JP3163700B2 - Image recording device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image recording apparatus, and more particularly, to an image recording apparatus for compressing a moving image by variable length coding and recording the compressed moving image on a recording medium.

[0002]

2. Description of the Related Art A digital video tape recorder (VTR) is known as an image recording apparatus for digitally compressing a moving image and recording it on a recording medium such as a magnetic tape. The compression method includes a method using a fixed-length coding method and a method using a variable-length coding method.

FIG. 3 is a block diagram showing the configuration of a conventional example of an image recording / reproducing apparatus using a fixed-length encoding system. Input terminal 10
Is input with an image signal of a moving image to be recorded.
The A / D converter 12 digitizes the analog image signal from the input terminal 10, and the fixed-length encoding circuit 14
The output data of the converter 12 is encoded with a fixed length. The modulation circuit 16 performs low-frequency suppression modulation on the output of the fixed-length encoding circuit 14, and the output is amplified to a predetermined level by the recording amplifier 18. The switch 20 is connected to the contact a during recording and to the contact b during reproduction. The output of the recording amplifier 18 is applied to the magnetic head 22 via the switch 20 and
4 recorded.

At the time of reproduction, the recording signal on the magnetic tape 24 is reproduced by the magnetic head 22 and its output is
0 and is applied to a demodulation circuit 28 via a reproduction amplifier 26 and demodulated. The fixed-length decoding circuit 30 is a decoding circuit corresponding to the fixed-length encoding circuit 14, and includes a demodulation circuit 2
8 is decoded to output a digital reproduction image signal. The output of the decoding circuit 30 is converted to an analog signal by a D / A converter 32, and an analog reproduced image signal is output from an output terminal 34 to the outside.

[0005] The fixed-length encoding circuit 14 generally employs an encoding system in which the amount of encoded data is constant within one screen (one field or one frame). Pulse code modulation (PC
In addition to the M) method, there are a so-called sub-sampling method for digital compression, a differential coding (DPCM) method, and the like. The former does not compress the image information at all, so there is no image quality degradation, but the amount of recording data is enormous and the recording rate has to be increased, and therefore the recording density and recording time of the hardware and the recording medium There are many problems in such points.

When digital compression is performed by fixed-length encoding, a compression ratio of about 1/4 to 1/6 can be achieved. However, in the case of a high-definition television signal such as HDTV, the amount of recorded information is still large. Pass. Further, deterioration of the image quality is noticeable.

On the other hand, for example, in a variable length coding system using a Huffman code or an arithmetic code, a high compression ratio of about 1/10 to 1/20 can be achieved without deteriorating image quality much. An example of such a high-efficiency variable-length coding method is an adaptive discrete cosine transform (ADCT) method (for example, Takahiro Saito et al., “Coding Method for Still Images”, Journal of the Institute of Television Engineers of Japan, Vol. 44, No. 2 (1990)). Tomohiro Ochi et al., “International Standard Trends in Still Image Coding,” Proceedings of the 14th Annual Meeting of the Institute of Image Electronics Engineers of 1988.

[0008]

However, in the variable length coding system, the amount of recording information per screen is basically not constant. Therefore, there is a problem that it is difficult to perform the editing, for example, the joint photographing (or the joint recording), and the reproduced image is largely disturbed before and after the joint photographing. Such a problem,
This is particularly noticeable in a digital VTR using a magnetic tape as a sequential access medium as a recording medium.

An object of the present invention is to provide an image recording apparatus which solves such a problem.

[0010]

According to the present invention, there is provided an image recording apparatus comprising: an input unit for inputting image data;
Intra-screen coding that performs variable-length coding using
Select between-screen coding and variable-length coding using correlation
Image data input by the input means
A plurality of first encodings encoded and intra-coded
Screen and a plurality of second encoded screens inter-coded
Encoding means for outputting encoded image data consisting of:
Forming a plurality of tracks on a tape like recording medium, the marks
And recording means for recording the encoded image data output from Goka means, of the encoded image data by said recording means
Instruction means for instructing the start of recording;
The encoding unit and the recording unit based on a recording start instruction
And control means for controlling the door, said control means, said
The first recording screen according to the recording start instruction is the first code
And the encoded image data of the first recording screen
Recording from the beginning of the track and the first recording
The first output from the encoding means after the recording surface
Converting the encoded image data of the encoded screen into the encoded image data
According to the data amount of the first and second encoded screens
The encoding to be recorded at an arbitrary position in the track
Means for controlling the means and the recording means .

