KR20090030778A - Apparatus and method for multi-pass encoding - Google Patents

Abstract

A multi-path encoding apparatus and a method thereof are provided to shorten the time required for the multi-path encoding. The first encoder and an input delay unit receive the original image data at the same time(S300). The first encoder encodes the inputted image data and then sequentially stores the encoded result in a buffer(S310). A bit rate control unit sets up and stores the quantization coefficient to be applied to the second encoder(S320). The input delay unit outputs image data after a certain time elapses from the time when the firs frame of the original image data is inputted(S330). The second encoder applies the quantization coefficient stored in a memory to the quantization step to encode the inputted image data(S340).

Description

Apparatus and method for multi-pass encoding

The present invention relates to an apparatus and a method for performing multipath encoding, and more particularly, to a multipath encoding apparatus and a method capable of bitrate control.

Multimedia data is internally composed of audio data and video data. In general, in a multimedia service, such audio data and video data are processed in a bitstream form and transmitted and received on a network according to a predetermined bitrate. Bitrate means how many bits (bits) the multimedia data corresponding to one second. Thus, the larger the bitrate, the more bits the video has and the higher the quality, but that requires a higher amount of capacity.

In general, a method of encoding image data of multimedia data in the form of a bit stream includes an I frame (Infra Frame) and a P frame (Predictive Frame).

I-frames are independent frames that can be compressed and reconstructed on their own without reference to other images. They can come anywhere in the datastream during production and allow arbitrary access to the data. The I frame is a full screen image compressed directly from the original source, but has a good quality, but also has a large disadvantage.

The P frame performs encoding or decoding based on the amount of change between front and rear frames. That is, focusing on the fact that the blocks of the image are moved sideways instead of changing the entire image among successive images, encoding and decoding are performed using the information of the previous I frame information and the previous P frame. Therefore, the capacity is relatively small compared to the I frame.

At this time, as a method of controlling the bit rate of the video data to provide a multimedia service on the network, there are a constant bit rate (CBR) and a variable bit rate (VBR).

Since CBR uses only fixed bits, network transmission efficiency is high. However, in case of moving images, I frame is damaged and image quality deteriorates. Therefore, it is inappropriate for use in a moving video such as sports, action movies, and the like.

VBR is more efficient because it can produce better image quality than CBR method because the bitrate changes variably according to screen movement. However, when the scene changes or there is a lot of information on the screen, the bitrate may peak at a moment, causing a packet loss or other data communication.

In other words, if the instantaneous non-peak peaks during multimedia data transmission within a limited bandwidth on the network, a packet transmission error occurs, resulting in packet loss of multimedia data and deterioration in image quality.

In order to solve this problem, a multi-pass encoding technique has been developed. Multi-pass encoding is a method of first encoding image data through a first encoder and then performing two encodings of image data through a second encoder using a first encoding result.

Such multi-pass encoding enables even distribution of bitrates and encoding of even image quality, but there is a problem in that encoding processing time is increased by performing a second encoding after all of the first encoding is completed.

Accordingly, an object of the present invention is to provide a multi-pass encoding apparatus and method for reducing encoding time required for multi-pass encoding.

In order to achieve the above-described problem, the apparatus for multi-pass encoding the original image data consisting of a plurality of frames according to an aspect of the present invention, the first encoder receiving the original image data to perform a first encoding, A bit rate controller configured to set a quantization coefficient for each frame according to the amount of bits for each frame detected from the first encoding result, an input delay unit for outputting the original image data by delaying a first time from a time point at which the first encoding starts; And a second encoder configured to perform second encoding on the image data output from the input delay unit by applying the set quantization coefficient.

In addition, according to another aspect of the present invention, a method for multi-pass encoding original video data consisting of a plurality of frames includes a frame using a bitstream generated as a result of first encoding the original video data sequentially from the first frame. Detecting a bit amount for each frame, comparing the bit amount for each frame with a reference bit amount, setting and storing a quantization coefficient of each frame according to a result of the comparison, and a time point at which the first frame starts to be encoded And after the first time elapses from the second frame, sequentially encoding the original image data from the first frame by applying the set quantization coefficient.

