JPH114445A - Image coder and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To code image data by changing over any of plural video sources, without destroying a video buffering verify VBV buffer. SOLUTION: After coding for 1st path by an image coder 10, a host computer 1 decides an object code quantity for each image, so that a VBV buffer is not defective through the changeover of any of plural video sources, while simulating the VBV buffer and the image coder 10 conducts 2nd path coding according to the object code quantity. The image coder 10 corrects an object code quantity, so that the generated code quantity is lower than the object code quantity given by the host computer 1 and conducts coding processing according to the object code quantity after the correction.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image encoding apparatus and method for encoding image data.

[0002]

2. Description of the Related Art In recent years, a DVD (digital versatile disk or digital video disk) which is an optical disk capable of recording a large amount of digital data.
Has been put to practical use. Among DVDs, a DVD video recording video data and the like is an MPEG (Moving Picture).
Experts Group 2) is recorded.

[0003] In the MPEG2 standard, a bidirectional predictive coding method is adopted as a coding method. In this bidirectional predictive coding method, three types of coding are performed: intra-frame coding, inter-frame forward predictive coding, and bidirectional predictive coding. An image of each coding type is an I picture ( intra coded picture), P picture (pr
edictive coded picture) and B picture (bidirect
Ionally predictive coded picture). Also, a GOP (Group of Picture) serving as a unit of random access is obtained by appropriately assembling the I, P, and B pictures.
Is configured.

In an encoding method such as the MPEG standard, in which the bit generation amount differs for each picture, in order to appropriately transmit and decode the obtained bit stream to obtain an image, data in an input buffer in an image decoding apparatus is required. Occupancy,
The information must be known on the image encoding device side. Therefore, according to the MPEG standard, VBV (Vi) is a virtual buffer corresponding to an input buffer in an image decoding device.
Assuming a video buffer (de-buffering verifier) buffer, the image encoding device must generate a stream so that the VBV buffer does not break down, that is, does not cause underflow or overflow. The same applies to a DVB (Digital Video Broadcast) system for transmitting compression-coded broadcast material.

[0005]

By the way, the DVD is provided with a multi-angle function which enables any one of a plurality of images from different photographing directions to be selectively reproduced seamlessly (without seams). In a DVB system, commercial video (hereinafter, referred to as CM) may be inserted into the main broadcast material and transmitted. In order to realize such a multi-angle function and CM insertion, it is necessary to switch a plurality of streams corresponding to a plurality of video materials. However, in this case, even if each of the plurality of streams is encoded without breaking the VBV buffer, there is a problem that the VBV buffer may break down depending on how the streams are connected.

The above problem will be described in detail with reference to FIG. FIG. 7A shows the data occupancy of the VBV buffer of stream A corresponding to a certain video material (hereinafter referred to as V
It is called the BV buffer occupancy. FIG. 7B illustrates an example of a change in the VBV buffer occupancy of the stream B corresponding to another video material. Such a stream A and a stream B
Is connected at a connection point (hereinafter, referred to as a splice point) SP, after connection, the VBV buffer may underflow or may overflow as shown in FIG. 7C. is there.

The present invention has been made in view of such a problem, and an object of the present invention is to make it possible to switch a plurality of video materials without breaking a virtual buffer corresponding to an input buffer on the decoding device side. It is another object of the present invention to provide an image encoding apparatus and method capable of encoding image data.

[0008]

According to a first aspect of the present invention, there is provided an image coding apparatus, comprising: coding means for coding input image data by changing conditions according to a target code amount; With respect to the virtual buffer corresponding to the input buffer of the image decoding device that decodes the decoded image data, the data occupancy of the virtual buffer at the switching point is set to a fixed value after the switching point with another video material. When the switching point is reached, the data occupancy of the virtual buffer at the switching point is set to a value having a predetermined relationship with a constant value so that the virtual buffer does not break down due to the switching of the video material. When the code amount is given and the switching point is reached, the data occupancy of the virtual buffer at the switching point becomes a value having a predetermined relationship with a certain value. As described above, the apparatus further comprises coding control means for correcting the given target code amount in accordance with the code amount generated by the coding means and controlling the coding means in accordance with the corrected target code amount. .

