JPH1051767A - Image playback device - Google Patents

Abstract

(57) [Summary] An image is divided into a plurality of coding blocks,
Calculate a “reference value” and “range” indicating the dynamic range of the coding block, and calculate a “quantization step width” in the coding block based on the “range” and a predetermined “quantization bit number”. A memory reduction coding unit that calculates a “requantized value” of a pixel belonging to the coding block based on the “reference value” and the “quantization step width”; and the “reference value” and the “range”. And a memory means for storing the re-quantized value, and again calculates a `` quantization step width '' based on the `` range '' and the predetermined `` quantization bit number '' stored in the memory means, Memory reduction decoding that calculates a pixel value of a pixel belonging to the coding block based on the “reference value” and the “requantized value” stored in a memory unit and the calculated “quantization step width”. And a stage. [Effect] The memory capacity can be reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reproducing apparatus for decoding a stream coded by a coding method using motion compensation prediction, and a coding method for coding an interlace signal in a frame structure. Playback device that decodes the stream that has been decoded, or OSD (On S
The present invention relates to an image reproducing apparatus such as an image reproducing apparatus having a clean display function. In particular, MPE
The present invention relates to an image reproducing device such as a G2 decoder.

[0002]

2. Description of the Related Art A conventional image reproducing apparatus will be described with reference to FIGS. FIG. 4 is a diagram showing a hardware configuration of a conventional image reproducing apparatus. FIG. 5 is a block diagram showing the concept of the device shown in FIG.

In FIG. 4, 1 is an input terminal, 2 is an encoded stream interface means, 3 is a bus means, 4 is a shared memory means, 5 is an encoded stream decoding means, 6 is reproduced image data interface means, and 7 is a reproduced image data interface means. Input terminal, 8
Is an additional image data input interface means, 9 is an additional image data output interface means, 10 is an image synthesizing means, and 11 is an output terminal.

In FIG. 1, an encoded stream A inputted from an input terminal 1 is supplied to an encoded stream interface means 2. The coded stream interface means 2 is connected to the bus means 3 and supplies a coded stream B. Further, the additional image data C input from the input terminal 7 is supplied to the additional image data input interface means 8. This additional image data input interface means 8 is connected to the bus means 3 and supplies additional image data D.

[0005] Further, the encoded stream decoding means 5 comprises:
It is connected to the bus means 3 and exchanges an encoded stream or image data E. Similarly, shared memory means 4
Are connected to the bus means 3 for exchanging image data F.

On the other hand, the reproduced image data interface means 6 is connected to the bus means 3 and is supplied with the reproduced image data G.
Supply. Similarly, the additional image data output interface means 9 is connected to the bus means 3, and the additional image data I
Is supplied and the additional image data J is supplied to the image synthesizing means 10. The output image K supplied from the image synthesizing means 10 is output from an output terminal 11.

FIG. 5 is a block diagram conceptually showing an image reproducing apparatus for making it easier to understand the function of the shared memory means 4 of FIG. In the figure, reference numeral 4a denotes a receiving buffer means, 4b denotes a reproduced image memory means, 4c denotes a frame / field scan converting means, and 4d denotes an additional image memory means.

In FIG. 1, a coded stream A input from an input terminal 1 is supplied to a coded stream decoding means 5 via a reception buffer means 4a. The encoded stream decoding means 5 is connected to the reproduced image memory means 4b, exchanges image data, and supplies image data to the frame / field scan conversion means 4c. Frame / field scan conversion means 4
The output H of c is supplied to the image synthesizing means 10.

The additional image data C input from the input terminal 7 is supplied to the image synthesizing means 10 via the additional image memory means 4d. The output image K supplied from the image synthesizing means 10 is output from the output terminal 11 as described above.

Next, the operation of the conventional image reproducing apparatus will be described.

At present, MPEG2 (ISO / IEC138)
18-2) and other general-purpose video coding methods
This is a method using motion compensation prediction. FIG. 5 shows the concept of an image reproducing apparatus that decodes a stream encoded by an encoding method using this motion compensation prediction method.

In FIG. 5, an encoded stream A inputted from an input terminal 1 is decoded by an encoded stream decoding means 5 after being delayed by a reception buffer means 4a. At this time, in an encoding method using motion compensation prediction, in order to decode the encoded stream,
It has a frame memory for storing a reference image required for motion compensation. This frame memory is the reproduced image memory means 4b in FIG. The reproduced image decoded by the encoded stream decoding means 5 is:
The image data is supplied to the frame / field scan converter 4c and supplied to the image synthesizer 10.