According to the exemplary embodiment of the present invention, stable multimedia data transmission is possible, and encoding time can be shortened in encoding such multimedia data.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is said to "include" a certain component, it means that it can further include other components, without excluding other components unless specifically stated otherwise. In addition, the terms “… unit”, “… unit”, “module”, etc. described in the specification mean a unit that processes at least one function or operation, which may be implemented by hardware or software or a combination of hardware and software. have.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of a multi-pass encoding apparatus according to an embodiment of the present invention.

As shown in FIG. 1, a multi-pass encoding apparatus according to an embodiment of the present invention may include a first encoder 100, a buffer 110, a bit rate controller 120, an input delay unit 130, and a second encoder 140. ).

The first encoder 100 compresses the input original image data and encodes the original image data into a bitstream format. In this case, the original image data includes a plurality of frames, and the first encoder 100 encodes the current frame image data by using motion prediction and correction of the previous frame image. In addition, the first encoder 100 outputs the encoded bitstream to the buffer 110.

A detailed configuration and operation of the first encoder 100 will be described with reference to FIG. 3 later.

The buffer 110 stores the bitstream encoded by the first encoder 100. In this case, the buffer 110 stores the bitstream output from the first encoder 100 in a 'first in-first out' (FIFO) manner.

The bit rate controller 120 controls the bit rate when encoding the original image data to encode the input image data so that the input image data can be stably transmitted to a communication channel or a storage medium.

In detail, the bit rate controller 120 controls the image data to be input to the second encoder 140 after a predetermined time has passed since the image data is input to the first encoder 100 and the encoding is performed.

The bit rate controller 120 sets the quantization coefficient of the second encoder 140 and delays a predetermined time compared to the first encoder 100 so as to set the quantization coefficient of the image data input to the second encoder 140. It controls to encode using.

In this case, the bit rate controller 120 uses the bit amount of the bit stream stored in the buffer 110 and the quantization coefficient set in the first encoder 100 to quantize each frame to be applied to the quantization of the second encoder 140. Calculate the coefficients.

Specifically, in setting the quantization coefficient of the second encoder 140, the bit rate control unit 120 includes the bit amount and the buffer 110 of the bitstream until immediately before the quantization step of the frame currently input to the second encoder 140. The quantization coefficient of the frame currently input to the second encoder 140 is calculated by using the bit amount of the bitstream of the frame corresponding to the current frame stored in the " That is, the quantization coefficient of the current frame to be applied by the second encoder 140 is calculated using the quantization coefficient value set in the first encoder 100 to obtain the bit amount of the bitstream of the corresponding frame stored in the buffer 110. do.

The input delay unit 130 processes the original image data to be input to the second encoder 140 after a predetermined time elapses from the time when the original image data is input to the first encoder 100. Here, the predetermined time is preferably longer than the time taken for the at least one frame to be encoded through the first encoder 100 in the original image data consisting of a plurality of frames.

 The input delay unit 130 according to an embodiment of the present invention temporarily stores the input original image data by the FIFO method, and when a predetermined time elapses, the second encoder 140 sequentially stores the temporarily stored image data from the first frame. )

For example, when the original image data is composed of n frames, the input delay unit 130 temporarily stores the image data up to the i-th frame for a predetermined time from the time when the first frame of the original image data is input. Stored. In this case, the input delay unit 130 has a capacity of at least i capacity for storing image data of at least i frames.

The input delay unit 130 outputs the temporarily stored image data to the second encoder 140 sequentially from the first frame when a predetermined time comes. At this time, the input delay unit 130 sequentially outputs the temporarily stored image data from the first frame to the second encoder 140 and sequentially receives from the i + 1 th frame to the n th frame of the original image data. Temporarily save. The input delay unit 130 outputs the i-th frame to the second encoder 140 sequentially from the i + 1 th frame to the n th frame. As a result, the time point at which each frame of the original image data is input to the second encoder 140 is delayed by the time that i frames are input as compared to the first encoder 100.

In addition, the input delay unit 130 may include a timer (not shown) for outputting the original image data to the second encoder 140 after a predetermined time has elapsed since the input of the original image data.

The second encoder 140 compresses the input image data and encodes the image data into a bitstream format. In this case, the second encoder 140 quantizes the input image data by applying the quantization coefficient for each frame calculated and set by the bit rate controller 120. As such, the bitstream of the image data encoded by the second encoder 140 is transmitted through a communication channel on a network or transmitted to a storage medium.