According to a sixth aspect of the present invention, there is provided an image encoding method for a virtual buffer corresponding to an input buffer of an image decoding apparatus for decoding encoded image data after a switching point with another image material. Is a fixed value for the data occupancy of the virtual buffer at the switching point, and when the switching point is reached, the data occupancy of the virtual buffer at the switching point is set so that the virtual buffer does not break down due to the switching of the video material. A target code amount is given so as to be a value having a predetermined relationship with a constant value, and when reaching a switching point, the data occupancy of the virtual buffer at the switching point has a predetermined relationship with the constant value. A target code amount correction procedure for correcting a given target code amount according to the generated code amount so as to be a value, and a target code amount correction procedure. Under different conditions in accordance with the target code amount after the positive, the input image data is intended to include a coding procedure of coding.

In the image encoding apparatus according to the first aspect, the encoding control means sets the data occupancy of the virtual buffer at the switching point to a fixed value after the switching point with another video material, and When the point is reached, a target code amount is given so that the data occupancy of the virtual buffer at the switching point has a value having a predetermined relationship with a constant value. At this time, the given target code amount is modified according to the code amount generated by the encoding means so that the data occupancy amount of the virtual buffer at the switching point has a value having a predetermined relationship with a fixed value. The encoding means is controlled in accordance with the corrected target code amount, and the encoding means encodes the input image data by changing conditions according to the corrected target code amount.

According to the image coding method of the present invention, the data occupancy of the virtual buffer at the switching point is set to a constant value after the switching point with another video material by the target code amount correction procedure, Is reached, the target code amount is given so that the data occupancy of the virtual buffer at the switching point has a value having a predetermined relationship to a certain value. The given target code amount is modified in accordance with the generated code amount so that the data occupancy of the dynamic buffer becomes a value having a predetermined relationship with a fixed value. The input image data is encoded by changing the conditions according to the amount.

[0012]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of an image encoding device according to one embodiment of the present invention.

Image coding apparatus 10 according to the present embodiment
Receives an input image data S ₁ , performs an encoder control unit 11 for performing pre-processing for compression encoding and the like, and receives an output data of the encoder control unit 11 and performs encoding according to a picture type for each picture. An encoder 13 serving as an encoding unit that outputs the compressed image data S ₂ by performing compression encoding by a method; a motion detection circuit 14 that detects a motion vector based on output data of the encoder control unit 11 and sends the motion vector to the encoder 13; The encoder 13 is controlled by inputting the intra AC data S ₃ output from the encoder control unit 11, the ME residual data S ₄ output from the motion detection circuit 14, and the generated bit amount data S ₅ output from the encoder 13. And an encoding control unit 15 as encoding control means for performing the encoding. The host computer 1 is connected to the image encoding device 10.

Intra-AC refers to the absolute value of the difference between the pixel value of each pixel in an 8 × 8 pixel DCT (discrete cosine transform) block and the average value of the pixel values in the DCT block in an I picture. It is defined as a sum and can be said to represent the complexity of the picture. Intra AC data S ₃
This is data representing the intra AC. The ME residual is simply the sum of the absolute value or the square of the motion prediction error for the entire picture, and the ME residual data S ₄ is data for calculating the ME residual.
This will be described in detail later.

An encoding control unit 15 includes a CPU (central processing unit) 16 and a ROM connected to each other via a bus 19.
(Read-only memory) 17 and a computer having a RAM (random access memory) 18. The CPU 16 executes the image encoding control program stored in the ROM 17 using the RAM 18 as a working area. The functions of the encoding control unit 15 described below are realized. The ROM 17 may be an IC (integrated circuit) or an IC
It may be a card, a storage device using a magnetic disk such as a hard disk or a floppy disk as a recording medium, a storage device using an optical disk such as a CD (compact disk) -ROM as a recording medium, or other types of recording. A storage device using a medium may be used.

The host computer 1 includes an encoding control unit 1
The CPU 16 receives the generated bit amount data S ₅ and the encoding difficulty data S ₆ from the CPU 16 and sends the target code amount data S ₇ to the CPU 16.

FIG. 2 shows the image encoding apparatus 10 shown in FIG.
FIG. 3 is a block diagram showing a detailed configuration. As shown in this figure, the encoder controller 11 controls the input image data S ₁
And an image rearranging circuit 21 for rearranging the order of pictures (I picture, P picture, and B picture) in accordance with the encoding order, and inputting output data of the image rearranging circuit 21 to input a frame structure or a field structure. A scan conversion / macroblock conversion circuit 22 for performing scan conversion and macroblock formation of 16 × 16 pixels according to the determination result;
Is input, and intra A
C is calculated, and the intra AC data S ₃ is
5 and a scan conversion / macroblock conversion circuit 22
Output data from the encoder 13 and the motion detection circuit 14
And an intra AC operation circuit 23 for sending the data to the computer.