The frame / field scan conversion means 4c has the following role. At present, many videos that are widely used have an interlaced structure. However, video encoding is often performed in a frame structure. That is, at the time of encoding, the interlaced input image is converted into a frame structure, and thereafter, the encoding is performed with the frame structure. At the time of decoding, the encoded stream is decoded, and then converted into the field structure again. Convert,
The output image will be displayed in an interlaced manner. The frame / field scan conversion unit 4c converts the data into a field structure for decoding.

On the other hand, in many current image reproducing apparatuses, O
Equipped with SD function. The OSD function is a function capable of displaying not only one image but also combining and displaying several images, graphics, character information, and the like. From the input terminal 7, this OSD
An additional image for the function is input. The additional image is
After being temporarily stored in the additional image memory unit 4d, it is read out in real time as appropriate and supplied to the image synthesizing unit 10. At this time, FIGS. 4 and 5 show that only one system of additional image is synthesized, but usually, the present invention is not limited to this, and a plurality of systems of additional images may be synthesized.

In FIG. 4, the function of the shared memory means 4 as one shared memory is as follows: the receiving buffer means 4a, the reproduced image memory means 4b, the frame / field scan converting means 4c, and the additional image memory means 4d in FIG.
Corresponds to each function. In order to share the memory, an encoded stream interface unit 2, a reproduced image data interface unit 6, an additional image data input interface unit 8, and an additional image data output interface unit 9 are provided, and each unit is connected by a bus unit 3. Then, the coded stream or the image data is transferred. Generally, the coded stream interface means 2, the reproduced image data interface means 6, the additional image data input interface means 8, and the additional image data output interface means 9 are FIFOs having a relatively small memory capacity.
(First In First Out) memory.

[0016]

In the conventional image reproducing apparatus as described above, the shared memory means 4 includes the receiving buffer means 4a, the reproduced image memory means 4b, the frame / field scan converting means 4c, and the additional image memory means 4d.
Therefore, there is a problem that the total memory capacity of the shared memory means 4 is considerably large.

Further, in the conventional image reproducing apparatus, the additional image is not always synthesized and displayed. When the additional image is not used, the function of the additional image memory means 4d is not involved in the operation, and is not necessarily mounted. There is a problem that it is difficult to say that the used memory is used effectively.

The present invention has been made to solve the above-mentioned problems, and has as its object to provide an image reproducing apparatus capable of reducing the total memory capacity. It is another object of the present invention to provide an image reproducing apparatus capable of effectively utilizing a mounted memory.

[0019]

An image reproducing apparatus according to the present invention divides an image into a plurality of encoded blocks, calculates a parameter of the encoded block, and calculates a pixel belonging to the encoded block based on the parameter. A memory reduction encoding unit for calculating a requantized value of the memory, a memory unit for storing the parameter and the requantized value, and the encoding based on the parameter and the requantized value stored in the memory unit. Memory reduction decoding means for calculating the pixel value of the pixel belonging to the block.

Further, in the image reproducing apparatus according to the present invention, the memory reduction encoding means calculates a “reference value” and a “range” indicating a dynamic range of the encoded block, and calculates the “range” and a predetermined value. A “quantization step width” in the coding block is calculated based on the “quantization bit number”, and “re-quantization” of pixels belonging to the coding block is calculated based on the “reference value” and the “quantization step width”. Value, the memory means stores the "reference value", the "range", and the "requantized value", and the memory reduction decoding means stores the "value" stored in the memory means. The “quantization step width” is calculated again based on the “range” and the predetermined “quantization bit number”, and the “reference value” and the “requantized value” stored in the memory unit are calculated. "Quantization Tsu and calculates the pixel values of the pixels belonging to the encoded block based on the flop width ".

Further, in the image reproducing apparatus according to the present invention, the memory reduction encoding means calculates a “reference value” and a “range” indicating a dynamic range of the encoded block, and calculates the “range” and a predetermined value. A “quantization step width” in the coding block is calculated based on the “quantization bit number”, and “re-quantization” of pixels belonging to the coding block is calculated based on the “reference value” and the “quantization step width”. Value, the memory means stores the 'reference value', the 'quantization step width', and the 'requantized value', and the memory reduction decoding means is stored in the memory means. And calculating a pixel value of a pixel belonging to the coding block based on the "reference value", the "quantization step width", and the "requantized value".