2 is a block diagram illustrating a bit rate controller according to an exemplary embodiment of the present invention.

As shown in FIG. 2, the bit rate controller 120 includes a bit amount detection module 121, a quantization coefficient generation module 122, and a memory 123.

The bit amount detection module 121 detects a bit stream of image data stored in the buffer 110, extracts a bit amount of each frame, and transfers the extracted bit amount of each frame to the quantization coefficient generation module 122. . At this time, the bit amount detection module 121 calculates the bit amount by detecting the bit stream of the image data stored in the buffer 110 in the FIFO method on a frame basis in real time. In addition, the bit amount detection module 121 may sequentially store the calculated bit amount for each frame in a memory (not shown).

Meanwhile, the bit amount detection module 121 may calculate the bit amount by detecting the bit stream when the bit stream stored in the buffer 110 becomes more than a predetermined amount. That is, the bit amount detection module 121 may sequentially detect the bit-by-frame bitstream after the bitstream is stored in the buffer 110 for a predetermined time.

Hereinafter, for convenience of description, the bit amount of each frame calculated by the bit amount detection module 121 is referred to as a 'first encoding result value'.

The quantization coefficient generation module 122 compares the first encoding result value for each frame with the reference bit amount, respectively, and resets the quantization coefficient of the frame having a bit amount exceeding the reference bit amount. The quantization coefficient generation module 122 generates a table of frame-specific quantization coefficients including the reset quantization coefficients and stores them in the memory 123.

Specifically, the quantization coefficient generation module 122 may determine the quantization coefficient to be applied by the second encoder 140 in the case of a frame in which the first encoding result value for each frame transmitted from the bit amount detection module 121 exceeds the reference bit amount. Calculate.

That is, the quantization coefficient generation module 122 checks the first encoding result value for each frame transmitted from the bit amount detection module 121, and then encodes a first encoding of a frame corresponding to the frame currently input to the second encoder 140. The resulting value is compared with the reference bit amount. When the first encoding result exceeds the reference bit amount as a result of comparing with the reference bit amount, the quantization coefficient generation module 122 determines the bit amount of the bit stream from the second encoder 140 until immediately before the quantization step of the current frame. And a quantization coefficient of a frame currently input to the second encoder 140 using the quantization coefficient applied by the first encoder 100 to obtain a first encoding result.

Meanwhile, if the first encoding result of the frame currently input to the second encoder 140 is less than or equal to the reference bit amount, the quantization coefficient generating module 122 may convert the quantization coefficient applied to the corresponding frame from the first encoder 100 to the second. The same may be applied to the encoder 140.

That is, the quantization coefficient generation module 122 generates a quantization coefficient table by calculating and setting the quantization coefficient of the second encoder 140 with reference to the frame-specific quantization coefficient applied to the quantization in the first encoder 100.

In this case, the quantization coefficient of the frame in which the first encoding result calculated by the quantization coefficient generation module 122 exceeds the reference bit amount is set to a value larger than the quantization coefficient applied to the frame in the first encoder 100. . That is, the final bit amount of the frame is calculated to satisfy the target bit amount (or less than the reference bit amount).

Here, the reference bit amount means a maximum bit amount that does not exceed a range of bit amounts that may be hostile to the bandwidth required for image data to be transmitted through a communication channel or transmitted to a transmission medium. In general, when the bit amount of the image data to be transmitted exceeds the maximum bit amount, the data to be transmitted is lost or a transmission error occurs, thereby degrading the quality of the multimedia service.

Accordingly, the quantization coefficient generation module 122 according to the embodiment of the present invention calculates the bit amount of a specific frame having a bit amount exceeding the reference bit amount to be equal to or less than the reference bit amount.

The memory 123 stores the quantization coefficient for each frame set by the quantization coefficient generation module 122. In this case, the quantization coefficient set in the quantization coefficient generation module 122 is applied to the quantization of the second encoder 140, and may be stored in a table form in which the quantization coefficient set according to each frame is matched.

Meanwhile, in the embodiment of the present invention, the input delay unit 130 outputs the image data to the second encoder 140 when a predetermined time elapses after the first frame of the original image data is input to the first encoder 100. It was shown. However, in another embodiment, the bit rate controller 120 may include a timing control module (not shown), and may instruct the input delay unit 130 to output image data after a predetermined time elapses.