The encoder 13 includes an encoder control unit 11
A subtraction circuit 31 for obtaining a difference between the output data of the subtraction circuit and the predicted image data.
A DCT circuit 32 that performs DCT in units of T blocks and outputs DCT coefficients, a quantization circuit 33 that quantizes output data of the DCT circuit 32, and a variable circuit that performs variable length encoding of output data of the quantization circuit 33. A long encoding circuit 34;
The output data of the variable length encoding circuit 34 is temporarily held,
A buffer memory 35 to output as compressed image data S ₂ made from the bitstream, an inverse quantization circuit 36 for inversely quantizing the output data of the quantization circuit 33, inverse DCT on the output data of the inverse quantization circuit 36 Inverse DCT
Circuit 37, an addition circuit 38 for adding the output data of the inverse DCT circuit 37 and the prediction image data and outputting the added data, and holding the output data of the addition circuit 38,
And a motion compensation circuit 39 that performs motion compensation in accordance with the motion vector sent from the controller and outputs predicted image data to the subtraction circuit 31 and the addition circuit 38. The buffer memory 35 sends generated bit amount data S ₅ representing the bit amount generated by the variable length coding circuit 34 to the coding control unit 15.

The motion detection circuit 14 includes the encoder control unit 1
1 is a macro that minimizes the sum of absolute values or the sum of squares of pixel value differences between the macroblock of interest of the picture to be compression-encoded and the macroblock of interest in the referenced picture based on the output data of A block is searched, a motion vector is detected, and sent to the motion compensation circuit 39. The motion detecting circuit 14 sends when obtaining the motion vector, the absolute value sum or square sum of differences of pixel values between the minimum and since macroblock, the encoding control unit 15 as the ME residual data S ₄ It has become.

The encoding control unit 15 calculates an ME residual value which is a value obtained by adding the ME residual data S ₄ from the motion detecting circuit 14 for the entire picture.
Based on the ME residual calculated by the ME residual calculator 41 and the intra AC data S ₃ from the intra AC operation circuit 23, an encoding difficulty indicating the encoding difficulty of the picture is calculated. , The encoding difficulty level calculating unit 42 that sends the encoding difficulty level data S ₆ to the host computer 1, and the target bit rate data S ₅ from the host computer 1 based on the generated bit rate data S ₅ from the buffer memory 35.
₇ , and a quantizer 33 in the quantization circuit 33 so that the generated code amount in the encoder 13 becomes the target code amount after being corrected by the target code amount corrector 44. A quantization index determination unit for determining a quantization index corresponding to the quantization characteristic value and sending the quantization index to the quantization circuit.

Here, the encoding difficulty will be described.
The encoding difficulty represents the difficulty of encoding a picture, which can be rephrased as a ratio of the amount of data required to maintain the same image quality. There are various methods for digitizing the encoding difficulty, but in the present embodiment,
For I-pictures, the coding difficulty is obtained using intra AC, and for P-pictures and B-pictures, ME
The encoding difficulty is determined using the residual. As described above, the intra AC represents the complexity of the picture, and the ME
The residual indicates the speed of movement of the video and the complexity of the picture, and since these have a strong correlation with the difficulty of encoding, the intra AC and the linear function using the ME residual as a variable, etc. It is possible to calculate the encoding difficulty from the ME residual.

Next, the operation of the image coding apparatus according to this embodiment will be described. The following description also serves as a description of the image encoding method according to the present embodiment.

First, the basic operation of the image encoding device 10 will be described. First, the input image signal S ₁ is input to the image rearranging circuit 21 of the encoder control unit 11. The image rearranging circuit 21 rearranges the order of pictures according to the order of encoding. The output image data of the image rearranging circuit 21 is supplied to a scan conversion / macroblocking circuit 2
2 is input. Scan conversion / macroblock circuit 22
Discriminates between a frame structure and a field structure, and performs scan conversion and macroblock conversion according to the discrimination result. The output data of the scan conversion / macroblock conversion circuit 22 is input to the intra AC operation circuit 23. In the case of an I picture, the intra AC operation circuit 23 calculates an intra AC and sends the intra AC data S ₃ to the encoding control unit 15. The output data of the scan conversion / macroblock conversion circuit 22 is sent to the encoder 13 and the motion detection circuit 14 via the intra AC operation circuit 23.