Further, in the image reproducing apparatus according to the present invention, the memory reduction encoding means calculates a “minimum value” and a “maximum value” of the encoded block, and calculates the “minimum value” and the “maximum value”. And a predetermined "quantization bit number" to calculate a "quantization step width" in the coding block,
Based on the “minimum value” and the “quantization step width”, calculate a “requantized value” of the pixel belonging to the coding block, and the memory unit calculates the “minimum value” and the “maximum value”. The “re-quantized value” is stored, and the memory reduction decoding unit stores the “minimum value”, the “maximum value”, and the predetermined “quantized bit number” stored in the memory unit.
The "quantization step width" is calculated again based on the "minimum value" and the "requantization value" stored in the memory means and the calculated "quantization step width". This is to calculate the pixel values of the pixels belonging to the block.

Further, in the image reproducing apparatus according to the present invention, the memory reduction coding means obtains a "maximum value" and a "minimum value" of a pixel in the coded block, and obtains a "maximum value".
Is defined as max and the “minimum value” is defined as “min”.
Is expressed as “level” and “range” is expressed as “range”. Each level is defined as follows: level = min range = max−min + 1 Next, when a predetermined “quantization bit number” is depth, “quantization step When the “width” is expressed as “step”, the quantization step width in the coding block is determined as follows: step = range / (2 ^depth ) (step)
≧ 1) Next, when “pixel value” is set to “value” and “requantized value” is expressed as “index”, the requantized value of the pixel having the pixel value is determined as follows. index = (value−min) / ste
p

Further, in the image reproducing apparatus according to the present invention, when the memory reduction decoding means represents the decoded "pixel value" as decodedvalue, it is determined as follows. decodedvalue = level + st
ep × index + step / 2

Further, in the image reproducing apparatus according to the present invention, the memory reduction decoding means calculates the quantization step width in the same manner as in the encoding procedure.

Further, in the image reproducing apparatus according to the present invention, the memory reduction encoding means obtains a “maximum value” and a “minimum value” of the pixel in the encoded block, and obtains a “maximum value”.
Is defined as max, the “minimum value” is defined as min, and the predetermined “number of quantization bits” is defined as depth, and the “quantization step width” is expressed as “step”. And step = (max−min) / (2 ^depth
-1) (step ≧ 1) Next, when “pixel value” is set to “value” and “requantized value” is expressed as “index”, the requantized value of the pixel having the pixel value is determined as follows. Things. index = ((value-min) + step
/ 2) / step

Further, in the image reproducing apparatus according to the present invention, the memory reduction decoding means calculates the quantization step width in the same manner as in the coding procedure, and decodes the decoded "pixel value".
Is represented as decodedvalue, which is defined as follows. decodedvalue = min + step
× index

Also, in the image reproducing apparatus according to the present invention, the memory means stores a reference image for motion compensation prediction, performs frame / field scan conversion, and stores an additional image.

In the image reproducing apparatus according to the present invention, the coding block has a frame structure when the input image is a non-interlaced material or a still image.

Further, an image reproducing apparatus according to the present invention is
When the input image is an interlaced material or a moving image, the coding block has a field structure.

[0031]

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1 FIG. First Embodiment A first embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a diagram showing a hardware configuration according to the first embodiment of the present invention. FIG.
FIG. 2 is a block diagram showing the concept of the device shown in FIG. In the drawings, the same reference numerals indicate the same or corresponding parts.

In FIG. 1, 4A is a shared memory means,
2 is a memory reduction encoding means, and 13 is a memory reduction decoding means. The other configuration is the same as that of the conventional example, and the description is omitted.

In the figure, the shared memory means 4A comprises a memory reduction encoding means 12 and a memory reduction decoding means 13
Is connected to the bus means 3 through which image data is exchanged.

FIG. 2 shows the shared memory unit 4A, the memory reduction encoding unit 12, and the memory reduction decoding unit 13 of FIG.
FIG. 2 is a block diagram conceptually showing an image reproducing apparatus to make it easy to understand the function of FIG. In the figure, 12b, 12
c, 12d are memory reduction coding means, 13b, 13c,
13d is a memory reduction decoding means. The other configuration is the same as that of the conventional example, and the description is omitted. It should be noted that the reason why there is no memory reduction encoding means and memory reduction decoding means before and after the reception buffer means 4a is that the encoded stream is not further encoded.

In FIG. 2, reproduced image memory means 4b
Is connected to the encoded stream decoding means 5 via the memory reducing encoding means 12b and the memory reducing decoding means 13b, and exchanges image data. Further, the encoded stream decoding unit 5 supplies the image data to the image synthesizing unit 10 via the memory reduction encoding unit 12c, the frame / field scan conversion unit 4c, and the memory reduction decoding unit 13c. On the other hand, the additional image data input from the input terminal 7 is stored in the memory reduction encoding unit 12d.
Additional image memory means 4d, memory reduction decoding means 13d
Is supplied to the image synthesizing means 10 via

Next, the operation of the first embodiment will be described.

