JP2833235B2 - Orthogonal transform coding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an orthogonal transform coding apparatus used for increasing a compression ratio when a video signal is coded with high efficiency.

[0002]

2. Description of the Related Art A band compression technique (or a high-efficiency code) for reducing the data amount of an original video signal for a long time recording is used to configure a video recording / reproducing apparatus (VTR, video disk, etc.) for recording / reproducing a video signal. Technology) is used.
One of the band compression techniques is orthogonal transform coding. Orthogonal transform coding is a technique in which a video signal is divided into blocks, and frequency decomposition is performed on a block basis to reduce the amount of data to be transmitted (recorded) as the high-frequency component is degraded. This makes use of the visual characteristic that it is difficult to detect the image.

FIG. 7 shows a block diagram of a conventional orthogonal transform coding apparatus for performing the above orthogonal transform coding.
The operation will be described.

In FIG. 7, reference numeral 1 denotes an input terminal for inputting a block signal of a video signal according to the conventional example, 2 denotes an orthogonal transform circuit for orthogonally transforming the block signal, and 3 denotes an orthogonal transform circuit obtained from the orthogonal transform circuit 2. To encode the orthogonal transform coefficients,
A reordering circuit 4 for reordering the coefficient arrangement, an encoding circuit 4 for encoding the orthogonal transform coefficients output from the reordering circuit 3, and an output terminal 5 of the orthogonal transform device according to the conventional example, which outputs encoded data. . Further, the encoding circuit 4
Is a quantizer 41 for quantizing the orthogonal transform coefficients by a certain step width, a quantization selection circuit for selecting a quantizer having a step width suitable for keeping a desired data amount, and the quantizer 41 And an encoder 43 for encoding the output quantized data.

The encoder 43 employs variable length coding in which a code word having a higher occurrence frequency is assigned a shorter code word (small data amount) with respect to the occurrence frequency of an encoded code word. Therefore, the quantization selection circuit 42 selects a quantizer having an optimal step width by calculating the data amount as a result of the variable length coding.

FIG. 8 shows the arrangement of coefficients in the block A at the output A of the orthogonal transformation circuit 2 and the output B at the rearrangement circuit 3 for explaining the operation of the rearrangement circuit 3. In the coefficient arrangement shown in the figure, the orthogonal transform coding apparatus of the conventional example adopts a technique of two-dimensional orthogonal transform in the horizontal direction and the vertical direction, and the block has a block size of 4 horizontal pixels and 4 vertical pixels. Things. Therefore (FIG. 8) In the signal A, 16 coefficients of 4 coefficients in the horizontal direction and 4 coefficients in the vertical direction are used.
One block is composed of coefficients. In the block in the figure,
The frequency component represented by each coefficient corresponds to a horizontal low band as leftward, and corresponds to a vertical low band as upward. The configuration of the orthogonal transformation circuit 2 that performs two-dimensional orthogonal transformation is as follows.
Since the orthogonal transformation in one direction is performed after the orthogonal transformation in one direction is performed in the horizontal direction and the vertical direction, the coefficients are arranged like the signal A.

However, in order to perform encoding for the two-dimensional orthogonal transform, the coefficient of the signal B and a two-dimensional low frequency band called zigzag scan as shown (both horizontal and vertical low bands and upper left in the figure) It is suitable to line up from to high frequencies (lower right in the figure). In other words, the lower frequency component including the DC component has a greater effect on vision, and the lower frequency component is a more important component. Therefore, in the encoding circuit 4, it is only necessary to reduce the step width sequentially from the first modulus of the coefficient arrangement until the data amount reaches a desired data amount, starting from a state where quantization is performed with the coarsest step width.

[0008]

However, the above configuration has the following problems.

In terms of the amount of data obtained by encoding, some blocks having a large data amount include blocks containing large-amplitude frequency components, and some blocks having a small data amount include blocks containing only small-amplitude frequency components. When a similar step width is assigned to these two types of blocks in order to reduce the data amount, most of the latter blocks containing only small-amplitude frequency components are deleted (coefficients are assigned to 0). Although this is a block with a flat screen, important information is lost and block noise occurs (for example, noticeable in the background vegetation and sky). On the other hand, the former block including the large-amplitude frequency component is not deleted by the step width because of the amplitude of the frequency component, and thus the influence of the quantization error is visually smaller than the latter block.

