JP4079530B2 - Video coding method and apparatus with variable bit rate - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an encoding method for compressing and encoding a moving image and an apparatus for executing the same, and more particularly, to enable encoding at a variable bit rate (VBR) that changes the compression rate with time, in real time. It is.
[0002]
[Prior art]
When moving images are transmitted through a network or recorded on a recording medium, the moving image data is compressed and encoded in order to reduce the data amount.
[0003]
In the MPEG system, which is a typical compression encoding system for moving images, compression encoding is performed by the encoder shown in FIG.
[0004]
The encoder includes a subtractor 11 that outputs a difference between the input image frame data and the previous image frame data stored in the image memory 17, a DCT converter 12 that performs DCT conversion on the difference, and a DCT conversion. A quantizing unit 13 for quantizing the DCT coefficients, a variable length encoding unit 19 for variable-length encoding the data, a buffer 20 for temporarily storing the variable-length encoded data, and a quantized value An inverse quantization unit 14 that performs inverse quantization, an inverse DCT transformation unit 15 that performs inverse DCT transformation on the output of the inverse quantization unit 14, and the previous image frame data stored in the image memory 17 and the inverse DCT transformation unit 15 The adder 16 that reproduces the input image frame data by adding the output, the image memory 17 that stores the frame data, and the input image and the image of the previous frame are compared to obtain a motion vector (MV). Motion vector to calculate And a le detection unit 18.
[0005]
In this encoder, when image data is input, the motion vector detection unit 18 compares the input image data with the image data of the previous frame stored in the image memory 17 to calculate a motion vector (MV), and the image memory 17 and the variable length coding unit 19.
[0006]
The subtraction unit 11 to which the input image data has been input reads the image data of the previous frame with compensated motion from the image memory 17, and the subtraction unit 11 reads the input image data and the previous frame read from the image memory 17. The difference from the image data is output.
[0007]
The difference data is output to the DCT conversion unit 12, which performs DCT conversion on the difference data, outputs the DCT coefficient to the quantization unit 13, and sets the type of DCT conversion to the variable length encoding unit 19. Tell.
[0008]
The quantization unit 13 quantizes the DCT coefficient at a quantization step specified by the quantization characteristic information (Quant) from the buffer 20. The quantized result and the Quant are output to the variable length encoding unit 19, and the quantized result is also output to the inverse quantization unit 14Quant.
[0009]
The variable length coding unit 19 performs variable length coding on the input motion vector, DCT type, Quant, and the output of the quantization unit 13. The encoded data is temporarily recorded in the buffer 20 and then output as a stream.
[0010]
On the other hand, the output of the quantization unit 13 is inversely quantized by the inverse quantization unit 14, is inversely DCT transformed by the inverse DCT transformation unit 15, and is added to the previous image data stored in the image memory 17 by the addition unit 16. Then, the input image data is reproduced, this image data is newly stored in the image memory 17, and used for subtracting the next input image data.
[0011]
When the quantization step specifies a fine quantization step, the encoding compression rate decreases and the data amount increases, but the image quality improves. Conversely, when Quant designates a coarse quantization step, the compression rate of encoding increases and the amount of data decreases, but the quality of the image decreases.
[0012]
When coding a scene with intense motion or an image with a complicated pattern, if an image quality comparable to that of an image with a simple pattern with little motion is maintained, the compression rate decreases and the amount of data increases.
[0013]
In the DVD package media, in order to improve the image quality while keeping the recording time within a predetermined time, the encoding compression rate is lowered when encoding a scene with intense movement or an image of a complex picture, In the case of an image with a simple pattern with little motion, encoding by VBR is performed to increase the encoding compression rate.
[0014]
Therefore, at the time of recording, once the moving image to be recorded is encoded to the end, a distribution of a complex pattern image that needs to reduce the compression rate and a simple pattern image with little movement that can increase the compression rate Then, divide the recording capacity of the recording medium by the recording time to obtain the average bit rate, assign a bit rate exceeding the average bit rate to the image of the complex pattern (lower the compression rate), and then calculate this average bit rate In order to cancel out the number of bits exceeding 1, a bit rate equal to or lower than the average bit rate is assigned to an image with a simple pattern with little motion (the compression rate is increased), and then encoding by VBR is performed again.
