JP5626272B2 - Quantization controller - Google Patents

Description

  The present invention relates to a quantization control device that controls a quantization parameter when encoding a video signal.

  Quantization control in video encoding is to control quantization parameters used in quantization so that the amount of generated code generated by encoding falls within a predetermined encoding bit rate (target code amount) within a certain time width. Say. As this quantization control method, a method called a test model defined by the MPEG (Moving Picture Experts Group) method is generally used (for example, see Non-Patent Document 1).

  In the test model, a parameter called an image complexity index is used to estimate an approximate amount of data before compression. The complexity index is represented by the product of the generated code amount and the quantization parameter, and becomes large for a video with a high degree of difficulty in encoding, and becomes small for a video with a low degree of difficulty in encoding.

ISO / IEC JTC1 / SC29 / WG11 / NO400 (1993), “Test Model 5”

  In video coding, quantization is a process of thinning data, and is a process of dividing data by a product of a quantization matrix and a quantization parameter and truncating after the decimal point. Therefore, the compression rate at the time of encoding is determined by the product of the quantization parameter and the quantization matrix.

  However, conventionally, the quantization parameter is regarded as a compression ratio, and a complexity index that does not consider the quantization matrix is used when obtaining a target code amount for an image before encoding. The quantization matrix is selected according to the characteristics of the video. For this reason, when the quantization matrix used in the encoded frame for which the complexity index is calculated differs from the quantization matrix used in the frame to be encoded from now, based on the target code amount calculated from the complexity index When encoding is performed, encoding degradation occurs.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to obtain a quantization control apparatus capable of reducing video coding deterioration.

  The quantization control device according to the present invention is a quantization control device that controls a quantization parameter for encoding a video signal. A quantization matrix determination unit, a complexity index calculation unit that calculates a product of a generated code amount of an encoded video signal and a quantization parameter as a complexity index, and a correction factor that is a function of the determined quantization matrix A complexity index correction unit that multiplies the complexity index to calculate a corrected complexity index; a target code amount calculation unit that calculates a target code amount to be targeted from the generated code amount and the corrected complexity index; A quantization parameter determining unit that determines a quantization parameter so that a generated code amount when the video signal is encoded falls within the target code amount, and the correction factor is the plurality of quantization matrices. Compared to the variation in the complex index when the same video signal is encoded using hex respectively, the variation of the corrected complex index of each case, characterized in that it is set to be smaller.

  According to the present invention, video encoding degradation can be reduced.

It is a block diagram which shows the video coding apparatus which concerns on embodiment of this invention. It is a figure which shows the complexity parameter | index without correction | amendment. It is a figure which shows the corrected complexity parameter | index. It is the figure which piled up the complexity parameter | index with and without correction | amendment. It is a figure which shows the target code amount.

FIG. 1 is a block diagram showing a video encoding apparatus according to an embodiment of the present invention. This video encoding apparatus is an MPEG encoder, and includes a quantization control apparatus that controls a quantization parameter when encoding a video signal.

  The frame order changing unit 1 divides the input video signal into frames (frames) and rearranges them in the order of encoding. The motion detector 2 detects a motion vector of the video signal.

  The orthogonal transform unit 3 transforms the difference between the video signal and the predicted video from a pixel component to a frequency component. The quantization unit 4 divides the frequency component of the video signal by the product of the quantization matrix and the quantization parameter and rounds off the decimal part. The variable length coding 5 replaces the quantization result with codes of various lengths.

  The multiplexer 6 combines the variable length coding results and header information such as motion vectors into one signal. The buffer memory 7 accumulates the output signal of the multiplexer 6 and outputs it as a video stream.

  The inverse quantization unit 8 performs inverse quantization by multiplying the quantization result by the product of the quantization matrix and the quantization parameter. The inverse orthogonal transform unit 9 performs inverse orthogonal transform that transforms the inverse quantization result into pixel components. The frame accumulation and prediction unit 10 accumulates the sum of the inverse orthogonal transform result and the predicted image, and creates a predicted image from the accumulated data and motion vector.