However, since the optimal quantization performed in the coding circuit 4 of the conventional orthogonal transform coding apparatus is performed for each block, it is difficult to select a quantizer between blocks (data amount control). Therefore, the block including only the small-amplitude frequency component is subjected to the quantization selection operation similarly to the block including the large-amplitude frequency component, thereby deteriorating the image quality of a flat portion such as the background. .

The present invention has been made in view of the above-mentioned problems of the prior art, and has as its object to provide an orthogonal transform coding apparatus which can easily control the amount of data over blocks and has a very small increase in circuit scale.

[0012]

SUMMARY OF THE INVENTION The present invention provides a coefficient for encoding orthogonally transformed transform coefficients and a reordering means for reordering, and a maximum for detecting a maximum value of the transform coefficients in parallel with the reordering means. A maximum value detecting means for detecting the maximum value of the transform coefficient in parallel with the value detecting means or the rearranging means, and a maximum value obtained by the maximum value detecting means is subjected to a predetermined classification according to a plurality of threshold values. A classifying means for classifying the transform coefficients into predetermined classes in accordance with a plurality of threshold values in parallel with the classifying means or the reordering means; and detecting a class corresponding to a maximum transform coefficient of the class obtained by the classifying means. The orthogonal transform coding apparatus is provided with a maximum class detecting means for performing the orthogonal transform coding.

[0013]

With the above arrangement, the present invention obtains the maximum value of the transform coefficient corresponding to the maximum amplitude of the frequency components included in each block itself or the result of classifying it. The amplitude of the frequency component can be easily understood, and even if the amplitude is small, the data amount can be controlled without deleting important frequency components. Further, since the maximum value or the class of the frequency component is obtained in parallel with the reordering circuit, no new memory is required for detecting the maximum value or the class.

[0014]

FIG. 1 is a block diagram of an orthogonal transform coding apparatus according to a first embodiment of the present invention. Reference numeral 1 in FIG. 1 denotes an input terminal of this embodiment, which has a block size of 4 pixels horizontally and 4 pixels vertically per block as in the conventional example. Numeral 2 denotes an orthogonal transformation circuit which performs two-dimensional orthogonal transformation on the input block of 4 × 4 pixels, and numeral 3 designates an orthogonal transformation circuit 2 which performs two-dimensional frequency quantization and encoding in a 4-encoding circuit.
This is a rearrangement circuit that rearranges output conversion coefficients so as to be arranged from a low band to a high band in a two-dimensional frequency. Reference numeral 5 denotes an output terminal for outputting the orthogonal transform coefficient encoded by the encoding circuit 4. Here, reference numeral 60 denotes a maximum value detection circuit for calculating the maximum value of the coefficient for each block with respect to the conversion coefficient (A) output from the orthogonal transformation circuit 2.

The operation of the above configuration will be described with reference to FIG. 2 (FIG. 2), using a change diagram of the amplitude value of the transform coefficient in the block.

(A) of FIG. 2 is a block composed of only those having relatively small amplitudes of frequency components included in the block, and (b) is a conversion of a block including frequency components having relatively large amplitude. The change in the amplitude of the coefficient is shown. In the figure, A indicates a change in the output of the orthogonal transformation circuit 2, and C indicates a change in the output of the maximum value detection circuit 60. In each of (a) and (b), the horizontal axis indicates the order of the coefficients, which corresponds to the order A of the conversion coefficients in the conventional example (FIG. 8). The vertical axis represents the magnitude of the amplitude and the absolute value of the conversion coefficient.

The maximum value detection circuit 60 detects the absolute values of the coefficients in the order of the coefficients of the block signal A, but the first conversion coefficient of one block is a DC component and is excluded from the detection of the maximum value.
Therefore, on the horizontal axis 1, the maximum value candidate in the maximum value detection circuit 60 is set to the initial value 0. Thereafter, the maximum value candidate is updated by comparing the absolute value of the conversion coefficient with the maximum value candidate. When the comparison with the last transform coefficient (in the case of the figure, the horizontal axis 16) is completed, the block is output as the maximum value of the transform coefficient of the block.

According to this embodiment as described above, the maximum value of the transform coefficient (excluding the DC component) for each block obtained by the maximum value detection circuit 60 is determined by the quantization selection circuit in the encoding circuit 4 at the subsequent stage. 44. In this way, the quantization selection circuit 44 determines, based on the maximum value of the transform coefficient, the transform coefficient of a block containing only a small-amplitude frequency component as compared to a block containing a large-amplitude frequency component. Low rate of reduction
As a result, it is possible to prevent the image quality from being deteriorated in the flat portion.