[0011]

[Operation] The screen is decoded by the variable length coding method.
The amount of recording data differs for each
Although not specified, according to the present invention, for example,
Also in this case, the recording start position becomes clear.

[0012]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic block diagram of a digital VTR according to an embodiment of the present invention.

An input terminal 40 receives an image signal of a moving image to be recorded. The A / D converter 42 digitizes the analog image signal from the input terminal 40. The variable length encoding circuit 44 controls the A / D converter 12 before starting recording.
Are encoded in the frame, and the A / D converter 1
The variable-length coding of the output data in the frame 2 and between the frames is started, and the variable-length coding within and between the frames is stopped according to the recording stop instruction. Details of the variable length encoding circuit 44 will be described later.

The modulation circuit 48 performs low frequency suppression modulation on the output of the variable length coding circuit 44, and the output is amplified to a predetermined level by the recording amplifier 50. The switch 52 is connected to the contact a during recording and to the contact b during reproduction.
The output of 0 is applied to the magnetic head 54 via the switch 52, and is recorded on the magnetic tape 56.

At the time of reproduction, the recording signal of the magnetic tape 56 is reproduced by the magnetic head 54, and its output is
2 and is applied to the demodulation circuit 60 via the reproduction amplifier 58 and demodulated. The variable length decoding circuit 62 is a decoding circuit corresponding to the variable length coding circuit 44,
0 is decoded, and a digitally reproduced image signal is output. The output of the decoding circuit 62 is converted into an analog signal by a D / A converter 64, and an analog reproduced image signal is output from an output terminal 66 to the outside.

FIG. 2 is a block diagram showing the configuration of the variable length coding circuit 44 which is a characteristic part of the present embodiment, and particularly shows the quantization and prediction coding unit in detail. 70 is A / D
The input terminal 72 to which the output of the converter 42 is input is a device that converts the raster-scanned image data from the input terminal 70 into horizontal i pixels ×
This is a blocking circuit for blocking into blocks of vertical j pixels. i and j are usually about 8 to 16. A delay circuit 74 delays the output of the blocking circuit 72 by three frames.

Reference numerals 76, 78 and 80 denote subtracters for calculating a prediction error for predictive difference encoding. The subtracter 76 subtracts the local decoded value three frames before from the output of the blocking circuit 72, and A subtractor 78 subtracts image data obtained by interpolating and synthesizing the local decoded values of one frame before and four frames before from the output of the three-frame delay circuit 74. And the image data obtained by interpolating and synthesizing the local decoded value five frames before.

A switch 82 selects the output of the blocking circuit 72 (a contact), the output of the subtractor 76 (b contact), the output of the subtractor 78 (c contact), or the output of the subtractor 80 (d contact). It is.

Reference numeral 84 denotes a DCT circuit for performing discrete cosine transform of the data selected by the switch 82, and reference numeral 86 denotes a DCT circuit.
A quantization circuit 88 quantizes the output (frequency coefficient) of the T circuit 84 at a different quantization step for each frequency coefficient, and an inverse quantization circuit 88 inversely quantizes the output of the quantization circuit 86. The quantization step size in the quantization circuit 86 greatly affects the information compression ratio. The characteristics of the quantization circuit 86 and the inverse quantization circuit 88 can be changed by a control coefficient from the input terminal 90. Normally, the control coefficient is determined according to the occupancy of the subsequent data buffer, and the quantization characteristics of the quantization circuit 86 and the inverse quantization circuit 88 are feedback-controlled.

Reference numeral 92 denotes an entropy coding circuit for performing entropy coding (for example, Huffman coding) on the output of the quantization circuit 86 by using the statistical property of zero continuous data. Reference numeral 94 denotes an output of the entropy coding circuit 92. This is an output terminal supplied to the modulation circuit 48 of FIG.