Conceptually, the following processing is performed to reduce the memory capacity of the shared memory means 4A. In FIG. 2, when writing the reproduced image data to the reproduced image memory means 4b, the memory reduction encoding means 12b encodes the reproduced image data, reduces the data amount, and then writes the reproduced image data to the reproduced image memory means 4b. Do. When the reproduced image data is read from the reproduced image memory means 4b, the encoded reproduced image data is decoded by the memory reduction decoding means 13b and then supplied to the subsequent stage. Thus, the capacity of the reproduced image memory means 4b is reduced.

Similarly, the reduction of the memory capacity of the frame / field scan conversion unit 4c is realized by the memory reduction encoding unit 12c and the memory reduction decoding unit 13c. Further, the reduction of the memory capacity of the additional image memory unit 4d is realized by the memory reduction encoding unit 12d and the memory reduction decoding unit 13d.

As described above, the reduction of the memory capacity of the shared memory unit 4A is realized by the memory reduction encoding unit 12 and the memory reduction decoding unit 13.

Hereinafter, the memory reduction encoding means 12 and the memory reduction decoding means 1 for performing the operation of reducing the memory capacity will be described.
3 will be described. As described above, currently, general-purpose video encoding methods include motion compensation prediction such as MPEG2 and DCT (D
iscrete Cosine Transform).

In these coding methods, coding is usually performed for each small area. For example, in the above MPEG2, a macroblock corresponds to the small area. Hereinafter, the above-mentioned small area in such a video coding method is referred to as a “macro-coded block”.

By the way, considering the consistency with the video coding method or the simplicity of processing, it is desirable that the memory reduction coding in this embodiment is completed at least in the macro coding block. In consideration of random access to the memory, it is preferable that the amount of encoded data in the macro-encoded block is a fixed amount. With this in mind, three encoding and decoding procedures will be described below.

The first encoding and decoding procedures in the memory reduction encoding means 12 and the memory reduction decoding means 13 are as follows.

Normally, an image signal is quantized by 8 bits (dynamic range: 0 to 255). However, if the image is viewed locally, the quantization bit number of 8 bits is not necessarily required. Therefore, the code amount can be reduced by locally defining the dynamic range and requantizing with a smaller number of quantization bits. Specifically, the following encoding procedure is performed.

The macro coded block is divided into a plurality of small areas. Hereinafter, the small area is referred to as an “encoded block”. The memory reduction encoding is basically performed for each encoded block.

The dynamic range of each encoded block is obtained. That is, the maximum value and the minimum value of the pixel in the coding block are searched, and from this, the reference value (level
l: For example, the minimum value or the maximum value, and a range (range: a width of a quantization value that can include at least the maximum value to the minimum value). For example,
When the maximum value is “max” and the minimum value is “min”, the reference value (level) and the range (range) can be defined by the following equations 1 and 2.

Level = min Equation 1 range = max−min + 1 Equation 2

Based on the range and a set number of quantization bits (depth), a quantization step width (step) in the coding block is calculated. For example,
Equation 3 below can be defined.

Step = range / (2 ^depth ) (step ≧ 1) Equation 3

The data amount can be reduced by setting the quantization bit number (depth) to 3 bits or 4 bits. At this time, the quantization step width (step) is considered to be a real number.

Based on the reference value (level) and the quantization step width (step), a requantization value (index) of a pixel having a pixel value is calculated as in the following equation 4, for example.

Index = (value−min) / step (4)

For the above-mentioned coded block, the above-mentioned reference value
(level), range (range), requantized value (i
ndex) is stored in the shared memory means 4A. At this time,
All of the above are fixed length.

The following decoding procedure is performed. The reference value (level), the range (range), and the requantized value (index) of each pixel in the coding block stored in the shared memory unit 4A are read.

Based on the range and a set number of quantization bits (depth), the quantization step width (step) in the coding block is calculated again. This calculation procedure needs to be the same as the encoding procedure.

The reference value (level) and the quantization step width (st
ep), decoding is performed based on the requantized value (index) of each pixel. For example, a pixel value (decoded value) is decoded by the following Expression 5.

Decodedvalue = level + step × index + step / 2 Expression 5

Next, the second encoding procedure and decoding procedure in the memory reduction encoding means 12 and the memory reduction decoding means 13 will be described. Here, the following encoding procedure is performed.