[0015]
It should be noted that a scene with intense motion that requires a reduction in compression rate can be obtained from a motion vector, and an image with a complicated pattern can be obtained from the degree of dispersion of DCT coefficients.
[0016]
[Problems to be solved by the invention]
However, this encoding method cannot be applied when an input moving image is encoded in real time.
[0017]
When the input moving image is encoded in real time, for example, as shown in FIG. 3A, if the input moving image is an image with a complicated pattern, the compression rate is lowered from the average level (bits). On the other hand, for a simple picture image with little motion, a bit rate control method is conceivable in which the compression rate is increased above the average level (the bit rate is decreased). However, if this control is continued, as shown in FIG. 3B, at the end time t1 of the recording time, it is necessary to finally set the accumulated value (dotted line) of the number of bits exceeding the average level to zero. Therefore, in a moving image in which a complex pattern frequently appears, even when a complex pattern image continues in the latter half of recording (t2 → t1), the compression rate is forced to be significantly higher than the average level (FIG. 3 (a ) As a result, the introduction of VBR causes a problem that the image quality deteriorates as compared with the case of a fixed rate.
[0018]
The present invention solves such problems, and an object of the present invention is to provide an encoding method and an encoding apparatus capable of performing real-time encoding by VBR on an input moving image.
[0019]
[Means for Solving the Problems]
Therefore, in the VBR encoding of the present invention, a reference bit rate slightly lower than the average bit rate when encoding at a fixed bit rate is set, and the difference between the average bit rate and the reference bit rate is multiplied by a time component. When the reserve capacity is obtained and the encoding is performed at a bit rate higher than the reference bit rate, the reserve capacity is reduced. When the encoding is performed at a bit rate lower than the reference bit rate, the reserve capacity is increased, and when the reserve capacity becomes zero. The encoding bit rate is set so as not to exceed the reference bit rate.
[0020]
Therefore, when the input moving image data is encoded in real time, the encoding at the reference bit rate is assured even when a complicated pattern frequently appears and the reserve capacity becomes 0. Can be avoided.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
According to the first aspect of the present invention, in the encoding method for encoding a moving image at a variable bit rate, a reference bit rate slightly lower than an average bit rate when encoding at a fixed bit rate is set. The spare capacity is obtained by multiplying the difference between the average bit rate and the reference bit rate by the time component, and when encoding at a bit rate higher than the reference bit rate, the reserve capacity is reduced and the bit rate lower than the reference bit rate is reduced. When encoding, the reserve capacity is increased, and when the reserve capacity reaches 0, the encoding bit rate is set so as not to exceed the reference bit rate. Since the encoding can be performed at the reference bit rate, deterioration in image quality can be avoided.
[0022]
According to the second aspect of the present invention, when the reserve capacity is not 0, an image of a scene with a high motion or a complicated pattern is encoded at a bit rate higher than the reference bit rate, and when the reserve capacity is 0, the reference bit Since it is coded at a rate, scenes with high motion and complex picture images can be coded at the reference bit rate even when the reserve capacity becomes 0, thus avoiding image quality degradation. can do.
[0023]
According to the third aspect of the present invention, a simple picture image is encoded at a bit rate lower than the reference bit rate regardless of the reserve capacity. It is possible to increase the bit rate allocation for images of intense scenes and complex patterns.
[0024]
According to a fourth aspect of the present invention, recording time on a recording medium is set as a time component, and encoded moving image data is stored in the recording medium. Can be encoded and stored in a recording medium.
[0025]
According to the fifth aspect of the present invention, the transmission time to the transmission line is set as a time component, and the encoded moving image data is transmitted to the transmission line. The input moving image data is transmitted in real time in the VBR. Can be encoded and transmitted over a network.
[0026]
According to the sixth aspect of the present invention, the encoded moving image data is transmitted at a fixed bit rate from the transmission line, and the moving image data encoded by the VBR is transmitted at a fixed bit rate for a predetermined transmission time. Can be transmitted within.
[0027]
According to the seventh aspect of the invention, the encoded moving image data is transmitted at a variable bit rate from the transmission line, and the moving image data encoded by the VBR is transmitted using the ATM VBR service. Can be transmitted.
[0028]
According to an eighth aspect of the present invention, in an encoding device for encoding a moving image at a variable bit rate, an encoding unit for encoding the moving image and a VBR setting for setting an encoding bit rate of the encoding unit Means, a reference bit rate setting means for setting a reference bit rate slightly lower than the average bit rate in the case of encoding at a fixed bit rate, and a difference between the average bit rate and the reference bit rate multiplied by a time component and spare When the capacity is obtained and the encoding is performed at a bit rate higher than the reference bit rate, the reserve capacity is reduced, and when the encoding is performed at a bit rate lower than the reference bit rate, a reserve capacity management means for increasing the reserve capacity is provided, and VBR setting means However, when the reserve capacity reaches 0, the encoding bit rate should be set so as not to exceed the reference bit rate. Are those, even spare capacity becomes 0, because it can be encoded with reference bit rate, it is possible to avoid the deterioration of image quality.
[0029]
According to the ninth aspect of the present invention, when the reserve capacity is not 0, the VBR setting means sets a bit rate higher than the reference bit rate as a bit rate for encoding a scene with intense motion or an image of a complex picture. When the reserve capacity is 0, the reference bit rate is set, and a scene with intense movement or an image of a complex picture can be encoded at the reference bit rate even if the reserve capacity becomes 0. Therefore, it is possible to avoid degradation of image quality.
[0030]
According to the tenth aspect of the present invention, the VBR setting means sets a bit rate lower than the reference bit rate as the bit rate for encoding a simple picture image regardless of the reserve capacity. Yes, the reserve capacity can be increased to increase the bit rate allocation for scenes with intense movement and images with complex patterns.
[0031]
The invention according to claim 11 is such that the reserve capacity management means calculates the initial reserve capacity by multiplying the difference between the average bit rate and the reference bit rate by the recording time to the recording medium, The moving image data to be input can be encoded by VBR in real time and stored in a recording medium.
[0032]
The invention according to claim 12 is such that the reserve capacity management means calculates the initial reserve capacity by multiplying the difference between the average bit rate and the reference bit rate by the transmission time to the transmission line, Input video data can be encoded by VBR in real time and transmitted through a network.
[0033]
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0034]
In this embodiment, a case where input moving image data is encoded by VBR in real time and recorded on a recording medium will be described.
[0035]
As shown in FIG. 2, an encoding apparatus that performs this encoding method includes an average bit rate / recording time setting unit 31 that sets an average bit rate and recording time based on a recording medium and a moving image to be recorded, and a VBR. A reference bit rate setting unit 32 for setting a reference bit rate, a spare capacity management unit 34 for managing spare capacity, a VBR setting unit 33 for setting VBR, and an encoding unit 35 for encoding moving image data And. The encoding unit 35 has the internal configuration shown in FIG.
[0036]
As shown in FIG. 1A, the reference bit rate setting unit 32 of the encoding apparatus sets a reference bit rate r1 that is slightly lower than the average bit rate r0 (thus, the compression rate is slightly higher).
[0037]
The average bit rate r0 is a bit rate when recording at a fixed rate, and the average bit rate / recording time setting unit 31 calculates the value r0 by dividing the recording capacity of the recording medium by the recording time t1. For example, when recording a moving image having a recording time of 133 minutes on a recording medium having a recording capacity of 4.7 GB, the average bit rate r0 is about 5 Mbps. The reference bit rate r1 is set to a value slightly lower than r0, for example, 4.9 Mbps.
[0038]
The spare capacity management unit 34 manages “(average bit rate r0−reference bit rate r1) × recording time t1” as the initial spare capacity.
[0039]
The VBR setting unit 33 identifies the state of the moving image from the motion vector and DCT coefficient information input from the encoding unit 35. If the input moving image has a normal pattern, the VBR setting unit 33 encodes the image. When the bit rate is set to the reference rate r1 and a scene with intense movement or complex images is input, the spare capacity is used to set the bit rate higher than the reference rate r1 (to reduce the compression rate) Set.
[0040]
When a simple pattern image with little motion is input, the bit rate is set lower than the reference rate r1 (in order to increase the compression rate).
[0041]
The encoding unit 35 performs encoding at the bit rate specified by the VBR setting unit 33 and notifies the reserve capacity management unit 34 of the number of bits used in the encoding.
[0042]
When encoding is performed at a bit rate higher than the reference bit rate, the reserve capacity management unit 34 subtracts the difference bit number from the reference rate r1 from the reserve capacity, and at a bit rate lower than the reference bit rate. When encoding is performed, the number of bits corresponding to the difference from the reference rate r1 is incorporated in the reserve capacity.
[0043]
The spare capacity management unit 34 notifies the VBR setting unit 33 when the spare capacity becomes zero. When the reserve capacity is 0, the VBR setting unit 33 sets the encoding to the reference rate r1 even when a scene with intense motion or an image with a complicated pattern is input.
[0044]
Also, if the reserve capacity remains at the end of the recording time, even if the input image is a simple picture, encoding is performed with a reduced compression rate so that the reserve capacity is used up. .
[0045]
FIG. 1B shows a temporal change in the reserve capacity when the bit rate is controlled as shown in FIG. In this example, the reserve capacity has reached 0 at time t2, so even if a complex picture image (dotted line in FIG. 1A) is input from time t2 to time t1, encoding at the reference rate r1 is performed. Is done.
[0046]
Since the reference rate r1 is only slightly lower than the average rate r0, the image quality when encoded at the reference rate r1 is almost the same as the image quality when encoded at the average rate r0.
[0047]
On the other hand, the reserve capacity can secure a considerable amount because the difference between r1 and r0 is multiplied by the recording time. In the case of a recording medium with a recording time of 133 minutes, if r0−r1 = 0.1 Mbps, the initial reserve capacity is 798 Mbit.
[0048]
Therefore, it is possible to encode an image with many complicated patterns at a bit rate higher than the reference rate. Even if the reserve capacity becomes 0, encoding at the reference rate r1 is guaranteed, so there is no possibility of incurring image quality degradation.
[0049]
As described above, in this encoding method, it is possible to execute real-time VBR encoding on input moving image data, and it is possible to improve image quality as compared with encoding at a fixed rate.
[0050]
In this embodiment, the case where the encoded data is stored in a recording medium such as a disk or a magnetic tape has been described. However, the encoded data may be transmitted through a transmission path. When this transmission is performed at a fixed bit rate, when the average rate is equal to the fixed rate, the recording time is the transmission time, and the recording capacity is the product of the fixed bit rate and the transmission time.
[0051]
Also, the encoded moving image data can be transmitted by an ATM variable speed (VBR) service. In this VBR service, transmission is started by declaring a peak value and an average value of the amount of information generated from the transmission side. If the transmission time is predetermined, the average value of moving image data to be transmitted is set. In order to maintain, this encoding method is applied.
[0052]
In this case, as in the previous explanation, a reference bit rate slightly lower than the average value is set, and the spare capacity is obtained by multiplying the difference between the reference bit rate and the average value by the transmission time, and this spare capacity has a margin. During a certain period, an image with a complicated pattern is encoded at a bit rate higher than the reference bit rate and transmitted at that bit rate. In addition, in order to increase the reserve capacity, a simple picture image is encoded at a bit rate lower than the reference bit rate and transmitted at the bit rate. When the reserve capacity runs out, even an image with a complicated pattern is encoded at the reference bit rate and transmitted at the reference bit rate.
[0053]
By doing so, it is possible to encode the moving image data to be input with a variable compression rate in real time and transmit it by VBR.