Reference numeral 96 denotes an inverse DCT circuit for performing an inverse discrete cosine transform of the output of the inverse quantization circuit 88, and 98 an inverse DCT circuit 96
Adder that adds zero or a predetermined predicted value to the output of
00 is a delay circuit for delaying the output of the adder 98 by three frames, 102 is a product-sum operation for performing weighted product-sum operation on the output of the adder 98 and the output of the three-frame delay circuit 100, and outputting interpolated synthesized data. A circuit 104 is a delay circuit for delaying the output of the product-sum operation circuit 102 by one frame, and a delay circuit 106 is for delaying the output of the delay circuit 104 by one frame.

Reference numeral 108 denotes a zero (a contact), the delay circuit 10
A switch 110 for selecting an output of 0 (b contact), an output of the delay circuit 104 (c contact), or an output of the delay circuit 106 (d contact) is an operation signal input from the operation switch 46 via the input terminal 112. Switches 82 and 1
08 is a control circuit. Switches 82 and 108
Is connected to the contact a regardless of the frame before the start of recording, but is switched for each frame from the start of recording in the switching sequence shown in FIG.

Although details will be described later, the switches 82 and 108
Is connected to the contact a, intra-frame coding when connected to the contact b, inter-frame coding of a difference of three frames when connected to the contact b, and a frame based on an interpolated composite value of two frames before and after when connected to the contact c or d. It becomes inter coding (bidirectional coding).

The operation of the circuit of FIG. 2 will be described with reference to FIG. 5A shows the frame order of image data input to the input terminal 70, and FIG. 5B shows the frame order of image data recorded on the magnetic tape 56.

Frame # immediately after switch operation for starting recording
At 7, the switches 82 and 108 are connected to the a contacts as shown in FIG. The blocking circuit 72 converts the raster-scanned image data from the input terminal 70 into a block row of i × j pixels, and the output is applied to a contact a via a switch 82, a subtractor 76 and a delay circuit 74. . When the block circuit 72 outputs the image data of the frame # 7,
The delay circuit 74 outputs the image data of the frame # 4 three frames before.

The DCT circuit 84 converts the image data blocked by the blocking circuit 72 into the frequency domain by discrete cosine transform, and outputs a transform coefficient. The quantization circuit 86 quantizes the output of the DCT circuit 84 with a different quantization step size for each transform coefficient. The quantization step size of the quantization circuit 86 is controlled by a control coefficient from the input terminal 90.

The entropy encoding circuit 92 entropy encodes the output of the quantization circuit 86, and the output is supplied to the modulation circuit 48 of FIG. The image at this time is an intra-coded frame (hereinafter, referred to as an I-frame) that has been compression-coded in one frame.

The inverse quantization circuit 88 inversely quantizes the output of the quantization circuit 86, and the inverse DCT circuit 96 performs the inverse DCT transformation on the output of the inverse quantization circuit 88. Since the switch 108 is connected to the contact a, the adder 98 outputs the output of the inverse DCT transform circuit 96 as it is. The output of the adder 98 is applied to the three-frame delay circuit 100 and the product-sum operation circuit 102.

At this point, the output of the delay circuit 100 is the locally decoded image data of the frame # 4, and the product-sum operation circuit 102 outputs the interpolated synthesized image data by the weighted product-sum operation of the frames # 7 and # 4. Is output.

When the frame # 8 of the second frame is input, the switches 82 and 108 are connected to the c-contact as shown in FIG. That is, the switch 82 selects the output of the subtractor 78. At this point, the delay circuit 74 sets the frame # 5
And the delay circuit 100 outputs the frame # 5
And the delay circuit 104 outputs interpolated and synthesized data of the frames # 7 and # 4. The subtractor 78 subtracts the interpolated combined data (bidirectional predicted image data) of the frames # 7 and # 4 from the image data of the frame # 5, and the output is applied to the DCT circuit 84 via the switch 82. .

The output of the subtractor 78 is subjected to discrete cosine transform and quantization by a DCT circuit 84 and a quantization circuit 86. The entropy encoding circuit 92 entropy encodes the output of the quantization circuit 86, and the output is supplied to the modulation circuit 48 of FIG. The image at this time is differentially coded using the composite value of frame # 4 and frame 7 sandwiching frame of interest # 5 as the predicted value,
Hereinafter, this is referred to as a bidirectional prediction / interpolation frame (abbreviated as B frame).