That is, the timing control module (not shown) may instruct the output delay unit 130 to output the image data when the bitstream of at least one frame is stored in the buffer 110.

3 is a block diagram illustrating a configuration of a first encoder illustrated in FIG. 1.

As shown in FIG. 3, the first encoder 100 includes a transformer 101, a quantizer 102, a coder 103, an inverse quantizer 104, an inverse transformer 105, a motion estimator 106, and a motion. Compensator 107 is included.

In the exemplary embodiment of the present invention, the original image data input to the first encoder 100 is divided into macroblock units every frame and input to the converter 101.

The transformer 101 receives a result of the operation 108 of the motion compensated image data and the input current original image data, and performs a discrete cosine transform (hereinafter, referred to as 'DCT'). At this time, the motion compensated image data is obtained by referring to the image data of the previous frame.

The quantizer 102 quantizes the image data passed through the DCT. In this case, the quantizer 102 performs quantization by applying a quantization coefficient set every frame. That is, although not shown in FIG. 3, the quantizer 102 may separately include a memory (not shown) that stores a table of quantization coefficients for each frame. In this case, the quantization coefficient may be fixed to a predetermined value or updated and stored as an optimized value.

The coder 103 converts the image data into a bitstream by performing entropy coding to improve compression efficiency of the quantized image data. At this time, the bitstream for each frame that is a result of entropy coding of the coder 103 is stored in the buffer 110 in the FIFO method.

Inverse quantizer 104 dequantizes quantized image data.

Also, the inverse transformer 105 restores the image data by performing inverse DCT transformation on the inversely quantized image data.

In this case, the image data of the current frame reconstructed by the inverse transformer 105 and the motion compensated image data of the previous frame are combined 109 so that the image data is output to the motion estimator 106 and the motion compensator 107. .

The motion estimator 106 receives the resultant image data combined with the original input image data 109 and outputs a motion vector and a differential image for motion prediction.

The motion compensator 107 receives the differential image output through the combination image 109 and the motion estimator 106 to compensate for the motion.

4 is a block diagram illustrating a configuration of a second encoder illustrated in FIG. 1.

The second encoder 140 according to an embodiment of the present invention includes a transformer 141, a quantizer 142, a coder 143, an inverse quantizer 144, an inverse transformer 145, a motion estimator 146, and a motion. Compensator 147.

In this case, each configuration and operation of the second encoder 140 are identical to those of the first encoder 100 except that the quantization coefficient for each frame referred to for application to quantization in the quantizer 142 illustrated in FIG. 4 is different. Therefore, duplicate descriptions will be omitted.

The converter 141 receives a result of the operation 148 of the current original image data and the motion compensated image data output from the input delay unit 130 and performs DCT. In this case, the input original image data is input after a predetermined time delay from the time when the original image data of the same frame is input to the first encoder 100.

The quantizer 142 quantizes the image data passed through the DCT. In this case, the quantizer 142 refers to a quantization coefficient table stored in the memory 123 in quantizing image data of each input frame.

That is, the quantizer 142 may be configured to reset the quantization coefficient that is relatively larger than the quantization coefficient applied to the specific frame by the quantizer 102 of the first encoder 100 in the case of a specific frame in which the amount of bits per frame exceeds the reference bit amount. Apply to perform quantization. As a result, the second encoder 140 may perform encoding by predicting a bit rate of image data of a frame to be currently encoded.

The remaining components of the second encoder 140 perform the same operations as the corresponding components in the first encoder 100.

Hereinafter, a multipath encoding method according to an embodiment of the present invention will be described with reference to FIGS. 5 and 6. Hereinafter, an encoding process through the first encoder 100 is referred to as a first encoding, and an encoding process through the second encoder 140 is referred to as a second encoding.

5 is a flowchart illustrating a method of generating a quantization coefficient according to an embodiment of the present invention.

First, the bit amount detection module 121 of the bit rate controller 120 detects the bit amount for each frame from the buffer 110 (S200). In this case, the bitstream stored in the buffer 110 is a bitstream of at least one frame obtained through the first encoding, and is stored in the FIFO method in the case of a plurality of frames. Therefore, the bit amount detection module 121 may sequentially detect the bit amount for each frame obtained through the first encoding from the first encoded frame.