In the case of an I picture, the encoder 13 directly inputs the output data of the encoder control unit 11 to the DCT circuit 32 to perform DCT without making a difference from the predicted image data in the subtraction circuit 31 to perform DCT. The DCT coefficient is quantized by the quantization circuit 33, the output data of the quantization circuit 33 is variable-length encoded by the variable length encoding circuit 34, and the output data of the variable length encoding circuit 34 is temporarily held by the buffer memory 35, and the compressed image and outputs it as the data S _2. Further, the output data of the quantization circuit 33 is inversely quantized by the inverse quantization circuit 36, and the inverse DCT circuit 37
Inverse DC with respect to the output data of the inverse quantization circuit 36
T is performed, and the output image data of the inverse DCT circuit 37 is input to the motion compensation circuit 39 via the addition circuit 38 and held.

In the case of a P picture, the encoder 13 in the encoder 13 stores the past I
It generates predicted image data based on the image data corresponding to the picture or the P picture and the motion vector from the motion detection circuit, and outputs the predicted image data to the subtraction circuit 31 and the addition circuit. Further, the difference between the output data of the encoder control unit 11 and the predicted image data from the motion compensation circuit 39 is calculated by the subtraction circuit 31, DCT is performed by the DCT circuit 32, and DT is performed by the quantization circuit 33.
The CT coefficients quantizes the output data of the quantization circuit 33 by the variable-length coding circuit 34 variable-length coding, as to temporarily holds the output data of the variable length coding circuit 34 by the buffer memory 35 the compressed image data S ₂ I do. Further, the output data of the quantization circuit 33 is inversely quantized by the inverse quantization circuit 36, inverse DCT is performed on the output data of the inverse quantization circuit 36 by the inverse DCT circuit 37, and the inverse DCT circuit 37 The output data and the predicted image data are added and input to the motion compensation circuit 39 to be held.

In the case of a B picture, the encoder 13 uses the motion compensation circuit 39 to store two pieces of image data corresponding to the past and future I pictures or P pictures and the two motions from the motion detection circuit 14. It generates predicted image data based on the vector and outputs the predicted image data to the subtraction circuit 31 and the addition circuit. Further, the difference between the output data of the encoder control unit 11 and the predicted image data from the motion compensation circuit 39 is calculated by the subtraction circuit 31, the DCT is performed by the DCT circuit 32, and the DCT coefficient is quantized by the quantization circuit 33, and The output data of the quantization circuit 33 is subjected to variable length encoding by the long encoding circuit 34, and the output data of the variable length encoding circuit 34 is temporarily held by the buffer memory 35, and the compressed image data S
Output as ₂ . Note that the B picture is the motion compensation circuit 3
9 is not held.

The buffer memory 35 sends the generated bit amount data S ₅ representing the bit amount generated by the variable length coding circuit 34 to the coding control unit 15.

The encoding control unit 15, and sends the generated bit amount data S ₅ to the host computer 1 will be described later 1
In the coding of the pass, the coding difficulty calculator 42 calculates the coding difficulty, and sends the coding difficulty data S ₆ representing the coding difficulty to the host computer 1. The encoding control unit 15, the encoding of the second pass to be described later, by the target code amount correcting unit 44 corrects the target code amount given as the target code amount data S ₇ from the host computer 1, quantization The index determination unit 45 determines a quantization index corresponding to the quantization characteristic value in the quantization circuit 33 so that the generated code amount in the encoder 13 becomes the target code amount after being corrected by the target code amount correction unit 44. , To the quantization circuit 33.

Image coding apparatus 10 according to the present embodiment
Is applicable to both fixed rate coding and variable rate coding. In encoding data to be recorded on a storage medium such as a DVD, a variable rate method is adopted. In the fixed-rate encoding, encoding is performed so that the VBV buffer does not break down, that is, does not cause underflow or overflow. In contrast, in variable rate coding, data transmission to the VBV buffer is performed at the maximum rate, and transmission can be stopped when the VBV buffer is full. FIG. 3 shows an example of a change in VBV buffer occupancy in variable rate coding. As shown in this figure, in the case of the coding of the variable rate method, the transmission can be stopped when the VBV buffer is full. No overflow occurs. However, when the generated code amount becomes equal to or larger than the target code amount, as in the case of the fixed rate method, VB
An underflow of the V buffer occurs.

In the present embodiment, in order to realize a multi-angle function in DVD and CM insertion in DVB, image data corresponding to a video material switchable with another video material is used by using the image encoding device 10. Encode. In the multi-angle function of the DVD, switching of a plurality of streams is performed at a portion called an angle point. This angle point is, for example, GOP
At the beginning. When switching streams at this angle point, it is necessary to prevent the VBV buffer from breaking down. Similarly, in the case of CM insertion in DVB, when switching a plurality of streams,
The VBV buffer must not be broken.

