression result.
It is not directly intended to avoid overflows. However, in applying the present invention, it is desirable to prevent the occurrence of overflow as much as possible in advance. A mechanism for preventing the occurrence of overflow can be provided in, for example, the controllers 5 and 5 'in the configuration examples of FIGS.

More specifically, the controllers 5, 5 '
The configuration shown in FIG. In the example of FIG. 9, the controller 5, 5 ′ uses the compression ratio for the image data of the (n + 1) th frame, that is, the quantization parameter Q _{n + 1} for the image data of the nth frame. Prediction using the quantization parameter Q _n , the data amount D _n detected by the data amount detection unit 11 and the target data amount,
A compression ratio control unit 12 to be provided to the compressors 6, 6 'is further provided. In such a configuration, the maximum data amount D _max
A data amount somewhat smaller than the target data amount (optimal data amount) _Dobj is set, and the quantization parameter Q is determined by feedback control so that the compression result becomes a data amount as close as possible to the target data amount _Dobj . Overflow can be prevented as much as possible.

Further, the controllers 5 and 5 ′ are, for example, a patent application filed on Dec. 21, 1993 by the present applicant.
As described in (Reference number 4149), a configuration as shown in FIG. 10 may be employed.

That is, in the configuration shown in FIG. 10, the data at one time point (n + 1) is compressed at a predetermined compression rate Q ₀ in the first compression processing section C ₁ , and the first compression processing section C ₁ responds to the compression result at that time. An optimal first compression ratio Q _{n + 1} ″ is determined, and the second compression processing unit C ₂ performs compression processing on the data at the previous time point (n) to obtain a compression result. An optimal second compression ratio Q _{n + 1} ′ is determined based on the _first compression ratio Q _{n + 1} ′ and the first compression ratio Q _{n + 1} ″. Optimal compression ratio Q for data at (n + 1)
_The compression ratio determination unit 16 determines _{n + 1} .
As a result, even when the time variation of the data is large, the data amount is quickly converged to an optimal data amount as large as possible without causing overflow, thereby suppressing the deterioration of the data quality and performing the compression processing in real time. be able to.

Alternatively, as shown in a patent application filed on the same date as the present application (reference number 4136) by the applicant of the present invention, the constitution shown in FIG. 11 may be adopted.

In the configuration shown in FIG. 11, the compression ratios predicted by the compression ratio prediction unit 35 are stored in the storage unit 36 in the range of several points in time from the current time to the past, and are stored in the storage unit 36. Among the compression ratios, the one with the maximum compression ratio is determined by the compression ratio determination unit 16 as the compression ratio for the current data and is given to the compressor, and the compressor compresses the current data at the determined compression ratio. It has become. As a result, such as scenes burning flash and strobe,
Even when the image greatly changes other than when the scene is switched, it is possible to reduce the occurrence of overflow and obtain a reproduced image that does not cause a sense of incongruity during reproduction.

Further, as described in a patent application filed on December 21, 1993 by the applicant of the present invention (reference number 4138), the overflow described above which occurs when the quantization parameter Q is in a small range. In order to prevent this, the range in which the value of the quantization parameter Q does not become too small, that is, the fluctuation range of the data amount when compression is performed by the predicted quantization parameter (the target data amount D
_obj, the limit value (minimum value) Q _min of the quantization parameter Q is set in a range that can be absorbed by the difference (margin) between the limit (maximum) data amount D _max and the target data amount D _obj , When the quantization parameter Q _{n + 1} becomes smaller than Q _min , a configuration may be adopted in which the quantization parameter Q _{n + 1} is determined again so as to have a constant value Q _min .

In the above embodiment, the image data is
Although the data compression (quantization) is performed after performing the CT transformation, the present invention can be similarly applied to a case where an arbitrary transformation other than the DCT transformation is used.

In the above-described embodiment, the compressors 5,
5 'is such that the compression processing by the quantization parameter is actually performed by the quantization matrix (i, j) as shown in Expression 1. At this time, when a quantization parameter is given, it takes a very long time to calculate the quantization matrix (i, j) by Equation 1 each time, which causes a problem in real-time processing. For the quantization parameter Q, a quantization matrix (i, j) is determined in advance, and these are stored in, for example, a ROM so as to correspond to the value of the quantization parameter Q. When the value of the quantization parameter Q is given, Is preferably read out from the ROM. In the above embodiment, the JPE
A quantization matrix (quantization parameter) defined by G
The compression is performed according to the following equation. However, as long as the data is compressed at a predetermined compression rate, not limited to the quantization matrix (quantization parameter) specified by JPEG, an arbitrary one can be used.