  The quantization matrix determination unit 11 determines one from a plurality of quantization matrices according to the characteristics of the video signal. The generated code amount calculation unit 12 calculates the generated code amount G for each frame from the signal accumulated in the buffer memory 7. The complexity index calculation unit 13 calculates the product of the generated code amount G of the encoded video signal and the quantization parameter Qp as the complexity index X (X = G * Qp).

  The complexity index correction unit 14 calculates the corrected complexity index X ′ by multiplying the complexity index X by the correction rate S that is a function of the determined quantization matrix (X ′ = X * S). The target code amount calculation unit 15 calculates a target code amount targeted for each frame from the generated code amount G and the corrected complexity index X ′. The quantization parameter determination unit 16 determines the quantization parameter so that the generated code amount when the video signal is encoded falls within the target code amount.

  The correction rate S is such that the variation in the corrected complexity index X ′ in each case is smaller than the variation in the complexity index X when the same video signal is encoded using a plurality of quantization matrices. Set to

Specifically, the correction rate S is expressed by the following mathematical formula.
S = A * log (V) + B
Here, V is an encoding bit rate, and A and B are constants set for each quantization matrix.

  FIG. 2 is a diagram illustrating a complexity index without correction. FIG. 3 is a diagram illustrating the corrected complexity index. 2 and 3, as an example, in addition to the reference quantization matrix, three kinds of quantization matrixes of the quantization matrix 1 to the quantization matrix 3 are used. The values of A and B in each case are as follows. (A, B) = (0.1, 1.7), (0.2, 2.4), (0.4, 3.6). A “field” on the horizontal axis is a collection of only odd-numbered or even-numbered lines in a frame, which is one frame with two fields. It can be seen that the complexity index without correction varies in the order of several times depending on the quantization matrix, but the corrected complexity index does not depend on the quantization matrix.

  FIG. 4 is a diagram in which the complexity indices with and without correction are superimposed. The quantization matrix has a large value at a place where the complexity index is large. However, since the compression rate is high and the generated code amount is suppressed, the variation of the complexity index without correction is small compared to the variation of the quantization matrix. On the other hand, since the corrected complexity index is multiplied by a correction factor that is a function of the quantization matrix, it greatly varies with the variation of the quantization matrix.

  FIG. 5 is a diagram illustrating the target code amount. In the case of no correction, the target code amount is inappropriately reduced around the horizontal axis 196 and around the horizontal axis 361. On the other hand, when there is a correction, such a decrease can be prevented.

  Therefore, by controlling the quantization parameter using the corrected complexity index obtained by correcting the complexity index with the quantization matrix as in the present embodiment, it is possible to reduce video coding degradation.

  The present invention can also be applied to an apparatus that performs encoding a plurality of times, such as a 2-pass encoding apparatus.

DESCRIPTION OF SYMBOLS 11 Quantization matrix determination part 13 Complexity parameter | index calculation part 14 Complexity parameter | index correction | amendment part 15 Target code amount calculation part 16 Quantization parameter determination part

Claims (1)

In a quantization controller for controlling a quantization parameter when encoding a video signal,
A quantization matrix determining unit that determines one from a plurality of quantization matrices according to the characteristics of the video signal;
A complexity index calculation unit that calculates the product of the generated code amount of the encoded video signal and the quantization parameter as a complexity index;
A complexity index correction unit that calculates a corrected complexity index by multiplying the complexity index by a correction factor that is a function of the determined quantization matrix;
A target code amount calculation unit that calculates a target code amount to be targeted from the generated code amount and the corrected complexity index;
A quantization parameter determining unit that determines a quantization parameter so that a generated code amount when the video signal is encoded falls within the target code amount;
The correction rate is such that variations in the corrected complexity index in each case are smaller than variations in the complexity index when the same video signal is encoded using each of the plurality of quantization matrices. A quantization control apparatus characterized by being set to:

Images