Here, with reference to FIG. 7, a description will be given of conditions under which the VBV buffer is not broken when the streams are switched. Here, as shown in FIGS. 7A and 7B, the streams A and B are switched at a certain splice point (corresponding to the switching point in the present invention) SP, and each stream A at the splice point SP is switched. , And B occupy the VBV buffers, respectively.
_OCA , VBV _OCB . Further, the VBV _OCB, the depth of the VBV buffer stream B is used for the next splice point SP and D _B. At this time, if the following equation (1) is satisfied, V
Stream switching can be performed without breaking the BV buffer.

VBV _OCA = VBV _OCB (1)

In the case of variable rate coding, there is no fear of overflow. By satisfying the following expression (2), it is possible to switch streams without breaking the VBV buffer.

VBV _OCA ≧ D _B (2)

[0035] In practice, the depth of the VBV buffer stream B is utilized, so do not know if I actually coded, the value of D _B, the stream B and the depth of the VBV buffers available I do. Also, simply, D
_The following equation (2 ′) may be used as _B式 VBV _OCB instead of equation (2).

VBV _OCA ≧ VBV _OCB (2 ')

Next, the operation of the image encoding apparatus 10 according to the present embodiment will be described in detail, taking as an example a case where a multi-angle function of a DVD is realized by performing encoding by a so-called two-pass encoding method. . Note that the two-pass encoding method means that the image encoding device 10
The first (first pass) encoding is performed with the quantization scale kept constant, and the encoding difficulty (Difficulty) representing the encoding difficulty of the picture is calculated by the host computer 1.
And the like, and a target code amount is determined for each picture based on the information.
This is a method of encoding the (pass-th). In this case, in the present embodiment, the host computer 1 calculates a target code amount for each picture so as to satisfy Expression (1) or Expression (2) while simulating a VBV buffer. DV
In the case of D, variable-rate encoding is performed, so that equation (2) may be satisfied.

In order to satisfy the equation (2) at any splice point SP, the host computer 1 firstly sets the splice point SP and thereafter after the splice point SP as shown in FIG. It is the VBV buffer occupancy at the splicing point SP as a constant value V _C, to compute the subsequent target code amount. The host computer 1, when reaching the splicing point SP, if VBV buffer occupancy at the splicing point SP satisfies a predetermined relationship to a predetermined value V _C (Equation (2) is VBV buffer occupancy as but a value having a relationship) exceeding a predetermined value V _C, calculates the target code amount. Incidentally, a constant value V _C, for example equal to 2/3 of the total capacity of the VBV buffer can be set appropriately.

Here, the encoding process of the two-pass encoding method in the present embodiment will be described with reference to FIG. In this process, first, initialization for inputting and setting encoding conditions to the host computer 1 is performed (step S101).
The encoding of the first pass is performed with the quantization scale kept constant, and the encoding difficulty is measured. Coding difficulty is determined by the coding difficulty calculating unit 42 in FIG. 2, as the encoding difficulty data S _6, is provided to the host computer 1. Next, the host computer 1 allocates a picture type (step S103), and further allocates bits to each picture (calculates a target code amount) (step S104). Next, the host computer 1 gives the target code amount for each picture in the picture coding apparatus 10 by the target code amount data S _7, the image encoding device 1
0, the second pass is encoded (step S10).
5) The obtained compressed image data is displayed on the monitor device and previewed (step S106), and the image quality is checked (step S107). If the image quality is good (OK), the data necessary for authoring is compiled. And the like (step S110), and the image encoding operation ends. In the image quality check, if the image quality is not good (NG), customization for partially changing the image quality by partially changing the encoding conditions is performed (step S108). (Step S10)
9) Returning to step S105, the second pass encoding is performed again.

Here, an example of a process of allocating bits (calculating a target code amount) in the host computer 1 will be described. In this process, a target code amount of the encoding process is set according to the encoding difficulty. First, a block of image data that can be encoded continuously (hereinafter, referred to as a block) according to a weighting table set in advance. It is possible to weight the encoding difficulty for each encoding unit.) The host computer 1 sequentially detects and arranges the maximum value of the encoding difficulty between the corresponding GOPs of the encoding unit that realizes the seamless multi-angle function, thereby providing the encoding unit that realizes the seamless multi-angle function. On the other hand, a common encoding difficulty is set. As a result, the data amount becomes equal between the corresponding encoding units.