In the above embodiment, the controllers 5, 5 '
Is designed to determine whether or not an overflow has occurred based on the amount of compressed data detected by the data amount detection unit 11. Instead of this, for example, an overflow may occur from the state of writing to the memory. It is also possible to determine whether or not it has occurred. In this case, the data amount detection unit 11 does not necessarily have to be provided.

[0057]

As described above, according to the first aspect of the present invention, when one frame is composed of two fields of an odd field and an even field, the data of the odd field of the predetermined frame is obtained. If the compression result does not cause an overflow and the compression result of the even field data of the predetermined frame causes an overflow, the compression of the odd field data of the predetermined frame is performed on the even field of the predetermined frame. Since the result is substantially copied and recorded, the reproduced image of the even field of this frame is the same as the reproduced image of the odd field of this frame, and time inversion does not occur unlike the conventional method. Unnaturalness to the viewer (discomfort)
Need not be given.

According to the second aspect of the present invention, when one frame is composed of two fields, an odd field and an even field, an overflow occurs in the compression result of the data of the odd field of the predetermined frame. In this case, in the odd field of the predetermined frame, the result of compressing the data of the odd field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded,
In addition, since the compression result of the data of the even field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded in the even field of the predetermined frame. Since the compression result of the data of the odd-numbered field and the even-numbered field in the immediately preceding frame in which no overflow occurs is copied and recorded, the viewer does not have to feel unnatural (unnatural).

According to the third aspect of the present invention, when one frame is composed of two fields of an odd field and an even field, when performing compression processing on data of a predetermined frame, the predetermined frame is used. It is determined whether or not the compression result of the data of the odd field of the frame has caused an overflow. As a result, when it is determined that the compression result of the data of the odd field of the predetermined frame has caused the overflow, the overflow of the predetermined frame is determined. In the odd field, the result of compressing the data of the odd field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded, and in the even field of the predetermined frame, the immediately preceding field in which the overflow has not occurred is recorded. While the result of compressing the data of the even field of the frame is substantially copied and recorded, When it is determined that the compression result of the odd field data of the predetermined frame does not cause an overflow, it is further determined whether or not the compression result of the even field data of the predetermined frame has an overflow. When this occurs, the result of compressing the data of the odd field of the predetermined frame is substantially copied and recorded in the even field of the predetermined frame. Also, it is possible to prevent the reproduced image from being unbalanced between the odd field and the even field of this frame, and to give the viewer no unnaturalness (unnaturalness).

According to the fourth aspect of the present invention, when one frame is composed of two fields of an odd field and an even field, when performing compression processing on data of a predetermined frame, the predetermined It is determined whether or not the compression result of the data of the odd field of the frame has caused an overflow. As a result, when it is determined that the compression result of the data of the odd field of the predetermined frame has not caused the overflow, It is determined whether or not the compression result of the data of the even field of the predetermined frame has caused an overflow. When the overflow has occurred, the compression result of the data of the odd field of the predetermined frame is substantially added to the even field of the predetermined frame. While copying and recording the data, the odd field data of the predetermined frame is recorded. When it is determined that the compression result of the data has caused an overflow, the compression result of the data of the odd field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded in the odd field of the predetermined frame, Further, it is determined whether or not the compression result of the data of the even field of the predetermined frame has overflowed. When the overflow has occurred, the compression result of the data of the even field of the immediately preceding frame where no overflow has occurred is determined. Since the data is virtually copied and recorded, it is possible to prevent the reproduced image from being unbalanced between the odd field and the even field of this frame even if any overflow occurs in this frame. , Without giving the viewer unnaturalness (discomfort).

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration example of a data compression system to which a data recording method according to the present invention is applied.

FIG. 2 is a diagram illustrating a hardware configuration example of a compression processing unit for realizing the data recording method of the present invention.

