JP2861328B2 - High efficiency coding device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-efficiency encoding device necessary for transmitting an image signal or the like at a low rate or recording for a long time.

2. Description of the Related Art In order to perform a plurality of orthogonal transforms using a conventional high-efficiency code, each orthogonal transformer requires an encoder. FIG. 3 shows a block diagram of a conventional high efficiency coding apparatus. In the figure, 1 is an input signal, 2a and 2b are orthogonal transformers, 3a and 3b are orthogonal transform components, 4a and 4b are encoders, 5a and 5b are encoded signals, 6 is a selector, 7 is control information, 8 is a coded signal.

The operation of the conventional high-efficiency coding apparatus configured as described above will be described below. The input signal 1 is orthogonally transformed by each of the orthogonal transformers 2a and 2b, and orthogonal transform components 3a and 3b are generated. Next, the orthogonal transform components 3a and 3b are respectively encoded by encoders 4a and 4b, and one of them is selected by a selector 6, and an encoded signal 8
Becomes This is performed by the selection control information 7 of the selector 6,
The control information 7 is provided from outside. For example, when the orthogonal transformer 2a is an intra-field orthogonal transform and the orthogonal transformer 2b is an inter-field orthogonal transform, the coded signal 5a is selected for a moving image block, and the coded signal 5b is selected for a still block. The control information 7 is determined.

However, in the above-described configuration, a plurality of encoders are required, and the hardware scale of the encoder is large, so that the hardware scale of the high-efficiency encoding device is considerably large. Become.

In view of the foregoing, an object of the present invention is to provide a high-efficiency coding apparatus that has almost the same coding distortion as the conventional example without substantially increasing the hardware scale.

Means for Solving the Problems A high-efficiency encoding apparatus according to the present invention comprises: a block of a digital image input signal as an input signal; and at least two orthogonal transformers for performing different orthogonal transforms on the input signal; A primary converter that performs only the permutation of the sum, difference, and data position with respect to the output component of the transformer; An encoder for selecting and encoding by one kind of encoding method is provided.

Operation According to the present invention, the output of each primary transformer can be converted into a random variable having a certain distribution by performing a linear transformation on the orthogonal transformation components according to a rule that is a primary transformation. Therefore, even if the same encoder is used for different orthogonal transforms, the performance of the encoder is slightly reduced, and the encoding distortion increases even when compared with an encoder optimized for each orthogonal transform. Is slight.

Embodiment FIG. 1 is a block diagram of a high efficiency coding apparatus according to an embodiment of the present invention. In the figure, 1 is an input signal, 2a, 2b
Is an orthogonal transformer, 3a and 3b are orthogonal transform components, 10a and 10b are primary converters, 11a and 11b orthogonal transform components, 12 is a selector, 13 is control information, 14 is an orthogonal transform component, 4 is an encoder, 8 is a coded signal.