Although the maximum value detection circuit 60 can be configured in the quantization selection circuit 44, all the conversion coefficients for one block must be compared and inspected when the maximum value is detected. For that purpose, a delay circuit or memory for at least one clock is newly required. However, the configuration of the present embodiment is effective in terms of circuit scale because it is executed in parallel with the rearrangement circuit 3 and does not involve an increase in delay circuits or memories.

In the maximum value detection circuit 60, the DC component is excluded from the target of the maximum value detection because the DC component is an offset of the entire AC component block composed of other frequency components, which is a problem. This is because, unlike the amplitude value of the frequency component, it does not relate to the deterioration of the image quality of the flat portion in question.

FIG. 3 is a block diagram of an orthogonal transform apparatus according to a second embodiment of the present invention, in which 70 components are classified immediately after the maximum value detection circuit 60 with respect to the configuration of the first embodiment. A circuit is provided.

The operation of the present embodiment will be described with reference to FIG. The change A of the transform coefficient of the output of the orthogonal transform circuit 2 in FIG. 4 and the change C of the maximum value candidate by the maximum value detection circuit 60 are the same as in the first embodiment. Now, a class to be classified by the classifying circuit 70, using the three threshold P _1, P _2, P ₃ as shown, are prepared four types of classes. In the figure, these four types of classes are assigned 0,
They are numbered 1, 2, and 3. In the case of the change A of the orthogonal transform coefficient shown in FIG. 4, since the maximum value is larger than the threshold value P ₃ , it is detected by the classifying circuit 70 as the class 3, and the class is quantized in the encoding circuit 4 at the subsequent stage. The selecting circuit 45 is used to select quantizers having different step widths.

In the first embodiment, the maximum value of the orthogonal transform coefficient is sent to the encoding circuit 4 as it is, but the extent to which the amplitude of the frequency component of interest is small is in a fairly limited range. In other words, even if the step width is the same between blocks where the maximum value of the transform coefficient is large to a certain extent, the effect is small. Therefore, the quantization selection function in the quantization selection circuit 45 does not require the number of bits capable of expressing the entire range of the orthogonal transform coefficients.

As described above, in the present embodiment, the effect of solving the above-described problem is the same as that of the first embodiment, and the maximum value of the transform coefficient to be sent to the encoding circuit 4 is further classified. Since the class may be transmitted, the number of transmission bits may be small, which is more effective.

Next, a third embodiment of the present invention will be described. (FIG. 5) is a block diagram of the orthogonal transform coding apparatus according to the present embodiment, and (FIG. 6) is a change diagram of transform coefficients.

In FIG. 3, an input terminal 1, an orthogonal transformation circuit 2, a reordering circuit 3, an encoding circuit 4, and an output terminal 5 have the same configuration as that of the second embodiment (FIG. 3). There are newly provided 71 classification circuits and 61 maximum class detection circuits.

The classifying circuit 71 receives the orthogonal transform coefficient A (FIG. 6) as an input, and converts the transform coefficients having small absolute values to 0, 1, 2, 3 according to the magnitude of the absolute value of each orthogonal transform coefficient.
Are classified into four types. The classification method is the same as that of the second embodiment in the same three types of thresholds (P ₁ , P ₂ , P ₃ )
Is used. The result is shown in E of FIG. 6.

Next, the maximum class detecting circuit 61 detects a class corresponding to the maximum conversion coefficient based on the class E. At this time, if a class for a large transform coefficient is assigned to a large number as in the present embodiment, it is only necessary to find the maximum value of the number of the assignment. This can be easily realized with the same function as the maximum value detection circuit 60 of the first and second embodiments. Moreover, as compared with the maximum value detection circuit 60, the range that the input class can take (class 0 in this embodiment)
３3 for 2 bits), and the possible range of the orthogonal transform coefficient (generally, 8 bits or more are required). Therefore, a comparator for detecting the maximum class candidate and a register for storing the maximum class candidate itself are also provided. It is small and very effective.

As described above, according to this embodiment, the effect of solving the above-mentioned problem is the same as that of the first and second embodiments, which is effective. In addition, the conversion coefficients are classified. Since the maximum class detection is performed from, the circuit scale is small and it is very practical.