FIG. 3 is a diagram illustrating a configuration example of a controller of a compression processing unit illustrated in FIG. 2;

FIG. 4 is a diagram illustrating an example of a write / read control mode of the memory according to the present embodiment when no overflow occurs in the compression result of the data of each frame.

FIG. 5 is a diagram illustrating an example of a write / read control mode of a memory according to the present embodiment when an overflow occurs in a compression result of data of an even field of a certain frame.

FIG. 6 is a diagram for explaining a conventional recording method when an overflow occurs in a compression result of data of an even field of a certain frame.

FIG. 7 is a diagram illustrating an example of a write / read control mode of the memory according to the present embodiment when an overflow occurs in a compression result of data of an odd field of a certain frame.

FIG. 8 is a diagram for explaining another example of a recording method when an overflow occurs in a compression result of data of an odd field of a certain frame.

FIG. 9 is a diagram showing another configuration example of the controller.

FIG. 10 is a diagram showing another configuration example of the controller.

FIG. 11 is a diagram showing another configuration example of the controller.

FIG. 12 is a diagram showing QD curves for various images.

FIG. 13 is a diagram illustrating an example of a recording type of a recording apparatus.

[Explanation of symbols]

 Reference Signs List 1 frame memory 2 compression processing unit 2-1 first compression processing unit 2-2 second compression processing unit 3 control unit 4 selection output unit 5, 5 'controller 6, 6' compressor 7a, 7b, 8a, 8b memory 9, 9 'switch selector 11 data amount detection unit 13 discrimination control unit 22 buffer memory 23 recording device

Claims (4)

(57) [Claims] 1. When one frame is composed of two fields, an odd field and an even field, no overflow occurs in the compression result of the data of the odd field of the predetermined frame, and the even field of the predetermined frame does not overflow. If the result of the data compression occurs an overflow, the data recording direction, characterized in that recording to substantially duplicate the results of compression data of the odd field of the predetermined frame to the even field of the predetermined frame
Law . 2. When one frame is composed of two fields, an odd field and an even field, when an overflow occurs in a compression result of data of an odd field of a predetermined frame, an odd number of the predetermined frame is used. In the field, the result of compressing the data of the odd field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded, and in the even field of the predetermined frame, the immediately preceding field in which no overflow has occurred is recorded. A data recording method characterized by substantially copying and recording a compression result of data of an even field of a frame. 3. When one frame is composed of two fields, an odd field and an even field, when performing compression processing on data of a predetermined frame,
It is determined whether or not the compression result of the odd field data of the predetermined frame has caused an overflow, and as a result,
When it is determined that the compression result of the data of the odd field of the predetermined frame has caused the overflow, the compression result of the data of the odd field of the immediately preceding frame in which the overflow has not occurred is stored in the odd field of the predetermined frame. While substantially copying and recording, and in the even field of the predetermined frame, the result of compressing the data of the even field of the immediately preceding frame in which no overflow has occurred is substantially copied and recorded, and When it is determined that the compression result of the data of the odd field of the frame does not cause an overflow, it is further determined whether or not the compression result of the data of the even field of the predetermined frame has caused an overflow. Sometimes, the predetermined frame is stored in an even field of the predetermined frame. Data recording method characterized by the compression result of the data of the odd field recording and substantially copying. 4. When one frame is composed of two fields, an odd field and an even field, when performing compression processing on data of a predetermined frame,
It is determined whether or not the compression result of the odd field data of the predetermined frame has caused an overflow, and as a result,
If it is determined that the compression result of the odd field data of the predetermined frame has not caused an overflow, it is further determined whether or not the compression result of the even field data of the predetermined frame has caused an overflow. Occurs, the compression result of the data of the odd field of the predetermined frame is substantially copied and recorded in the even field of the predetermined frame,
When it is determined that the compression result of the data of the odd field of the predetermined frame has caused the overflow, the compression result of the data of the odd field of the immediately preceding frame in which the overflow has not occurred is stored in the odd field of the predetermined frame. It is substantially copied and recorded, and it is determined whether or not the compression result of the data of the even-numbered field of the predetermined frame has caused an overflow. When the overflow has occurred, the immediately preceding frame having no overflow has been determined. A data recording method characterized by substantially copying and recording a compression result of data of an even field.