The operation of the high-efficiency coding apparatus according to the present embodiment configured as described above will be described below. The input signal 1 is orthogonally transformed by each of the orthogonal transformers 2a and 2b, and orthogonal transform components 3a and 3b are generated. Next, the orthogonal transform components 2a and 2b are subjected to the primary transform of the orthogonal transform components by primary converters 10a and 10b, respectively, and one of the orthogonal transform components 11a and 11b is selected by the selector 12, and the orthogonal transform components 14 and Become. The control information 13 is the control information 7 of the embodiment of FIG.
And switches the selector 12 provided externally. The orthogonal transform component 14 is encoded by the encoder 4 and becomes an encoded signal 8. In general, one of the components
The transform subjected to the next transform is a beam orthogonal transform, but the orthogonal transformers 2a and 2b of the present invention are of a type that can be configured with simple hardware (that is, an organized configuration in which a high-speed algorithm or the like exists). . One example will be described with reference to FIG. The figure shows components obtained by performing orthogonal transformation of 8 × 8 points.
FIG. 3A shows an 8-point orthogonal transform similar to the MPEG 8 × 8-point discrete cosine transform (DCT) performed between fields, that is, within a frame, in the vertical direction of water, and the frequency of a component corresponding to a low frequency is calculated in the vertical direction. They are arranged in the order of the height of the business and industry. In a still image, since the correlation in the horizontal and vertical directions in a frame is large, energy is concentrated on components having low horizontal and vertical frequencies (the upper left component of the block). In FIG. 2B, orthogonal transformation of eight horizontal points and four vertical points is performed for each field, and components corresponding to low frequencies are arranged in order of frequency for each field. As above
When the orthogonal transform of DCT used in MPEG or the like is performed, the outputs shown in FIGS. 3A and 3B are obtained. However, since the correlation between the horizontal direction and the vertical direction is large even in the field, the energy is concentrated on the 0th line of the even field and the 0th line of the odd field in FIG. FIGS. 7A and 7B have different energy distributions as described above. Therefore, when encoding is performed by a common encoder, an input mismatch occurs and coding efficiency is reduced. If b) is rearranged in the order of frequency in the vertical direction as shown in FIG. 13C, the distribution of the input of the code transformer becomes similar to that of FIG. Encoders can be shared. In the case where the coding efficiency is improved when the motion is relatively small and the difference between the fields is taken to remove the pixel correlation between the fields to improve the coding efficiency, FIG. The sums and differences may be calculated as described above. In this figure, the i-th line field sum (i is an integer) represents the sum of the odd field i-th line and the even field i-th line, and the i-th line field difference is the difference between the odd field i-th line and the even field i-th line. And When the motion is small, the correlation between the fields becomes very large, and the energy of the field difference becomes much smaller than the energy of the field sum. Therefore, the energy distribution of FIG. 3D is close to the energy distribution of FIG. Become. Conversely, when the motion is large, the energy of the field difference and the value of the energy of the field sum are substantially equal, and therefore the energy distribution of FIG. It approaches the energy distribution. Therefore, it is also possible to switch between the arrangements shown in FIGS. 11D and 11E depending on the magnitude of the movement. Replace as above,
By performing a simple primary transformation of sum and difference after orthogonal transformation, it is possible to improve matching with an encoder and reduce an increase in encoding distortion when the encoder is shared. .

As described above, according to the present embodiment, the primary converter
By installing the encoders 10a and 10b after the encoders 2a and 2b to improve the matching with the encoder 4, it is possible to reduce an increase in encoding distortion when the encoder is shared.

Although two orthogonal transformers are used in the embodiment of the present invention, three or more orthogonal transformers can be similarly configured. In the embodiment of FIG. 1, the control information 7 is provided from the outside, but the control information 7 may be generated from the orthogonal transform components 3a and 3bb by using a control information generator or the like.

Effect of the Invention As described above, according to the present invention, the hardware scale can be significantly reduced without substantially increasing the coding distortion, and the practical drop effect is large.

[Brief description of the drawings]

FIG. 1 is a block diagram of a high-efficiency coding apparatus according to an embodiment of the present invention, FIG. 2 is an explanatory diagram of the effect of the high-efficiency coding apparatus of the present invention, and FIG. 3 is a block diagram of a conventional high-efficiency coding apparatus. FIG. 2 ... orthogonal transformer, 4 ... encoder, 10 ... primary converter.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) H04N 7/24-7/68 H04N 1/41-1/419

Claims (1)

(57) [Claims] 1. An image processing apparatus comprising: a block of digital image input signals; an input signal; and at least two orthogonal transformers for performing different orthogonal transformations on the input signal; and a sum, difference, and data for output components of the orthogonal transformer. A primary transformer that performs only permutation of a position, and one of a signal that is primary-transformed by the primary transformer and an output component of the orthogonal transformer is selected and encoded by one type of encoding method. A high-efficiency encoding device comprising an encoder that performs the encoding.