Next, the host computer 1 checks each GOP
The code amount that can be actually allocated is allocated to each GOP according to the coding difficulty of the unit, and further, in each GOP,
A target code amount is allocated to each picture according to the encoding difficulty of each picture. In this way, in the seamless multi-angle function, after the same target code amount is assigned to the corresponding GOP according to the standardized coding difficulty, the coding difficulty of each picture constituting each GOP is determined. , The same target code amount is sequentially allocated to each picture.

Next, while simulating the VBV buffer, the host computer 1 corrects the target code amount for each picture so that the VBV buffer does not underflow. Further, the host computer 1 corrects the target code amount for each picture so as to satisfy Expression (2), for example, in GOP units.

At the time of the second pass encoding, the image encoding device 10 executes the host computer 1 as described above.
The encoding process is performed according to the target code amount for each picture determined and given. However, since the actual generated code amount has an error with respect to the target code amount, the trajectory of the VBV buffer occupancy as simulated by the host computer 1 is not reproduced. Where,
When the generated code amount is smaller than the target code amount, Equation (2) is satisfied, so there is no problem.
If the generated code amount is larger than the target code amount, VBV
This can cause buffer underflow. Therefore, in the present embodiment, the image encoding device 10 performs the operation shown in the flowchart of FIG. 6 so as to satisfy Expression (2).

In this operation, first, the encoding control unit 15
Sets the initial value of the total error defined by the following equation (3) to 0, and sets the number of pictures up to the next splice point (step S201).

Total error = Total target code amount−Total generated code amount (3)

Next, the target code amount correction unit 44 of the encoding control unit 15 acquires the target code amount from the host computer 1 (step S202). Next, the target code amount correction unit 4
4 judges whether the total error is 0 or more (step S2).
03). If the total error is 0 or more (Y), step S2
Go to 05. If the total error is not equal to or greater than 0 (N), the target code amount correction unit 44 corrects the target code amount according to the following equation (4) (step S204), and step S205
Proceed to.

Target code amount after correction = target code amount before correction− | total error | / number of remaining pictures (4)

That is, the excess of the generated code amount with respect to the given target code amount is divided into the remaining pictures, and the excess amount divided into each picture is subtracted from the target code amount given to each picture to correct it. A subsequent target code amount is calculated.

Next, in step S205, the corrected target code amount is provided from the target code amount correction unit 44 to the quantization index determination unit 45, and the quantization index determination unit 45
Thus, the quantization index corresponding to the quantization characteristic value in the quantization circuit 33 is determined so that the generated code amount in the encoder 13 becomes the target code amount after being corrected by the target code amount correction unit 44. 33, and encodes one picture.

Next, the target code amount correction unit 44 acquires the generated code amount (step S206), calculates the total error (step S207), and reduces the number of remaining pictures by one (step S208). Next, the encoding control unit 15 determines whether or not the number of remaining pictures is greater than 0 (step S2).
09) If it is larger (Y), the process returns to step S202; if not (N), the encoding process ends.

By the above-described processing shown in FIG. 6, the target code amount is corrected so that the total generated code amount is smaller than the total target code amount. When the code amount falls below, it is possible to satisfy Expression (2) at the splice point or to make the excess of the total generated code amount with respect to the total target code amount fall within the allowable error range.

When there are a plurality of GOPs between adjacent splice points, the operation shown in FIG. 6 may be performed only for the GOP immediately before the splice point.

As described above, in the present embodiment,
In any splicing point SP, in order to satisfy equation (2), first, the host computer 1, for splicing point SP after, the VBV buffer occupancy at the splicing point SP as a constant value V _C, subsequent target the code amount is calculated and when reaching the splicing point SP, the predetermined relationship VBV buffer occupancy is for a fixed value V _C at the splicing point SP (relation as VBV buffer occupancy exceeds a predetermined value V _C) Is calculated and given to the encoding control unit 15 of the image encoding device 10.
In the present embodiment, further, in consideration of the fact that the actual generated code amount has an error with respect to the target code amount, the coding control unit 15 of the image coding device 10 sets the VBV buffer at the splice point SP. predetermined relationship quantity occupied for a fixed value V _C (VBV buffer occupancy is constant value V _C
Is corrected in accordance with the generated code amount so that encoding is performed in accordance with the corrected target code amount. As described above, according to the present embodiment, when a plurality of video materials can be switched, for example, when a multi-angle function is realized by variable-rate encoding in a DVD, the VBV buffer is broken. It is possible to encode image data so that a plurality of video materials can be switched without generating a stream that can be reproduced with minimum degradation of image quality.