The bidirectional predictive encoded data of frame # 5 is inversely transformed by an inverse quantization circuit 88 and an inverse DCT circuit 96, and an adder 98 interpolates and synthesizes data of frame # 7 and frame # 4 (bidirectional predictive encoded data). Image data) are added and decoded. The decoded image data of the frame # 5 is applied to the delay circuit 100 and the product-sum operation circuit 102.

When the third frame # 9 is input, the switches 82 and 108 are connected to the d contact as shown in FIG. That is, the switch 82 selects the output of the subtractor 80. At this time, the delay circuit 74 sets the frame # 6
And the delay circuit 100 outputs the frame # 6
And the delay circuit 106 outputs interpolated and synthesized data of the frames # 7 and # 4. The subtracter 78 subtracts the interpolated combined data (bidirectional predicted image data) of the frames # 7 and # 4 from the image data of the frame # 6, and the output is applied to the DCT circuit 84 via the switch 82. .

The output of the subtractor 78 is output from a DCT circuit 84, a quantization circuit 86, and an entropy encoding circuit 92.
Processing is performed in the same manner as in the previous frame, and is supplied from the output terminal 94 to the modulation circuit 48 of FIG. The image at this time is differentially coded with the combined value of frame # 4 and frame 7 sandwiching frame # 6 of interest as the prediction value.
This is an interpolation frame (B frame).

The bidirectional predictive encoded data of frame # 6 is supplied to an inverse quantization circuit 88, an inverse DCT circuit 96, and an adder 9
8 is decoded. The decoded image data of the frame # 6 is applied to the delay circuit 100 and the product-sum operation circuit 102.

At the point of time when the fourth frame # 10 is input, the switches 82 and 108 are set to b as shown in FIG.
Connect to contacts. That is, the switch 82 selects the output of the subtractor 76. At this point, the delay circuit 100 outputs the locally decoded data of the intra-frame encoding of the frame # 7.
The subtractor 78 subtracts the local decoded value (inter-frame predicted image data) of the frame # 7 from the image data of the frame # 10, and outputs the result via the switch 82 to the DCT circuit 8
4 is applied.

The output of the subtractor 78 is compression-coded by the DCT circuit 84, the quantization circuit 86 and the entropy coding circuit 92, and is supplied from the output terminal 94 to the modulation circuit 48 of FIG. The image at this time is frame # 1 of interest.
It is differentially coded with the decoded value of frame # 7 three frames before 0 as a prediction value, and is hereinafter referred to as an inter-frame coded frame (hereinafter abbreviated as a U-frame).

The inter-frame coded data of frame # 10 is supplied to an inverse quantization circuit 88, an inverse DCT circuit 96, and an adder 9
8 is decoded. The decoded image data of the frame # 10 is applied to the delay circuit 100 and the product-sum operation circuit 102.

Thereafter, the B frame is performed twice, the U frame and the B frame are performed twice, and becomes an I frame. Thereafter, an I frame, a U frame, and a B frame are formed by the same repetition.

Next, the operation when the recording is stopped will be described. In this embodiment, there is a case where the frame after the recording stop instruction time has to be recorded due to the existence of the bidirectional prediction / interpolation frame. For example, in FIG.
It is assumed that the recording stop is input by the operation switch 46 between the frame 5 and the frame # 16. In this case, frame # 15
Is a bidirectional prediction / interpolation frame, so data of frame # 16 is required for restoration. Therefore,
Frames # 16, # 14, 15, etc.
Record up to 16.

FIG. 6 shows the track data on the magnetic tape 56 of the image data of each frame which has been subjected to the variable length coding.
An example of the pattern is shown. Since each frame is variable-length coded, the amount of recording data differs for each frame, and data of one frame often extends over a plurality of tracks. A frame number is added after B, U, and I, which indicate a B frame, a U frame, or an I frame, respectively, and when a plurality of tracks is covered, a child number is further added.

Assume that it is desired to connect another image after frame # 15 in the pattern shown in FIG. In this case, in this embodiment, the user returns the magnetic head to track # 1, which is the first recording track, and reproduces and outputs frames # 8, # 9, and # 10 in order from frame # 7, and returns to frame # 15.
When the image is reproduced and output, the recording start switch of the operation switch 46 is operated. In response to this, the variable-length encoded data of the post-recorded image is sequentially recorded from the head of the track next to the track on which the image of frame # 15 is recorded (track # 6 in FIG. 6).