Next, the quantization coefficient generation module 122 of the bit rate controller 120 compares the detected bit amount of each frame with a preset reference bit amount (S210). At this time, the detected bit amount for each frame is sequentially compared with the reference bit amount. Here, the reference bit amount is the maximum bit amount that can be stably provided to the multimedia service within the bandwidth of the communication channel.

In operation S220, the quantization coefficient generation module 122 sets the quantization coefficient of the corresponding frame to a preset quantization coefficient in the case of a frame in which the detected bit amount is less than or equal to the reference bit amount. In this case, the preset quantization coefficient may be the same as the quantization coefficient applied in the first encoding process, or may be set to a reference quantization coefficient or an average value of the quantization coefficient.

In addition, when the detected frame bit amount exceeds the reference bit amount, the quantization coefficient generation module 122 resets the quantization coefficient for achieving the target bit amount (S230). In this case, the target bit amount is a bit amount less than or equal to the reference bit amount, and is a bit amount that can stably load video data in the bandwidth of the communication channel. In addition, the quantization coefficient to be reset is set to a relatively larger value than the quantization coefficient applied in the first encoding process.

Next, the quantization coefficient generation module 122 generates a quantization coefficient table to be applied to the second encoding by using the respective quantization coefficient values set in the two previous steps S220 or S230 and stores the quantization coefficient table in the memory 123 (S240). In this case, a quantization coefficient for each frame is stored in the quantization coefficient table, which can be updated in real time.

Thereafter, the quantization of the second encoder 140 is performed with reference to the quantization coefficient table generated as described above.

6 is a flowchart illustrating a multi-pass encoding method according to an embodiment of the present invention.

First, original image data is simultaneously input to the first encoder 100 and the input delay unit 130 (S300). In this case, the input delay unit 130 temporarily stores the input image data sequentially for each frame.

Next, the first encoder 100 encodes the input image data and sequentially stores the input image data in the buffer 110 (S310). At this time, the first encoder 100 stores the encoded bitstream from the first frame to the last frame in the FIFO method.

Thereafter, the bit rate controller 120 sets and stores a quantization coefficient to be applied to the second encoder 140 (S320). In this case, the bit rate controller 120 detects the bit amount of each frame of the encoded bitstream stored in the buffer 110 and calculates a quantization coefficient to be applied to the second encoder 140. In addition, the bit rate controller 120 sequentially detects the bit amount of each frame stored in the buffer 110 from the first encoded frame and sets the quantization coefficient in real time.

Then, the input delay unit 130 outputs the image data to the input delay unit 130 after a predetermined time elapses from the input time of the first frame of the original image data (S330). In this case, the predetermined time is a time that the bitrate controller 120 can detect the bitstream of at least one frame from the buffer 110.

In this case, the input delay unit 130 outputs the image data stored for a predetermined time to the second encoder 140 and receives the original image data sequentially and stores the image data.

For example, when the input delay unit 130 receives an image output instruction from the bitrate controller 120 while storing the first frame to the i th frame of the original image data input for a predetermined time, the input delay unit 130 sequentially starts from the first frame. And output to the second encoder 140. At the same time, the input delay unit 130 continuously receives and stores image data of the i + 1 th frame. That is, when the input delay unit 130 outputs the first frame of the stored image data to the second encoder 140, the first encoder 100 and the input delay unit 130 are i + 1 th of the original image data. Take a frame.

Then, the second encoder 140 encodes the input image data by applying the quantization coefficient stored in the memory 123 to the quantization step (S340).

In this case, each frame of the image data encoded by the second encoder 100 is encoded with a predetermined time delay from the time of encoding through the first encoder 100. That is, the first encoder 100 may predict the information necessary for encoding the second encoder 140 through the frame bit amount encoded before the predetermined time. Therefore, according to an embodiment of the present invention, while controlling the bit amount for each frame of the input image data, the second encoding result can be obtained after a predetermined time from the first encoding time, thereby reducing the time required for multi-pass encoding. Can be.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

1 is a block diagram of a multi-pass encoding apparatus according to an embodiment of the present invention.