The macro coded block is divided into a plurality of coded blocks. Hereinafter, each process is performed for each encoded block.

The dynamic range of each encoded block is obtained. That is, the maximum value and the minimum value of the pixel in the coding block are searched, and from this, the reference value (level
l: For example, the minimum value or the maximum value, and a range (range: a width of a quantization value that can include at least the maximum value to the minimum value). For example,
Assuming that the maximum value is “max” and the minimum value is “min”, the reference value (level) and the range (range) can be defined by Expressions 1 and 2.

Based on the above range and a set number of quantization bits (depth), a quantization step width (step) in the coding block is calculated. For example, Equation 3 above can be defined.

The data amount can be reduced by setting the quantization bit number (depth) to 3 bits or 4 bits. At this time, the quantization step width (step)
Is represented by a limited number of bits, not a real number.

Based on the reference value (level) and the quantization step width (step), the requantization value (index) of the pixel having the pixel value is calculated as in the above equation 4, for example.

The reference value for the encoded block
(level), the quantization step width (step), and the requantized value (index) for each pixel are stored in the shared memory unit 4A.
At this time, all of the above are of fixed length.

The following decoding procedure is performed. The reference value (level) and the quantization step width of the coding block stored in the shared memory means 4A
(step), the requantized value (index) of each pixel is read.

The reference value (level) and the quantization step width (st
ep), decoding is performed based on the requantized value (index) of each pixel. For example, the pixel value (decoded value) is decoded by Equation 5 above.

As is clear from the above-mentioned decoding procedure, the second procedure has a feature that the amount of calculation in decoding is smaller than that of the first procedure.

Next, the third encoding procedure and decoding procedure in the memory reduction encoding means 12 and the memory reduction decoding means 13 will be described. Here, the following encoding procedure is performed.

The macro coded block is divided into a plurality of coded blocks. Memory reduction coding is basically
Processing is performed for each encoded block.

The dynamic range of each encoded block is obtained. That is, the minimum value (min) and the maximum value (max) of the pixel in the coding block are searched.

The minimum value (min) and the maximum value (max)
A quantization step width (step) in the coding block is calculated based on a certain set number of quantization bits (depth). For example, the following equation 6 can be defined.

Step = (max−min) / (2 ^depth −1) (step ≧ 1) Expression 6

The data amount can be reduced by setting the number of quantization bits (depth) to 3 bits, 4 bits, or the like. At this time, the quantization step width (step) is considered to be a real number.

The minimum value (min) and the quantization step width
Based on (step), the requantized value of the pixel having the pixel value value
(index) is calculated, for example, as in the following Expression 7.

Index = ((value−min) + step / 2) / step Equation 7

For the coded block, the minimum value
(min), maximum value (max), requantized value (inde
x) is stored in the shared memory means 4A. At this time, all of the above are of fixed length.

The following decoding procedure is performed. The minimum value (min), the maximum value (max), and the requantized value (index) of each pixel in the encoded block stored in the shared memory unit 4A are read.

The minimum value (min) and the maximum value (max)
Based on a certain set number of quantization bits (depth), the quantization step width (step) in the coding block is calculated again. This calculation procedure needs to be the same as the encoding procedure.

The minimum value (min) and the quantization step width (ste
p), decoding is performed based on the requantized value (index) of each pixel. For example, a pixel value (decoded value) is decoded as follows.

Decodedvalue = min + step × index Expression 8

As is clear from the encoding and decoding procedures, the third procedure differs from the first and second procedures in that the minimum value (min) and the maximum value (max) are different. ,
The feature is that it is reproduced accurately. This feature shows good encoding characteristics for binary images such as character images. However, this feature is also possible in the first and second procedures. By the above procedures, the memory capacity of the shared memory unit 4A can be reduced by about 20% to 80% as compared with the conventional shared memory unit 4.

By the way, in the first, second, and third encoding and decoding procedures, the smaller the range, the smaller the image quality degradation due to the memory reduction encoding. Therefore, it is desirable to use a coded block in which the quantization value does not vary. In other words, the way in which the coded blocks are configured affects the image quality.

When the macro coded block has a frame structure, the coded block can be formed with either a frame structure or a field structure. This is shown in FIG. FIG. 3A shows a state in which the coding block has a frame structure, and FIG. 3B shows a state in which the coding block has a field structure.

In general, when the input image is a non-interlaced material, the coding block is constituted by a frame structure,
In the case of interlaced material, when the coding block has a field structure, variation in the quantization value in the coding block is smaller. That is, image quality degradation is small.
At this time, it should be noted that non-interlaced material may be interlaced or displayed.

Further, in general, it is better to configure a coding block with a frame structure in the case of a still picture, and to configure a coding block in a field structure in the case of a moving picture, because of a variation in quantization value in the coding block. Less is.

That is, as the structure of the coding block, it is desirable to use the frame structure and the field structure adaptively. For example, the above adaptive switching:
It can be performed for each program or for each sequence, and can also be performed for each macro coded block. However, in the case of performing for each macro coded block, for each of the macro coded blocks, it is appropriately determined whether to configure the coded block with the frame structure or the coded block with the field structure. Index indicating whether the encoding has been performed (hereinafter referred to as “encoded block structure index”)
Must be additionally stored in the shared memory means 4A.

By the way, when the memory reduction encoding means 12 and the memory reduction decoding means 13 are applied to an actual image reproducing apparatus, the above-mentioned encoding and decoding procedures are used as described below, so that the on-board memory (shared memory) is used. Memory means 4
A) can be effectively used.

Normally, an image reproducing apparatus is equipped with a memory having a predetermined capacity. At this time, for example, the OS
When the D function is not used, deterioration of the image quality of the reproduced image can be suppressed by increasing the capacity of the memory used for decoding the original reproduced image.

Alternatively, when displaying a plurality of additional images, optimal display of a plurality of images becomes possible by allocating an appropriate memory capacity to each image in consideration of the image quality required for each image. These can be realized by appropriately changing the number of quantization bits (depth) or the size of the coded block in the memory reduction coding means 12 and the memory reduction decoding means 13 described above.

In the first embodiment, the luminance signal,
Although the format of an image signal such as a color difference signal or an RGB signal has not been particularly described, the first embodiment can be applied to any case regardless of the format of the image signal.

Further, the first embodiment does not specify the calculation accuracy in the encoding procedure and the decoding procedure in the memory reduction encoding means 12 and the memory reduction decoding means 13. At this time, if the calculation accuracy is increased, the calculation amount is increased and the memory reduction amount is reduced.
It is possible to suppress the image quality deterioration depending on the memory reduction coding / decoding. Further, if the calculation accuracy is reduced, the image quality deterioration is increased, but the calculation amount can be reduced and the memory reduction amount can be increased.

[0092]

As described above, the image reproducing apparatus according to the present invention divides an image into a plurality of encoded blocks,
A memory reduction encoding unit that calculates a parameter of the encoded block, and calculates a requantized value of a pixel belonging to the encoded block based on the parameter; and a memory unit that stores the parameter and the requantized value. And memory reduction decoding means for calculating a pixel value of a pixel belonging to the coding block based on the parameters stored in the memory means and the requantized value, so that a memory capacity can be reduced. This has the effect.

Further, in the image reproducing apparatus according to the present invention, as described above, the memory reduction encoding means sets the "reference value" indicating the dynamic range of the encoded block.
And a “range”, and calculate a “quantization step width” in the coding block based on the “range” and a predetermined “quantization bit number”, and calculate the “reference value” and the “quantization step width”. , A "requantized value" of a pixel belonging to the coding block is calculated, and the memory means stores the "reference value", the "range", and the "requantized value", and The reduction decoding unit calculates again the “quantization step width” based on the “range” stored in the memory unit and the predetermined “quantization bit number”, and calculates the “reference” stored in the memory unit. Since the pixel values of the pixels belonging to the coding block are calculated based on the “value” and the “requantized value” and the calculated “quantization step width”, the effect that the memory capacity can be reduced is obtained. Play.

Further, in the image reproducing apparatus according to the present invention, as described above, the memory reduction encoding means sets the "reference value" indicating the dynamic range of the encoded block.
And a “range”, and calculate a “quantization step width” in the coding block based on the “range” and a predetermined “quantization bit number”, and calculate the “reference value” and the “quantization step width”. , And calculates the “requantized value” of the pixel belonging to the coding block, and the memory unit stores the “reference value”, the “quantization step width”, and the “requantized value”. The memory reduction decoding unit calculates a pixel value of a pixel belonging to the coding block based on the “reference value”, the “quantization step width”, and the “requantized value” stored in the memory unit. Since the calculation is performed, there is an effect that the memory capacity can be reduced.

Further, in the image reproducing apparatus according to the present invention, as described above, the memory reduction encoding means calculates the “minimum value” and the “maximum value” of the encoded block, and calculates the “minimum value”. Calculate a “quantization step width” in the coding block based on the “maximum value” and a predetermined “quantization bit number”, and perform the encoding based on the “minimum value” and the “quantization step width”. The memory means calculates the “requantized value” of the pixel belonging to the block, and the memory means stores the “minimum value”, the “maximum value”, and the “requantized value”, and the memory reduction decoding means The "quantization step width" based on the "minimum value", the "maximum value" and the predetermined "quantization bit number" stored in the memory means.
Is calculated again, and a pixel value of a pixel belonging to the coding block is calculated based on the “minimum value” and the “requantized value” stored in the memory unit and the calculated “quantization step width”. Therefore, there is an effect that the memory capacity can be reduced.

Further, in the image reproducing apparatus according to the present invention, as described above, the memory reduction encoding means obtains the “maximum value” and “minimum value” of the pixel in the encoded block, and obtains the “maximum value”. Is max and “minimum value” is min
, "Reference value" is level, and "range" is r
When expressed as “ange”, each is defined as follows, and level = min range = max−min + 1 Next, when a predetermined “quantization bit number” is set to depth, the “quantization step width” is expressed as “step”. , The quantization step width in the coding block is determined as follows: step = range / (2 ^depth ) (step
≧ 1) Next, when “pixel value” is set to “value” and “requantized value” is expressed as “index”, the requantized value of the pixel having the pixel value is determined as follows: index = ( value-min) / ste
p This has the effect of suppressing image quality degradation due to memory capacity reduction.

Further, in the image reproducing apparatus according to the present invention, as described above, when the memory reduction decoding means expresses the decoded "pixel value" as decodedvalue, it is determined as follows. decodedvalue = level + st
ep × index + step / 2 An effect is obtained that image quality deterioration due to a reduction in memory capacity can be suppressed.

Further, in the image reproducing apparatus according to the present invention, as described above, since the memory reduction decoding means calculates the quantization step width in the same manner as the encoding procedure, the image quality accompanying the memory capacity reduction is obtained. This has the effect that deterioration can be suppressed.

Further, in the image reproducing apparatus according to the present invention, as described above, the memory reduction coding means obtains the “maximum value” and “minimum value” of the pixel in the coded block, and obtains the “maximum value”. Is max and “minimum value” is mi
n, when the predetermined “number of quantization bits” is depth, and when the “quantization step width” is expressed as “step”, the quantization step width in the coding block is determined as follows: step = (max−min) ) / (2 ^depth
-1) (step ≧ 1) Next, when “pixel value” is set to “value” and “requantized value” is expressed as “index”, the requantized value of the pixel having the pixel value is determined as follows. Therefore, index = ((value-min) + step
/ 2) / step It is possible to suppress the image quality deterioration due to the reduction of the memory capacity.

Further, as described above, in the image reproducing apparatus according to the present invention, the memory reduction decoding means calculates the quantization step width in the same manner as in the encoding procedure, and outputs the decoded “pixel value”. Is represented as decodedvalue, which is defined as follows: decodedvalue = min + step
× index This has the effect of suppressing image quality degradation due to memory capacity reduction.

In the image reproducing apparatus according to the present invention, as described above, the memory means stores the reference image for motion compensation prediction, performs frame / field scan conversion, and stores the additional image. There is an effect that the memory capacity can be reduced.

As described above, in the image reproducing apparatus according to the present invention, the coding block has a frame structure when the input image is a non-interlaced material or a still image. This has the effect of preventing image quality degradation due to memory capacity reduction.

Further, the image reproducing apparatus according to the present invention
As described above, since the coding block has a field structure when an input image is an interlaced material or a moving image, it is possible to suppress an image quality deterioration due to a reduction in memory capacity. Play.

[Brief description of the drawings]

FIG. 1 is a diagram showing a hardware configuration according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a conceptual configuration of the first embodiment of the present invention.

FIG. 3 is a diagram illustrating a configuration of an encoded block in a memory reduction encoding unit and a memory reduction decoding unit according to the first embodiment of the present invention.

FIG. 4 is a diagram illustrating a hardware configuration of a conventional image reproducing apparatus.

FIG. 5 is a block diagram illustrating a conceptual configuration of a conventional image reproducing apparatus.

[Explanation of symbols]

1 input terminal, 2 coded stream interface means, 3 bus means, 4A shared memory means, 5 coded stream decoding means, 6 reproduced image data interface means, 7 input terminal, 8 additional image data input interface means, 9 additional image Data output interface means, 10 image synthesizing means, 11 output terminals, 12
Memory reduction encoding means, 13 Memory reduction decoding means.