Further, according to the present embodiment, the excess of the generated code amount with respect to the given target code amount is divided into at least a predetermined number of pictures before reaching the splice point, and the excess Is subtracted from the target code amount given to each picture to calculate the corrected target code amount, so that it is possible to prevent the image quality of a specific picture from deteriorating.

The present invention is not limited to the above-described embodiment, but can be modified in various ways as described below.

First, in the correction of the target code amount in step S204 in FIG. 6, the excess of the generated code amount with respect to the given target code amount is equally divided into all the remaining pictures. However, when predictive coding such as the MPEG standard is performed, it is necessary to allocate a larger code amount to a picture on the predicted side (I picture or P picture) in order to maintain the overall image quality. It is valid. Therefore, the ratio of dividing the excess into each picture may be changed according to the picture type, and a larger code amount may be assigned to the predicted picture. Here, an example of a method of changing the ratio of dividing the excess into each picture according to the picture type will be described. In this method, I
No excess subtraction is performed for pictures. Also, P
For picture and B-picture, the coefficient C _P corresponding to the percentage of dividing the excess and sets C _B. Then, in step S204 of FIG. 6, for the P picture, the corrected target code amount is calculated according to the following equation (5).

Target code amount after correction = Target code amount before correction−C _P · | Total error | / Number of remaining pictures (5)

Similarly, in step S204 of FIG.
For the picture, the corrected target code amount is calculated according to the following equation (6).

Target code amount after correction = Target code amount before correction−C _B · | Total error | / Number of remaining pictures (6)

[0060] Then, C _P = 1.0, as C _B = 1.4, if the C _B greater than C _P, with P-picture is a picture of the expected side, compared to the B picture ,
The code amount to be subtracted from the target code amount before the correction is reduced, and therefore, the allocated code amount is kept relatively large, so that deterioration of the image quality can be prevented as a whole.

The method of calculating the corrected target code amount by changing the ratio of dividing the excess into each picture in accordance with the picture type is not limited to the equations (5) and (6), but can be variously changed. .

In consideration of a digit loss at the time of calculation, 1
It is preferable that the corrected target code amount is conservative by raising the excess amount to be subtracted per picture so that the total generated code amount does not exceed the total target code amount.

The present embodiment is also applicable to the case of fixed rate coding. In the case of the fixed rate coding, since the overflow of the VBV buffer must not occur, the equation (1) must be satisfied at the splice point. In this case, the host computer 1 determines a target code amount so as to satisfy Expression (1) and supplies the target code amount to the coding control unit 15 of the image coding device 10. The encoding control unit 15 corrects the target code amount so that the generated code amount is lower than the target code amount by performing the same processing as in the case of the variable rate method, and generates the generated code for the given target code amount. Formula (1) can be satisfied by generating dummy data “0” to compensate for the shortage of the amount and performing stuffing to embed the dummy data in the bit stream. Accordingly, even when performing fixed-rate encoding, for example, as in the case of CM insertion in DVB, image data is encoded so that a plurality of video materials can be switched seamlessly without breaking the VBV buffer. It is possible to generate a stream that can be reproduced while minimizing deterioration of image quality.

[0064]

As described above, claims 1 to 5
According to the image encoding device according to any one of the above or the image encoding method according to any one of claims 6 to 10, after the switching point with another video material, the data of the virtual buffer at the switching point The occupancy is fixed,
When the switching point is reached, the target code is set so that the data occupancy of the virtual buffer at the switching point has a value having a predetermined relationship with a certain value so that the virtual buffer does not break down due to the switching of the video material. Given the amount, when the switching point is reached, the given target according to the generated code amount such that the data occupancy of the virtual buffer at the switching point has a value having a predetermined relationship with a constant value. Since the code amount is corrected and the condition is changed according to the corrected target code amount, the input image data is coded, so that the virtual buffer corresponding to the input buffer on the decoding device side is not broken. There is an effect that image data can be encoded so that a plurality of video materials can be switched.

According to the image coding apparatus of the second or third aspect or the image coding method of the seventh or eighth aspect, the excess of the generated code amount with respect to the given target code amount is determined at least at the switching point. Is divided into a predetermined number of pictures before reaching, and the excess divided into each picture is subtracted from the target code amount given to each picture to calculate the corrected target code amount. There is an effect that deterioration of the image quality of a specific picture can be prevented.

Further, according to the image encoding device of the third aspect or the image encoding method of the eighth aspect, the ratio of dividing the excess into each picture is changed according to the picture type. There is an effect that deterioration of the overall image quality can be prevented.