FIG. 7 shows a recording track pattern after the splicing shooting. For ease of understanding, it is assumed that the post-recording image starts recording from the frame # 7, and n is added to the frame number. That is, I7n indicates that frame # 7 of the post-record image is an I frame.

A mark or signal indicating that splicing has been performed after frame # 15 of the recorded image is recorded on a predetermined portion of the magnetic tape 56, for example, a control track extending in the longitudinal direction of the magnetic tape 56, or the like. deep.

At the time of reproduction, the image of the frame # 7n can be reproduced and output continuously after the reproduction and output of the image of the frame # 15 by the above-described splicing mark or signal.

In the above description, encoding is performed in units of frames, but may be performed in units of fields. Further, the arrangement and number of the intra-coded frames, the inter-coded frames, and the bidirectional prediction / interpolation frames are not limited to the above examples. Also, the variable length coding method is not limited to the method of the above embodiment.

Also, a case where intra-frame coding, inter-frame coding, and bidirectional predictive coding coexist has been described.
The present invention can of course be applied to the case where one or two of these are used.

Although an example has been described in which a magnetic tape is used as a recording medium, it is needless to say that other recording media, such as a magnetic disk, an optical disk, and a magneto-optical disk, are also included in the scope of the present invention. .

[0050]

As can be easily understood from the above description, according to the present invention, since the recording of an image is started from a predetermined position (usually at the head) of a track, the reproduced image can be reproduced even if there is a splicing shot. It is easy to control the reproducing operation so as not to be disturbed, and a continuous reproduced image can be obtained.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of an embodiment of the present invention.

FIG. 2 is a circuit example of a variable length encoding circuit 44;

FIG. 3 is a configuration block diagram of a conventional example.

FIG. 4 is a table showing a switching sequence of switches 82 and 108.

FIG. 5 is an explanatory diagram of an input image and a recorded image.

FIG. 6 is an example of a recording track pattern.

FIG. 7 shows a recording track pattern after a splicing image is taken.

[Explanation of symbols]

10: Image input terminal 12: A / D converter 14: Fixed length encoding circuit 16: Modulation circuit 18: Recording amplifier 2
0: switch 22: magnetic head 24: magnetic tape
26: playback amplifier 28: demodulation circuit 30: fixed length decoding circuit 32: D / A converter 34: playback image output terminal 40: image input terminal 42: A / D converter 44: variable length encoding circuit 46: operation Switch 48: Modulation circuit 50: Recording amplifier 52: Switch 54: Magnetic head 56: Magnetic tape 58: Reproduction amplifier 60: Demodulation circuit 62: Variable length decoding circuit 64: D / A converter
66: reproduced image output terminal 70: input terminal 72: blocking circuit 74: delay circuit 76, 78, 80: subtractor 82: switch 84: D
CT circuit 86: quantization circuit 88: inverse quantization circuit 9
0: control coefficient input terminal 92: entropy coding circuit 94: output terminal 96: inverse DCT circuit 98: adder 100: delay circuit 102: product-sum operation circuit 104,
106: delay circuit 108: switch 110: control circuit 112: operation signal input terminal

Claims (1)

(57) [Claims] (1)Input means for inputting image data; Intra-screen coding that performs variable-length coding using intra-screen correlation
And inter-frame code that performs variable-length coding using correlation between screens
Input by the input means by selectively using
Encodes image data, and encodes a plurality of
A first coded picture and a plurality of second coded pictures
Code to output encoded image data composed of
Encoding means;  Forming a large number of tracks on a tape-shaped recording medium,SaidMark
CodingFrom meansOutputRecord of encoded image data
Recording means,Starting recording of the encoded image data by the recording unit
Instruction means for instructing, The code based on the recording start instruction by the instruction means
Means The recording meansWhenControlControl means, The control unit performs the first recording according to the recording start instruction.
A screen as the first coded screen, the first recording screen
Is recorded from the beginning of the track.
With the encoding means after the first recording screen
The output encoded image data of the first encoded screen is
The first and second encoded images in the encoded image data
At any position in the track according to the amount of data on the surface.
Controlling the encoding means and the recording means to record
 An image recording apparatus, comprising: