JPH0918878A - Image encoding device and method - Google Patents

Abstract

PURPOSE: To encode the video signals of both high and standard fineness in a comparatively small scale constitution of an image encoding device. CONSTITUTION: The information content compression parts 41 to 44 are prepared for the standard fineness. When the video signals of high fineness of a single system are encoded, an image is divided into plural pieces by a division processing part 2 and the data on every divided image are compressed and encoded at the parts 41 to 44. When the quantization is carried out at the parts 41 to 44 under such conditions, the same quantization matrix is used among the data on those divided images. Then the detection range of movement compensation set for the data on the boundary part of every divided image includes even the boundary parts of other adjacent divided images.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image encoding device and an image encoding method for encoding a video signal of a television broadcast signal, for example.

[0002]

2. Description of the Related Art In the fields of compression and transmission of video signals and audio signals such as television broadcasting, the digitization of video signal processing is delayed compared with the digitization of audio signal processing due to the difference in the amount of information. It was But,
With the progress of digitization technology and the progress of standardization work for digitization processing, digitization of video signal processing has begun to spread rapidly.

By digitizing the video signal,
You will be able to enjoy many advantages such as improved image quality, enhanced noise immunity, addition of new services, coupling with other media, etc. Thought to transmit this digitized video signal Since the transmission band is sometimes limited, the transmission side often needs to compress and encode the video signal for transmission. This requires a large-scale hardware capable of high-speed processing.

However, for so-called standard definition video signals such as the NTSC system, the standardization of the encoding algorithm has progressed, and along with this, the necessary circuits have also been integrated into ICs, and by utilizing these, standard definition video is obtained. It is becoming easier to design an encoding device for signals.

By the way, analog high-definition television broadcasting such as high-definition broadcasting (analog HDT
V) is entering the stage of practical application. This analog H
Although it may be possible to handle DTV television broadcast waves in a digitized manner, analog HDTV video signals have a high information content that is about five times that of standard definition video signals such as the NTSC system. It is a video signal of high definition.

Therefore, in the case of compressing and coding an analog HDTV video signal, an expensive high-definition coding device for coding at a higher speed than the standard-definition coding device is required. Become. And analog HDTV
Since its practical application has just begun and its application fields are limited, it is the current situation that the introduction of the above-mentioned encoding device for high definition is not progressing at all.

[0007]

By the way, for example, in the case where the broadcasting side attempts to perform services of both standard definition and high definition television broadcasting by using a digital video signal, As described above, at present, a coding device for signal compression dedicated to standard definition and a coding device dedicated to high definition are required. For this reason, the broadcasting facility becomes large in scale and very expensive.

For example, when performing both standard definition and high definition television broadcasting services, it is conceivable to use the same transmission line at different times. In this case,
If the transmission line is capable of transmitting one line of high definition TV broadcast waves, 4 to 5 lines of standard definition TV broadcast waves with an information amount of about ⅕ of high definition will be transmitted on the transmission line. It can be transmitted. Now, if an attempt is made to transmit four standard definition television broadcasts through the above-mentioned transmission paths, there will be five encoding devices, one for high definition and four for standard definition. Will be required.

Therefore, not only is it costly to arrange the coding devices, but also a switching device and a broadcast signal adjusting device are required for each coding device, the equipment on the broadcasting side becomes large, and the installation space is also a problem. become.

Then, the video signal to be broadcast is
Two types of encoding devices must be used properly depending on whether standard definition or high definition. Therefore, when the operating methods are different between these two encoding devices, handling becomes complicated and erroneous operation is prevented. It can also be a cause.

In view of the above, the present invention provides a single image coding apparatus, which has high definition and standard definition.
It is possible to encode two types of video signals having different amounts of information, and to enable encoding even with a high definition video signal at the same encoding processing speed as a standard definition video signal. With the goal.

[0012]

In order to solve the above problems, in the image coding apparatus according to the present invention, an analog video signal is sampled to obtain image data in which each sample value is pixel data. Then, in an image coding apparatus for removing redundant components, quantizing, compressing information, and coding for transmission, in the image data, a plurality of pixel data in the horizontal direction and vertical direction are set as one block, and this block is set to the minimum. N (N is an integer greater than or equal to 2) information amount compression units for performing information compression as a processing unit, and the N information amount compression units are operated in the first mode or in the second mode. And a coding control unit that selectively controls whether to operate, and each of the N information amount compression units spatially removes the redundant component from the spatial signal of the video signal. Conversion means for performing data conversion for information compression by a function, a first parameter given for each spatial frequency according to a difference in quantization sensitivity for each spatial frequency in human visual characteristics, and a quantization step width And a quantizing means for quantizing the data from the transforming means based on a second parameter for determining
In addition to supplying the first and second parameters or control signals for determining these parameters, in the first mode, the same number of N-information amount compression units are used as the first parameters. In the second mode, the quantization is performed using the above-mentioned quantization.
It is characterized in that the quantization is performed by using independent ones.

In the image coding apparatus according to the second aspect of the present invention, one screen is divided into N parts, and a division processing part for outputting image data of each divided image and a preceding stage of the N information amount compression parts. Image data of each divided image and independent image data are supplied from the division processing unit, and one of the image data of the divided image and the independent image data is supplied to the encoding control unit. And N signal switching means that are taken out according to the switching control from
In the mode 2, the image data of each divided image is supplied from the signal switching means to each of the N information amount compression units to perform compression processing. In the second mode, the signal switching is performed. Independent image data is supplied from each means to each of the N information amount compression units, and compression processing is performed.

Further, in the image coding apparatus according to the third aspect of the present invention, in the image coding apparatus according to the second aspect, each of the N number of information amount compression units is provided from the coding control unit. In the first mode, with respect to the data of the boundary area between each of the divided images with the adjacent image, the second parameter is made substantially the same to perform quantization in the first mode. In the mode, the quantization is performed by the second parameter independent of each other.

Further, in the image coding apparatus according to the invention of claim 4, in the invention of claim 2, the N
Each of the N information amount compression units has a function of performing motion compensation processing in block units at the time of compression encoding, and in the first mode, each of the N information amount compression units is divided into each of the divisions. With respect to the data of the boundary area between the images adjacent to each other, the range spanning the adjacent image areas of each divided image is set as the search range of the motion vector of the motion compensation,
The motion compensation is performed, and in the second mode, the N
Each of the individual information amount compression units is characterized in that the motion compensation is performed by setting the inside of the image region of its own processing target as a search range of the motion vector of the motion compensation.

[0016]

With the image coding apparatus according to the invention described in each of the above claims, in the first mode, for example, compression coding of a high-definition video signal of one system, and in the second mode,
For example, N-system (for example, 4 system) standard-definition video signals are respectively compression-coded. That is, in one image encoding device, by switching between the first mode and the second mode, one system of high-definition compression encoding and N systems of standard-definition video signal compression encoding are performed. Will be possible.

That is, in the first mode, a high-definition video signal is converted into image data of each divided image obtained by dividing one screen thereof into N image data, distributed and supplied to N information amount compression units, and compressed. In the second mode, each of the N systems of standard definition video signals is encoded and supplied to each information amount compression unit for compression encoding.

According to the invention described in claim 1,
In the first mode, for example, the first parameter corresponding to the quantization matrix in MPEG1 or MPEG2 is N
The same information compression unit is used. Therefore, with respect to the divided images that are separately compression-encoded by the N information amount compression units, the encoding process is performed with the same characteristics in consideration of the difference in the quantization sensitivity for each spatial frequency in the human visual characteristics. , When combining, there is no difference between split screens.

Further, in the second mode, each of the N systems of standard definition video signals is supplied to each information amount compression unit and encoded by the independent first parameter.
As a result, the image data of the video signal of standard definition for each one screen is quantized without deteriorating the characteristics of each image, and each image is properly reproduced even when the image is reproduced. It

According to the image coding apparatus of the third aspect of the present invention, in the first mode, for example, for the data of the boundary area between each of the high-definition divided images and the adjacent image, The second parameter corresponding to the quantization scale in MPEG1 and MPEG2 is made substantially the same and the quantization is performed. In the second mode, the image data of the images of the respective standard definition are separated by the independent second data. Quantization is performed according to the parameters.

Therefore, for a high-definition video signal, in the boundary area corresponding to the joint between the divided images which originally form one screen, the quantization steps are almost the same and the quantization is performed, so N An image for one screen formed by the individual pieces of divided image data can be reproduced without making the boundary portion of each divided image conspicuous.

Further, according to the image coding apparatus of the present invention as defined in claim 4, in the first mode, when the motion compensation of each of the N divided image data is performed, the image data of the N divided image data is compared with other adjacent divided data. At the boundary part, the motion vector detection range for motion compensation is set including the boundary part of other adjacent divided image data. As a result, each of the N pieces of image data is properly encoded as image data forming one screen. Therefore, an image for one screen that is decoded and formed by N divided images can be reproduced as a natural image.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the image coding apparatus and the image coding method according to the present invention will be described below with reference to the drawings.

The image coding apparatus of this example is a high-definition video signal (hereinafter referred to as a high-definition television broadcast, for example).
A first mode for encoding an HD video signal (hereinafter referred to as an HD mode) and a video signal of a standard definition such as an NTSC television broadcast (hereinafter referred to as S).
A second mode (hereinafter referred to as SD mode) for encoding a D video signal) is provided, and both modes can be switched according to the user's request.

The image compression coding described below is as follows.
This is an example in the case of compression encoding conforming to MPEG1 or MPEG2, or close to them.

The image coding apparatus of this example will be described in detail later, but N pieces each having the ability to compress and code one system of SD video signal, four in this example.
It is provided with individual information amount compression units. Therefore, it is possible to simultaneously encode four different SD video signals.

In the image coding apparatus of this example, when the HD video signal is coded, the HD video signal is divided into image data of a divided image in which one screen is divided into four, and these 4 are divided. The divided image data is compressed and encoded in each of the four information amount compression units. That is, by dividing the HD video signal, the data amount of each divided image is reduced to such an extent that compression encoding processing can be performed by the information amount compression unit for SD video signals, and the data of each divided image is reduced to one. The information amount compression unit can process the information. That is, even HD video signals can be coded at the coding speed of SD video signals.

FIG. 1 is a block diagram for explaining an embodiment of an image coding apparatus according to the present invention. As shown in FIG. 1, the image coding apparatus of this example includes an A / D converter 1 for HD video signals and an A / D converter for four SD video signals.
D converters 11 to 14 and HD video signal division processing unit 2
, Four signal switches 31 to 34, four information amount compression units 41 to 44, an encoding control unit 5, and a multiplexer 6
And a channel encoding unit 7.

The A / D converter 1 samples the analog HD video signal V1 supplied thereto at a sampling frequency corresponding to the HD video signal, and converts the sampled pixel values into digital signals. Digital HD image data is formed as pixel data. Formed digital H
The D image data is supplied to the division processing unit 2.

The division processing section 2 divides the digital HD image data into four in units of data for one screen. That is, as shown in FIG. 2, one screen is equally divided into two parts in the horizontal direction and two parts in the vertical direction, and is divided into four divided image data H1, H2, H3, and H4 for 1/4 screens. However, in this example, as will be described later, each output divided image data from this division processing unit 2 is arranged so that the motion vector search area for the motion compensation processing of the image data can be extended to the adjacent divided screens. H1 to H4 are designed to include image data of a boundary portion between other adjacent divided image data.

FIG. 2 is a conceptual diagram for explaining the division processing of the HD video signal in the division processing section 2, which is an area 80.
Indicates an area for displaying an image corresponding to the HD video signal for one screen, and divided areas 81 to 84 indicate areas of each divided screen obtained by dividing the area 80 of one screen into four.

As shown in FIG. 2, the split screen 81 has a display area having a width a1 in the left half of one screen in the horizontal direction and a width b1 in the upper half in the vertical direction of one screen.
2 is the width a2 (a2 = a) of the right half of one screen in the horizontal direction.
1) and a display area having an upper half width b1 in the vertical direction. Similarly, the split screen 83 has a width a1 in the left half of one screen in the horizontal direction and a width b in the lower half in the vertical direction of one screen.
It has a display area of 2 (b2 = b1) and has a split screen 8
Reference numeral 4 has a display area having a width a2 in the right half of one screen in the horizontal direction and a width b2 in the lower half in the vertical direction of one screen.

The output divided image data H1 includes the pixel data in the divided area 81 as the pixel data to be processed, and the adjacent areas 82 to 84 of the area 81 in FIG. 2 are used for the motion compensation processing. It includes the pixel data of the boundary portion with diagonal lines, and becomes the image data of the area indicated by horizontal Y1 and vertical T1 in FIG. The output divided image data H2 includes the pixel data in the divided area 82 as the pixel data to be processed, and also the area 82 for the motion compensation processing.
2 including the pixel data of the boundary portion of the adjacent regions 81, 83, 84 of FIG.
It becomes the image data of the area shown by.

The output divided image data H3 includes the pixel data in the divided area 83 as the pixel data to be processed, and for the motion compensation processing, the adjacent areas 81, 82 and 84 of the area 83 in FIG. The image data of the area including the pixel portion of the boundary with hatching is shown by the horizontal Y1 and vertical T2 in FIG. The output divided image data H4 includes the pixel data in the divided area 84 as the pixel data to be processed, and the adjacent areas 81, 82, and 83 of the area 84 are shaded in FIG. 2 for the motion compensation processing. The image data of the area including the boundary portion and the area indicated by horizontal Y2 and vertical T2 in FIG. 2 is obtained.

Then, in the division processing section 2, the divided image data H1 of the HD video signal divided as described above.
Each of H4 is supplied to each of the signal switchers 31 to 34.

On the other hand, the A / D converters 11-14 are provided with SD video signals V11-V provided as analog signals thereto.
14 is sampled at a sampling frequency corresponding to the sampling frequency, and the pixel samples that are the sampled values are converted into digital signals to form digital SD image data S11 to S14. Then, the digital SD image data S11 to S14 from these A / D converters 11 to 14 are supplied to the signal switchers 31 to 34.

The signal switching units 31 to 34 are the coding control unit 5
Based on the switching control signal SW depending on the encoding mode of HD mode or SD mode, the divided HD video signal is supplied (HD mode) or the SD video signal is output (SD Mode). The output data from each of the signal switchers 31 to 34 is supplied to the information amount compression units 41 to 44, respectively.

The four information amount compression units 41 to 44 have exactly the same configuration. Then, the information amount compression unit 41
To 44 are motion compensation as an example of a process for reducing the redundancy of the input image data in the time direction, and an orthogonal transformation as an example of a process for reducing the redundancy of the image data in the spatial direction. , The transformation coefficient obtained by the orthogonal transformation is expressed by a small value, and the quantization processing for reducing the code amount,
A series of image data compression processing such as variable length coding processing in which the quantized transform coefficient is represented by a code length based on the occurrence frequency is performed.

Each of the information amount compression units 41 to 44 performs H processing for the divided image data of the HD video signal.
A control signal is received from the encoding control unit 5 according to the D mode and the SD mode for processing the image data of the SD video signal, and the processing according to each mode is performed. Although details will be described later, the basic compression processing is almost the same between the HD mode and the SD mode, but the setting of the search range for motion detection of motion compensation, the determination of the quantization step at the time of quantization, etc. The mode differs from the SD mode.

Although not shown, the encoding control unit 5 is a CPU.
And a microcomputer equipped with ROM, RAM. Then, the mode switching signal MD supplied thereto is used to determine whether the mode is the HD mode or the SD mode, and the signal switching units 31 to 34, the information amount compressing units 41 to 44, the multiplexer 6, and the communication according to each processing mode. In addition to controlling the path coding unit 7, the information amount compression units 41 to 44
Is also controlled so that the generated code amount can be transmitted through the communication path.

The multiplexer 6, under the control of the encoding control unit 5, in the HD mode, the information amount compression units 41 to 44.
A process of synthesizing the HD divided image data supplied from to obtain one screen of image data is performed, and in the SD mode, a process of generating data for transmitting the respective SD image data arranged in order, for example, is performed. To do.

Then, under the control of the encoding control unit 5, the communication encoding unit 7 is an additional unit for transmitting HD image data or SD image data respectively through the communication path in accordance with the HD mode and the SD mode. Processing such as addition of information and conversion into a predetermined data format is performed.

Next, the information amount compression units 41 to 44 will be described in more detail by taking an example of the internal configuration of the information amount compression units 41 to 44. As described above, the four information amount compression units 41 to 44 all have the same configuration, and therefore, in the following description, the information amount compression unit 41 will be described as a representative.

The information amount compression unit 41 performs the intra-frame compression process using the spatial correlation of the images and the inter-frame compression process using the correlation between the screens, that is, the correlation in the time direction. For the former spatial compression processing, data is orthogonally transformed to frequency axis data,
After quantizing the information according to the magnitude of each frequency component,
Further, variable length coding (entropy coding) is performed.
In this example, DCT (discrete cosine transform) is used for orthogonal transform.

For the latter compression, a motion compensation process is performed. For example, with a small area region within one screen as a unit, for example, images of the previous screen and the next screen are compared for that portion, and a motion vector indicating the amount and direction of movement of that area region and the amount of change in that area region are compared. And the difference is included in the transmission information. In this case, the difference is D
CT calculation is performed, compressed and transmitted.

FIG. 3 is a block diagram for explaining an example of the configuration of the information amount compression unit 41. The information amount compression unit 41 is shown in FIG.
Is a DCT / motion detector 401 and a quantizer 402.
And a variable length coding unit 403.

In the DCT and motion detector 401, the DC
T calculation and motion detection (motion vector detection and difference detection)
Before that, the image data is divided into image processing units of an area area of a predetermined size. In this example, as shown in FIG. 4, the image area is divided into blocks each having a total of 64 pixels, that is, 8 pixels in the horizontal direction and 8 pixels in the vertical direction. Hereinafter, the block of this image area unit is simply referred to as a block.

As shown in FIG. 4, an area of four blocks (horizontal × vertical = 16 pixels × 16 pixels = 256 pixels) adjacent to each other for the luminance component is called a macroblock. This macroblock includes information for one block for each of the two types of color difference signal components Cb and Cr.

The DCT / motion detecting section 401 converts the input data into DCT coefficients for each frequency component by DCT calculation for each block. In this case, there are two types of target data for DCT calculation.

That is, when the frame to be processed is a so-called intra frame in which the image data for one screen is compressed as it is, the image data in block units consisting of 64 pixel data are DCT-operated respectively. In addition, a so-called inter-frame in which a frame to be processed detects a motion between the frame and a preceding frame, a subsequent frame, or both frames before and after the frame, and compresses information according to the detected motion. In this case, the DCT calculation is performed on the difference between the preceding and following frames of the block unit and the corresponding blocks.

Therefore, the DCT and motion detector 401
Then, when the processing target frame is an inter-frame, motion detection is performed using the frames before and after the inter-frame.
This motion detection includes detection of a motion vector on a macroblock basis and detection of a difference on a block basis.

In this case, the motion detection processing is performed not only at the same position of the processing target frame and the preceding frame and / or the succeeding frame, but also as shown in FIG.
Based on the position of the target macroblock MB, a motion search is performed within a range of ± n pixels, for example, ± 15 pixels in the horizontal and vertical directions on the screen of the frame to be motion-detected. In FIG. 6, a range AS indicated by a dotted line is a motion search range.

In this case, motion detection is performed as follows. That is, the difference between the target macroblock MB and the corresponding macroblock area of the previous frame and / or the subsequent frame is calculated. In this case, in the search range AS near the target macroblock of the target frame, for example, the difference between each macroblock area whose position is different by one pixel and the corresponding macroblock area of the previous frame and / or the subsequent frame is calculated. Ask for everything. Then, the macro block area having the smallest difference is detected, and the motion amount and the motion direction between the detected macro block area and the target macro block MB are detected,
The motion vector Ve is detected.

FIG. 5 is a diagram for explaining the detection of this motion vector, which is a case of detecting the motion vector between the macroblock MB of the processing target frame FLo and the subsequent frame FLa. In the example of, the case where the macroblock MB moves to the position of MB ^* is shown.

As described above, the information amount compression units 41 to 44
In each of the above, the range of motion detection may differ between the SD mode and the HD mode. Therefore, the DCTs of the information amount compression units 41 to 44 and the motion detection unit 401 are
As shown in FIGS. 4 and 1, the control signals MC1 to MC4 are supplied from the encoding control unit 5, respectively.

In the SD mode, each of the information amount compression units 41 to 4
In 4, the information compression processing of the independent SD image data is performed, so that the motion detection area is performed within the range of each one screen area. Therefore, the macroblocks at the upper, lower, left, and right ends of one screen have a narrower range than the above-described area AS as a motion detection search area.

On the other hand, in the HD mode, each information amount compression unit 4
In 1 to 44, as described above, the data of the image area of ¼ of the HD image data is the processing target.
However, as shown in FIGS. 2 and 7, the four 1/4 HD image data are originally the information amount compression units 41 to 44.
The data handled in (1) should be handled as one processing unit.

Therefore, in the present invention, in the HD mode, as shown in FIG. 2 and FIG. 7, the motion detection is performed in the adjacent 1/4 HD image data in the mutually adjacent boundary areas. It should be performed in a range that straddles the range of image data.

That is, in the case of the information amount compression unit 41, the range shown by hatching in FIG. 7 is also used as a search area for motion detection. Therefore, as described above, the division processing unit 2 includes the data of the hatched area and outputs the data as 1/4 HD image data. In the case of this example, as shown in FIG.
Since the search area AS has an extra area of n pixels above, below, to the left and right of the area of the macroblock MB, the width of the hatched portion in FIG. The pixel data of the width area is included in the output data H1 and is output.

As a result, even in the macroblocks of the boundary areas adjacent to each other of the 1/4 HD image data, motion detection can be performed in the same search area as when processing one original screen (see FIG. 7). . Therefore, the motion compensation is appropriately performed even at the boundary portion of the divided images adjacent to the other divided images, and even if the four divided images are joined and the image of the HD video signal for one screen is reproduced. , Does not deteriorate the image.

As described above, D of the information amount compression unit 41
The motion vector Ve detected by the CT and motion detection unit 401 is output from the information amount compression unit 41. Further, the detected block unit difference is set as a processing target of the DCT calculation, and is converted into the DCT coefficient as described above.

In this case, the obtained DCT coefficient is composed of the DC coefficient of the DC component and the AC coefficient of 63 AC components (AC1 to AC63) as shown in FIG. They are,
The quantizer 402 quantizes each.

Of the DCT coefficients, the DC coefficient encodes the difference value with the DC coefficient of the immediately preceding block as the predicted value.
Further, as shown in FIG. 8, the AC components AC1 to AC63 are quantized sequentially by scanning in a zigzag order from the lowest frequency.

In the quantizing unit 402, at the time of this quantization,
The quantization signal and the quantization scale are determined by the control signal QC1 from the encoding control unit 5, the quantization step for each coefficient is determined, the quantization is executed, and the information amount is compressed.

The quantization matrix is used to realize efficient coding using the difference in the quantization sensitivity for each spatial frequency for the image on the human visual characteristic. 9 corresponding to the DCT coefficients shown in FIG. 8 obtained by the DCT calculation in FIG. 8, that is, the DC coefficient and the AC coefficients AC1 to AC63, respectively.
9 is a table of values Kdc and K1 to K63 set as shown in FIG. In this case, the values Kdc and K1 to K63 of the quantization matrix are set to be larger as the frequency component of the coefficients AC1 to AC63 is larger, and the final quantization step is set to be larger.

Each value of this quantization matrix is a DCT
Each coefficient is, so to speak, a coefficient for performing the first-stage quantization. Several kinds of this quantization matrix are prepared in a memory such as a ROM provided in the quantization unit 402, and are switched for each frame in this example according to the nature of the image by the control signal QC1 from the encoding control unit 5. It is possible.

It is also possible to provide a quantization matrix memory on the side of the encoding control unit 5, select it as appropriate, and send it to each of the information amount compression units 41 to 44.

The quantization scale is a parameter for the quantization unit 402 to determine the width of the final quantization step, and can be changed in macroblock units by the control signal QC1 from the encoding control unit 5. . As will be described later, the value of the quantization scale is determined by the encoding control unit 5 based on the code generation amount information Qt1 from the variable length encoding unit 403 and is included in the control signal QC1 to be included in the quantization unit 402. Supply to. By controlling the quantization scale in this way, the generated code amount is too large for the transmission capacity,
Further, the code amount is controlled so that the code amount is not too small and is always appropriate.

Quantized DCT from quantizer 402
The coefficient is supplied to the variable length coding unit 403. The variable-length encoding unit 403 encodes the quantized DCT coefficient into a variable-length code having a code length based on the occurrence frequency of Huffman code (a code having a higher occurrence frequency has a shorter code length), and further Compress information. Then, the variable length coding unit 40
3 calculates the generated code amount and sends the information Qt to the encoding control unit 5.

Then, as described above, in each of the four information amount compression units 41 to 44, the SD mode and H
The quantization mode is controlled by the control signals QC1 to QC4 from the encoding control section 5 so that the control in the quantization section 402 is different between the D mode and the D mode.

That is, in the SD mode, the quantization processing in the quantizing section 402 of each of the information amount compressing sections 41 to 44 can be performed independently, and the quantization matrix is Different independent quantizers 402 can be used in the respective information compressors 41 to 44. In addition, the quantization scale is also controlled so as to be a quantization step corresponding to the information Qt1 to Qt4 of the code generation amount from the variable length coding unit 403 of each of the information amount compression units 41 to 44, and the generated code amount. Will be appropriate.

On the other hand, in the HD mode, the quantizing matrix used by the quantizing section 402 of each of the four information amount compressing sections 41 to 44 is set to be the same by the encoding control section 5. It is controlled by the control signals QC1 to QC4.

In the HD mode, the quantization scale is the same as that in the SD mode except for the image data other than the hatched boundaries in FIG.
Quantization unit 4 of each of four information amount compression units 41 to 44
02, the code generation amount information Qt
The quantization step is controlled according to each of 1 to Qt4.

On the other hand, in the HD mode, with respect to the image data at the boundary portion indicated by the diagonal lines in FIG. 2, the encoding control unit 5 causes the generated code amounts Qt1 to Qt4 from the four information amount compression units 41 to 44. , All four quantizers 402 of the information amount compressors 41 to 44 set a common quantizer scale, or a quantizer scale corresponding to each generated code amount Qt1 to Qt4. However, substantially the same quantization scale is set, and the information of the quantization scale is included in the control signals QC1 to QC4 and supplied to each quantization unit 402.

As described above, the image coding apparatus of this example has an H video processing method in which the HD video signal for one screen is divided into four.
In the case of the D mode, each of the quantization units 402 of the four information amount compression units 41 to 44 is controlled by the encoding control unit 5 to use a common quantization matrix, and thus each of the four divided regions is divided. The DCT coefficient for each spatial frequency of the image data is quantized by the same quantization matrix, and the four divided images are treated as one screen of image data, but the processing is equivalent to four information amounts. It was done in the compression section.

Therefore, the average luminance level and color tone do not differ among the four divided image data, and the HD video signal for one screen can be made equal to that quantized by one information amount compression unit. it can. Therefore, even at the time of reproduction, each divided image can be reproduced as an image forming one screen and having a uniform brightness level and hue.

Further, since the quantization scale is controlled so as to be the same or substantially the same in the boundary area of each divided image, it is possible to prevent the seam of the boundary portion from being conspicuous. That is, when the quantization scale is used independently in each of the information amount compression units 41 to 44 according to the information Qt1 to Qt4 of the generated code amount, depending on the content of the image, Although the quantization step is greatly different between the divided images and the joints may be conspicuous, in this example, common information is obtained by referring to all the information Qt1 to Qt4 of the generated code amounts of the four information amount compression units 41 to 44. Since the quantization scale or almost the same quantization scale is determined and used by the four information amount compression units 41 to 44 to determine the quantization step width, the joint is not conspicuous.

Thus, the image coding apparatus of this example is
Both the HD video signal and the SD video signal can be encoded, and the HD video signal and the SD video signal have different motion compensation and quantization processing as described above. The flow of processing of the coding control unit 5 of the image coding apparatus of this example described above is shown in the flowchart of FIG. This flowchart will be described below.

For example, when an HD video signal or an SD video signal is supplied to the image coding apparatus of this example, the coding control section 5 determines whether the video signal to be processed is an HD video signal, that is, whether it is in the HD mode. , SD mode is determined (step 101). The determination process here is based on, for example, the information specifying the type of the video signal input by the operator of the image encoding device of this example.
It is determined by the mode switching signal MD supplied to the.

When it is determined in step 101 that the video signal to be processed is the HD video signal and is in the HD mode, the signal switches 31 to 34 are switched to select the divided HD video signal. (Step 102). As a result, each of the divided image data of the HD video signal divided into four is supplied to the corresponding one of the four information amount compression units 41 to 44.

Next, the encoding control unit 5 includes the information amount compression unit 4
The DCTs 1 to 44 and the motion detection unit 401 are controlled to set a search region for motion detection for motion compensation including adjacent divided image data at the time of motion compensation. And (step 10
3).

Then, the coding control unit 5 controls the quantizing unit 402 of each of the information amount compressing units 41 to 44 to use a common quantizing matrix, and at the boundary portion between adjacent divided images. , The quantization scales are equal, or
The control is performed so that they are almost the same (step 104).

As described above, in the processing of step 104, the setting of the quantization scale is controlled so as to be performed in consideration of the generation amount of the variable length coded by the variable length encoder 403. is there.

When the quantizer 402 completes the quantization, the quantized data (transform coefficient) is supplied to the variable length encoder 403, and variable length encoding is performed (step 105). As described above, the multiplexer 6
Is transmitted via the channel encoder 7.

On the other hand, when it is determined in step 101 that the video signal is not an HD video signal, it is determined that the video signal to be processed is the SD video signal and is in the SD mode, and the signal switches 31 to 34 are set to A / A. D converter 1
Switching is performed so as to select independent SD video signals supplied from 1 to 14 (step 106). As a result, each of the four SD video signals for one screen is supplied to the corresponding information amount compression units of the four information amount compression units 41 to 44.

Next, the encoding control unit 5 has the information amount compression unit 4
The DCTs 1 to 44 and the motion detection unit 401 are controlled to perform motion compensation only within the range of each SD image data (step 107).

Then, the encoding control unit 5 controls the quantizing unit 402 of each of the information amount compressing units 41 to 44 so as to perform the quantization by independently using the quantization matrix (step 108). . In the processing of this step 108,
In the quantizing unit 402 of each of the information amount compressing units 41 to 44, different quantizing matrices can be used depending on the image to be processed, and the image data corresponding to the SD video signal for each one screen is the other SD. It is quantized without being affected by the image data corresponding to the video signal. The quantization scale is also set for each SD video signal.

Then, the quantized data is supplied to the variable length encoder 403, and the processing from step 105 is performed.

As described above, on the receiving side that receives a signal in which the image of the HD video signal for one screen is divided into four, and is encoded, each divided image data is divided into four and encoded. Similar to the conversion side, the image of the HD video signal for one screen is reproduced by expanding and decoding in consideration of the boundary portion of the adjacent divided image data.

Further, when four different SD video signals are simultaneously transmitted through a transmission line having a capacity capable of transmitting one system of HD video signals having a wide transmission band, the receiving side
It is also possible to select and reproduce one SD video signal image among them, or divide the image display unit of the receiving device on the receiving side into four or two to divide a plurality of SD video signal images. Can also be displayed.

In the above example, the HD video signal for one screen is divided into four, but it may be divided into six, eight or the like depending on the capacity of the transmission path.

[0092]

As described above, according to the image coding apparatus and the image coding method of the present invention, one system of HD video signal (high definition video signal) and a plurality of systems of SD video signal (standard) Definition video signal).

In this case, according to the present invention, when the HD video signal is encoded, the HD video signal for one screen is divided into a plurality of screens and the image data to be processed is reduced to perform the encoding process. Therefore, even if it is an HD video signal, the processing according to the encoding of the SD video signal
It can be encoded.

Therefore, the HD video signal encoding device can be formed by using the conventional IC circuit for encoding the SD video signal. Further, it is possible to provide an image encoding device that is relatively small and easy to handle.

Further, when the HD video signal is encoded, since the quantization matrix is commonly used for each of the divided image data, even if the divided image data is divided, it is regarded as one image data. It will be the same as when encoding was performed.

In addition, since the quantization scale is set to be equal to or almost the same as the data in the boundary areas of the other divided screens in the boundary areas of the divided images, the quantization scales in the boundary areas of the respective divided areas are set. It is possible to reduce variations and prevent seams between divided images from becoming conspicuous.

Further, since the search area for motion detection is made to extend over a part of the adjacent divided image areas, the same motion detection as that when motion is detected as one image data is possible. Therefore, there is no variation in the boundary area of the divided images, and the joint is not noticeable.

[Brief description of the drawings]

FIG. 1 is a block diagram for explaining an embodiment of an image encoding device according to the present invention.

FIG. 2 is a conceptual diagram for explaining an image division process of an embodiment of an image encoding device according to the present invention.

FIG. 3 is a block diagram of a configuration of an example of a main part of an embodiment of an image encoding device according to the present invention.

FIG. 4 is a diagram for explaining an encoding processing unit for an image in the embodiment of the image encoding device according to the present invention.

[Fig. 5] Fig. 5 is a diagram for describing motion compensation executed in an embodiment of the image encoding device according to the present invention.

[Fig. 6] Fig. 6 is a diagram for explaining a search range for motion detection of motion compensation executed in an embodiment of the image encoding device according to the present invention.

FIG. 7 is a diagram for explaining a search range for motion detection of motion compensation executed in an embodiment of the image encoding device according to the present invention.