[0054]
【The invention's effect】
As is apparent from the above description, the moving image encoding method and apparatus according to the present invention enables input moving image data to be encoded at a variable compression rate in real time without causing deterioration in image quality. To do.
[Brief description of the drawings]
FIG. 1 is a diagram for explaining a VBR encoding method according to an embodiment of the present invention;
FIG. 2 is a block diagram showing a configuration of a VBR encoding apparatus according to an embodiment of the present invention.
FIG. 3 is a diagram for explaining a conventional encoding method;
FIG. 4 is a block diagram showing an MPEG encoder.
[Explanation of symbols]
11 Subtraction part
12 DCT converter
13 Quantizer
14 Inverse quantization section
15 Inverse DCT converter
16 Adder
17 Image memory
18 Motion vector detector
19 Variable length encoder
20 buffers
31 Average bit rate / recording time setting section
32 Reference bit rate setting section
33 VBR setting section
34 Reserve capacity management

Claims (12)

In an encoding method for encoding a moving image at a variable bit rate,
Set a reference bit rate slightly lower than the average bit rate when encoding at a fixed bit rate, and multiply the difference between the average bit rate and the reference bit rate by the time component to obtain the reserve capacity, which is higher than the reference bit rate. When encoding at a bit rate, the reserve capacity is reduced. When encoding at a bit rate lower than the reference bit rate, the reserve capacity is increased. When the reserve capacity becomes 0, the encoding bit rate is set as a reference. A moving image encoding method, wherein the setting is made so as not to exceed a bit rate. When the reserve capacity is not 0, a scene with intense movement or an image of a complicated pattern is encoded at a bit rate higher than the reference bit rate, and when the reserve capacity is 0, it is encoded at the reference bit rate. The moving image encoding method according to claim 1, wherein: 2. The moving picture encoding method according to claim 1, wherein a simple picture image is encoded at a bit rate lower than a reference bit rate regardless of the reserve capacity. 2. The moving image encoding method according to claim 1, wherein a recording time on a recording medium is set as the time component, and the encoded moving image data is stored in the recording medium. 2. The moving picture encoding method according to claim 1, wherein a transmission time to a transmission path is set as the time component, and the encoded moving picture data is transmitted to the transmission path. 6. The moving image encoding method according to claim 5, wherein the encoded moving image data is transmitted from the transmission path at a fixed bit rate. 6. The moving image encoding method according to claim 5, wherein the encoded moving image data is transmitted from the transmission path at a variable bit rate. In an encoding device that encodes a moving image at a variable bit rate,
Encoding means for encoding a moving image;
Variable bit rate setting means for setting an encoding bit rate of the encoding means;
A reference bit rate setting means for setting a reference bit rate slightly lower than the average bit rate when encoding at a fixed bit rate;
The spare capacity is obtained by multiplying the difference between the average bit rate and the reference bit rate by a time component, and when encoding is performed at a bit rate higher than the reference bit rate, the spare capacity is reduced to a bit lower than the reference bit rate. When encoding is performed at a rate, a reserve capacity management means for increasing the reserve capacity is provided, and the variable bit rate setting means uses the encoding bit rate as a reference when the reserve capacity becomes zero. An apparatus for encoding a moving image, wherein the setting is made so as not to exceed a bit rate. The variable bit rate setting means sets a bit rate higher than a reference bit rate when the reserve capacity is not 0 as a bit rate for encoding a scene with intense motion or an image of a complex picture, and the reserve capacity is 9. The moving picture encoding apparatus according to claim 8, wherein a reference bit rate is set when 0. 9. The variable bit rate setting means sets a bit rate lower than a reference bit rate as a bit rate for encoding a simple picture image regardless of the reserve capacity. The moving image encoding apparatus described. 9. The moving image code according to claim 8, wherein the reserve capacity management means calculates an initial reserve capacity by multiplying a difference between an average bit rate and a reference bit rate by a recording time on a recording medium. Device. 9. The moving picture code according to claim 8, wherein the spare capacity management means calculates an initial spare capacity by multiplying a difference between an average bit rate and a reference bit rate by a transmission time to a transmission path. Device.

Images