Claims (12)

[Claims] 1. Dividing an image into a plurality of coding blocks,
A memory reduction encoding unit that calculates a parameter of the coding block and calculates a requantization value of a pixel belonging to the coding block based on the parameter; and a memory unit that stores the parameter and the requantization value. And a memory reduction decoding means for calculating a pixel value of a pixel belonging to the coding block based on the parameter and the requantized value stored in the memory means. 2. The memory reduction encoding unit calculates a reference value indicating a dynamic range of the encoded block and a range, and calculates a quantization step width in the encoded block based on the range and a predetermined number of quantization bits. And calculates a re-quantized value of a pixel belonging to the coding block based on the reference value and the quantization step width. The memory unit stores the reference value, the range, and the re-quantized value. The memory reduction decoding unit calculates the quantization step width again based on the range and the predetermined number of quantization bits stored in the memory unit, and calculates the reference value and the reference value stored in the memory unit. The image according to claim 1, wherein a pixel value of a pixel belonging to the coding block is calculated based on a requantization value and the calculated quantization step width. Raw devices. 3. The memory reduction coding unit calculates a reference value and a range indicating a dynamic range of the coding block, and calculates a quantization step width in the coding block based on the range and a predetermined number of quantization bits. And calculating a re-quantized value of a pixel belonging to the coding block based on the reference value and the quantization step width. The memory means stores the reference value, the quantization step width, and the re-quantization value. A memory value, and the memory reduction decoding unit calculates a pixel value of a pixel belonging to the coding block based on the reference value, the quantization step width, and the requantized value stored in the memory unit. The image reproducing apparatus according to claim 1, wherein: 4. The memory reduction encoding means calculates a minimum value and a maximum value of the encoded block, and performs quantization on the encoded block based on the minimum value, the maximum value, and a predetermined number of quantization bits. Calculating a step width; calculating a requantization value of a pixel belonging to the coding block based on the minimum value and the quantization step width; and the memory means calculates the minimum value, the maximum value, and the requantization value. Storing the value, the memory reduction decoding means recalculates the quantization step width based on the minimum value, the maximum value and the predetermined number of quantization bits stored in the memory means, And calculating a pixel value of a pixel belonging to the coding block based on the minimum value and the re-quantized value stored in the memory and the calculated quantization step width. Reproducing apparatus. 5. The memory reduction coding means calculates a maximum value and a minimum value of a pixel in the coding block, and sets a reference value to 1 when a maximum value is max and a minimum value is min.
When the level and the range are expressed as “range”, they are respectively defined as follows: level = min range = max−min + 1 Next, when the predetermined number of quantization bits is depth,
When the quantization step width is expressed as step, the quantization step width in the coding block is determined as follows: step = range / (2 ^depth ) (step)
≧ 1) Next, when the pixel value is value, the requantized value is i
When expressed as nindex, the requantized value of the pixel having the pixel value is determined as follows: index = (value−min) / ste
The image reproduction device according to claim 2 or 3, wherein p is an integer. 6. When the decoded pixel value is represented as decodedvalue, the memory reduction decoding means determines it as follows: decodedvalue = level + st
6. The image reproducing apparatus according to claim 5, wherein ep * index + step / 2. 7. The image reproducing apparatus according to claim 6, wherein said memory reduction decoding means calculates said quantization step width in a manner similar to said encoding procedure. 8. The memory reduction encoding means calculates a maximum value and a minimum value of a pixel in the encoding block, and when the maximum value is max, the minimum value is min, and a predetermined number of quantization bits is depth, Set the quantization step width to step
, The quantization step width in the above-mentioned coding block is determined as follows, and step = (max−min) / (2 ^depth
-1) (step ≧ 1) Next, assuming that the pixel value is value, the requantized value is i
When expressed as nindex, the requantized value of the pixel having the pixel value is determined as follows: index = ((value-min) + step
The image reproducing apparatus according to claim 4, wherein / 2) / step. 9. The memory reduction decoding means calculates the quantization step width in the same manner as in the encoding procedure, and when the decoded pixel value is represented as decodedvalue, it is determined as follows: decodedvalue = min + step
The image reproducing apparatus according to claim 8, wherein: x index. 10. The apparatus according to claim 1, wherein said memory means stores a reference image for motion compensation prediction, performs frame / field scan conversion, and stores an additional image. An image reproducing device according to claim 1. 11. The coding block according to claim 1, wherein the coding block has a frame structure when the input image is a non-interlaced material or a still image. An image reproducing device according to claim 1. 12. The coding block according to claim 1, wherein the coding block has a field structure when the input image is an interlaced material or a moving image. The image reproducing apparatus according to any one of the preceding claims.