2 is a block diagram illustrating a bit rate controller according to an exemplary embodiment of the present invention.

3 is a block diagram illustrating a configuration of a first encoder illustrated in FIG. 1.

4 is a block diagram illustrating a configuration of a second encoder illustrated in FIG. 1.

5 is a flowchart illustrating a method of generating a quantization coefficient according to an embodiment of the present invention.

6 is a flowchart illustrating a multi-pass encoding method according to an embodiment of the present invention.

Claims (15)

In the device for multi-pass encoding the original image data consisting of a plurality of frames, A first encoder receiving the original image data and performing a first encoding; A bit rate controller configured to set a quantization coefficient for each frame according to the amount of bits for each frame detected from the first encoding result; An input delay unit configured to output the original image data by delaying a first time from a time point at which the first encoding starts; And  A second encoder that performs second encoding on the image data output from the input delay unit by applying the set quantization coefficient Multipath encoding apparatus comprising a. The method of claim 1, A buffer for storing the bitstream generated through the first encoding Multipath encoding apparatus further comprising. The method of claim 2, In the buffer, The bitstream for each frame is stored in a first in-first out (FIFO) manner. Multi-pass encoding device, characterized in that. The method of claim 2, The first time is a period shorter than a period during which the first encoding of the entire plurality of frames is performed. Multipass Encoding Device. The method according to any one of claims 2 to 4, The bitrate control unit, A bit amount detection module for detecting a bit amount of each frame by using the bit stream; A quantization coefficient generation module for comparing the detected bit quantity with a reference bit quantity and setting the quantization coefficient for each frame according to the result; And Memory for storing the quantization coefficient for each frame Multi-pass encoding device comprising a. The method of claim 5, The bitrate control unit, A timing control module for instructing the input delay unit to output the image data after the first time has elapsed since the first encoding has started; Multi-pass encoding device further comprising. The method of claim 5, The quantization coefficient generation module, In the case of the frame in which the detected bit amount exceeds the reference bit amount, the quantization coefficient is set as a quantization coefficient to satisfy a target bit amount, In the case of the frame in which the detected bit amount is less than or equal to the reference bit amount, the quantization coefficient is set to a preset quantization coefficient. Multipass Encoding Device. The method of claim 7, wherein The quantization coefficient generation module, In the case of the frame in which the detected bit amount exceeds the reference bit amount, the quantization coefficient is calculated by using the quantization coefficient applied during the first encoding and the detected bit amount. Multipass Encoding Device. In the method of multi-pass encoding the original image data consisting of a plurality of frames, Detecting a bit amount of each frame by using a bitstream generated as a result of first encoding the original image data sequentially from a first frame; Comparing the bit amount for each frame and the reference bit amount, and setting and storing a quantization coefficient of each frame according to the comparison result; And After the first time elapses from the time when the first frame starts to be encoded, applying the set quantization coefficient to second encoding the original image data sequentially from the first frame Multi-pass encoding method comprising a. The method of claim 9, The first time is, A period shorter than a first encoding period of the entire frame of the original image data Multipass Encoding Method. The method of claim 10, Before the step of detecting the bit amount per frame, Storing the bitstream generated as a result of the first encoding in a first in-first out (FIFO) manner Multi-pass encoding method further comprising. The method of claim 11, The first time is, A period of time until a bitstream of at least one frame of the plurality of frames is stored as a result of the first encoding; Multi-pass encoding method. The method of claim 10, Setting and storing the quantization coefficient, Setting a quantization coefficient by calculating a quantization coefficient value such that the bit amount of the i-th frame satisfies a target bit amount when the bit amount of the i-th frame of the plurality of frames exceeds the reference bit amount; Setting a quantization coefficient to a preset value when the bit amount of the i-th frame is less than or equal to the reference bit amount; And Storing the set quantization coefficient for each frame Multi-pass encoding method comprising a. The method of claim 13, If the bit amount of the i-th frame exceeds the reference bit amount, The quantization coefficient is set to be greater than the quantization coefficient of the i-th frame applied during the first encoding. Multipass Encoding Method. The method of claim 13, When the bit amount of the i-th frame is equal to or less than the reference bit amount, The quantization coefficient is set to the same value as the quantization coefficient applied to the i th frame in the first encoding Multi-pass encoding method.

Images