FIG. 8 is a DC example of orthogonal transformation for image data.
It is a figure for demonstrating the output obtained by T.

FIG. 9 is a diagram for explaining a quantization matrix.

FIG. 10 is a flowchart for explaining the flow of processing in an embodiment of the image encoding device according to the present invention.

[Explanation of symbols]

 1 A / D converter for HD video signal 2 Division processing part 11-14 A / D converter for SD video signal 31-34 Signal switcher 41-44 Information amount compression part 5 Encoding control part 6 Multiplexer 7 Communication Path coding unit 401 DCT and motion detection unit 402 Quantization unit 403 Variable length coding unit

Claims (7)

[Claims] 1. An image coding apparatus for sampling an analog video signal to form image data in which each sample value is pixel data, removing redundant components, quantizing and compressing information, and coding for transmission, With respect to the image data, a plurality of pixel data in the horizontal direction and the vertical direction are set as one block, and N (N is an integer of 2 or more) information amount compression units for performing information compression with this block as a minimum processing unit, An encoding control unit that selectively controls whether to operate the N information amount compression units in the first mode or the second mode, and each of the N information amount compression units, In order to remove the redundant component, conversion means for performing data conversion for information compression by spatial correlation of video signals for each block, and an amount for each spatial frequency on human visual characteristics. Quantizing means for quantizing the data from the converting means based on a first parameter given for each spatial frequency according to a difference in quantization sensitivity and a second parameter for determining a quantization step width. And the encoding control unit supplies the first and second parameters or control signals for determining these parameters to the N information amount compression units, and In the mode, the N information amount compression units use the same parameter as the first parameter to perform the quantization, and in the second mode, the N information amount compression units, An image coding apparatus, wherein independent quantization parameters are used as the first parameters to perform the quantization. 2. A division processing unit that outputs image data of each divided image obtained by dividing one screen into N, and a division processing unit that is provided in front of the N information amount compression units and that outputs each divided image from the division processing unit. Image data and independent image data are supplied, and N signal switching means for extracting one of the image data of the divided image and the independent image data according to switching control from the encoding control unit. In the first mode, the image data of each divided image is supplied from the signal switching unit to each of the N information amount compression units to perform compression processing, and the second mode 2. The image coding according to claim 1, wherein in the mode, independent image data is supplied from the signal switching means to each of the N information amount compression units to perform compression processing. Location. 3. The image coding apparatus according to claim 2, wherein each of the N information amount compression units receives the control signal from the coding control unit and, in the first mode, each of the divisions. For the data in the boundary area between the images adjacent to each other, the second parameter is made substantially the same and quantization is performed. In the second mode, the quantization is performed by the independent second parameters. An image coding apparatus characterized by being performed. 4. The N information amount compression units have a function of performing motion compensation processing in block units, and in the first mode, each of the N information amount compression units includes With respect to the data of the boundary area between the divided images adjacent to each other, the range spanning the adjacent image areas of each divided image is set as the search range of the motion vector of the motion compensation, and the motion compensation is performed. In the second mode, each of the N information amount compression units performs the motion compensation by setting the inside of the image region to be processed by itself as the motion vector search range of the motion compensation. The image encoding device according to claim 2. 5. An image coding method in which an analog video signal is sampled to form image data in which each sample value is pixel data, redundant components are removed, quantization is performed, information is compressed, and coding is performed for transmission. With respect to the image data, a plurality of horizontal and vertical pixel data are set as one block, and data conversion for information compression is performed for each block by spatial correlation of video signals. Based on the first parameter given for each spatial frequency according to the difference in the quantization sensitivity for each spatial frequency and the second parameter for determining the quantization step width, the data from the conversion means is converted. N (where N is an integer of 2 or more) information quantity compression units for quantization are prepared to encode an image signal having more information quantity than a standard image signal. Cormorants In the first mode, the image data of each divided images to one screen has been divided into N supplied by distributing to each of the information amount of compression unit in each information volume compression unit, the first
In the second mode in which the same quantization is performed in the N information amount compression units and the standard image signal is encoded in each of the N information amount compression units, Independent image data can be supplied to each of the N.sub.1 and N.sub.2, and each of the N information amount compression units uses the independent one as the first parameter to perform the quantization. Image coding method. 6. In the first mode, each of the N information amount compression units sets the second parameter at the time of quantization of data in a boundary area between each of the divided images and an adjacent image, Quantization is performed in substantially the same manner, and in the second mode, each of the N information amount compression units performs quantization with an independent second parameter. The image encoding method according to claim 5. 7. The N information amount compression units have a function of performing motion compensation processing in block units at the time of compression encoding, and in the first mode, the N information amount compression units. For each of the divided images, for the data of the boundary region between the adjacent images, the range that spans the adjacent image regions of each divided image is set as the search range of the motion vector of the motion compensation. Then, the motion compensation is performed. In the second mode, each of the N information amount compression units limits a search range of the motion vector of the motion compensation within an image region to be processed by itself. The image coding method according to claim 5, wherein the motion compensation is performed.