In the embodiment of the present invention, the block signal to be orthogonally transformed has a block size of 4 × 4 pixels, but may have a block size of 8 × 8 pixels or 16 × 16 pixels. Further, a three-dimensional orthogonal transform may be used instead of the horizontal and vertical two-dimensional orthogonal transform. As described above, as the block size or the number of dimensions increases and the number of transform coefficients forming a block increases, the detection of the maximum value or the maximum class of the transform coefficients is performed by the reordering circuit 3 as in this embodiment. Executing in parallel is more effective because a new delay device or the like is not required for the detection.

[0031]

As described above, according to the present invention,
Since the value of the maximum value of the transform coefficient corresponding to the maximum amplitude of the frequency components included in each block itself or the result of classifying is obtained, the amplitude of the frequency component for each block can be easily understood at the time of the subsequent encoding, Data volume can be controlled without deleting important frequency components even with small amplitude,
This is very effective because orthogonal transform coding without image quality degradation can be performed. Further, since the maximum value or the class of the frequency component is obtained in parallel with the reordering circuit, no new memory is required for detecting the maximum value or the class, which is advantageous in terms of circuit size. But its practical effect is great.

[Brief description of the drawings]

FIG. 1 is a block diagram of an orthogonal transform encoding device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a change in an absolute value of a conversion coefficient for explaining the first embodiment.

FIG. 3 is a block diagram of an orthogonal transform coding apparatus according to a second embodiment of the present invention.

FIG. 4 is a diagram illustrating a change in an absolute value of a conversion coefficient for explaining the second embodiment.

FIG. 5 is a block diagram of an orthogonal transform coding apparatus according to a third embodiment of the present invention.

FIG. 6 is a diagram illustrating a change in the absolute value of a conversion coefficient for explaining the third embodiment.

FIG. 7 is a block diagram of a conventional orthogonal transform encoding device.

FIG. 8 is an arrangement diagram of block coefficients for explaining the operation of the conventional example.

[Explanation of symbols]

 3 Reordering circuit 60 Maximum value detection circuit 61 Maximum class detection circuit 70, 71 Classification circuit

Continuation of front page (72) Inventor Iwao Hidaka 1006 Kazuma Kadoma, Osaka Prefecture Inside Matsushita Electric Industrial Co., Ltd. (56) References JP-A-3-42984 (JP, A) JP-A-4-178088 ( JP, A) (58) Field surveyed (Int. Cl. ⁶ , DB name) H04N 7/24-7/68 H04N 1/41-1/419

Claims (3)

(57) [Claims] 1. An apparatus for orthogonally transforming and encoding a video signal , comprising: means for orthogonally transforming the video signal; and encoding the orthogonally transformed transform coefficient.
Rearranging means for rearranging the transform coefficient array ; maximum value detecting means for detecting a maximum value of the orthogonally transformed transform coefficients; and, according to the maximum value, a larger value as the maximum value increases.
Quantize to the specified data amount by quantizing with the step width.
Quantization selecting means for selecting a quantizer to be converted , and the quantization selection means
Quantizing means for quantizing with a quantizer selected by the stage
And an encoding means for encoding the output of the quantization means . 2. An apparatus for orthogonally transforming and encoding a video signal , comprising: means for orthogonally transforming the video signal; and encoding the orthogonally transformed transform coefficient.
Rearranging means for rearranging the transform coefficient sequence; maximum value detecting means for detecting a maximum value of the orthogonally transformed transform coefficients; and a maximum value obtained by the maximum value detecting means in accordance with a plurality of thresholds. Classifying means for classifying, and the classifying means according to the class information obtained from the classifying means.
The larger the class value, the larger the step size
Quantization to quantize to a predetermined data amount
Means for selecting a quantizer, and the transform coefficient obtained by the reordering means,
Quantizing means for quantizing with a quantizer selected by the stage
And an encoding means for encoding the output of the quantization means . 3. An apparatus for orthogonally transforming and encoding a video signal , comprising: means for orthogonally transforming the video signal; and encoding the orthogonally transformed transform coefficient.
And conversion obtain permutation means arranged transform coefficient sequence, and classification means for performing predetermined classification according to the transform coefficients into a plurality of threshold values, detects the class of the maximum conversion coefficient of the resulting class by the classification means Maximum class detecting means, and a maximum class obtained by the maximum class detecting means.
The largest class that indicates a large conversion coefficient
Quantize to the specified data amount by quantizing with the step width.
Quantization selecting means for selecting a quantizer to be converted , and the quantization selection means
Quantizing means for quantizing with a quantizer selected by the stage
And an encoding means for encoding the output of the quantization means .