Further, according to the image coding apparatus of the fourth or fifth aspect or the image coding method of the ninth or tenth aspect, the generated code amount is smaller than the given target code amount. Since the obtained target code amount is corrected, there is an effect that the failure of the virtual buffer can be more reliably prevented.

According to the image coding apparatus of the fifth aspect or the image coding method of the tenth aspect, the dummy data for compensating for the shortage of the generated code amount with respect to the given target code amount is generated. As a result, even in the case of performing fixed-rate encoding, there is an effect that a plurality of video materials can be switched seamlessly without breaking the virtual buffer.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of an image encoding device according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a detailed configuration of an image encoding device in FIG. 1;

FIG. 3 is an explanatory diagram illustrating an example of a change in VBV buffer occupancy in variable rate coding.

FIG. 4 is an explanatory diagram showing an example of a change in VBV buffer occupancy according to an embodiment of the present invention.

FIG. 5 is a flowchart showing an encoding process of a two-pass encoding method according to an embodiment of the present invention.

FIG. 6 is a flowchart illustrating an operation of the image encoding device according to the embodiment of the present invention.

FIG. 7 is an explanatory diagram for describing a case where a VBV buffer fails due to connection of a stream.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Host computer, 10 ... Image coding apparatus, 11
... Encoder control unit, 13 encoder, 14 motion detection circuit, 15 encoding control unit, 16 CPU, 44 target code amount correction unit, 45 quantization index determination unit.

Claims (10)

[Claims] 1. An image coding apparatus for coding image data corresponding to a video material switchable with another video material, wherein the input image data is coded by changing conditions according to a target code amount. Encoding means, and a virtual buffer corresponding to the input buffer of the image decoding apparatus for decoding the image data encoded by the image encoding apparatus. The data occupancy of the virtual buffer at the switching point is set to the constant value so that the virtual buffer does not break down due to the switching of the video material when the switching point is reached. The target code amount is given to a value having a predetermined relationship with respect to the virtual buffer. The given target code amount is corrected in accordance with the code amount generated by the encoding means so that the data occupancy has a value having a predetermined relationship with the constant value. And a coding control means for controlling the coding means according to the following. 2. The encoding means encodes input image data for each picture by an encoding method corresponding to a picture type, and said encoding control means calculates an excess of a generated code amount with respect to a given target code amount. Is divided into at least a predetermined number of pictures before reaching the switching point, and the excess divided into each picture is subtracted from the target code amount given to each picture to calculate a corrected target code amount. The image encoding device according to claim 1, wherein: 3. The image coding apparatus according to claim 2, wherein said coding control means changes a ratio of dividing an excess into each picture according to a picture type. 4. The image code according to claim 1, wherein said coding control means corrects the given target code amount so that the generated code amount is lower than the given target code amount. Device. 5. The image coding apparatus according to claim 4, wherein said coding control means generates dummy data for compensating for a deficiency of a generated code amount with respect to a given target code amount. 6. An image encoding method for encoding image data corresponding to an image material switchable with another image material, comprising: an input buffer on an image decoding device side for decoding the encoded image data. For the virtual buffer corresponding to, after the switching point with another video material, the data occupancy of the virtual buffer at the switching point is set to a constant value,
When the switching point is reached, the target is set so that the data occupancy of the virtual buffer at the switching point has a value having a predetermined relationship with the constant value so that the virtual buffer does not break down due to the switching of the video material. Given the code amount, when reaching the switch point, so that the data occupancy of the virtual buffer at the switch point is a value having a predetermined relationship with the constant value, according to the generated code amount,
A target code amount correction procedure for correcting a given target code amount, and an encoding procedure for encoding input image data by changing conditions according to the target code amount after the correction by the target code amount correction procedure. An image coding method characterized by the above-mentioned. 7. The encoding step encodes input image data for each picture by an encoding method according to a picture type, and the target code amount correcting step includes a step of exceeding a generated code amount for a given target code amount. Calculating the corrected target code amount by dividing the minute into at least a predetermined number of pictures before reaching the switching point, and subtracting the excess divided into each picture from the target code amount given to each picture. The image encoding method according to claim 6, wherein: 8. The image encoding method according to claim 7, wherein said target code amount correction procedure changes a ratio of dividing an excess amount into each picture according to a picture type. 9. The image according to claim 6, wherein the target code amount correcting step corrects the given target code amount such that the generated code amount is lower than the given target code amount. Encoding method. 10. The image encoding method according to claim 9, further comprising a step of generating dummy data for compensating for a shortage of a generated code amount with respect to a given target code amount.