JPH02305284A - Variable transmission speed picture encoding device - Google Patents

Abstract

PURPOSE:To improve the picture quality and to efficiently operate a line by integrating the generated information volume due to encoding of each one frame for a prescribed time by an integrating means and applying the feedback control based on the controlled variable obtained from the integrated volume to control the generated information volume of the next frame. CONSTITUTION:A one-frame generated information volume calculating means 4 calculates a generated information volume 5 of encoded data 3 of one frame which is already encoded. In a means 6 to integrate the generated information volume for t-number of seconds, the integrated volume is obtained by an adder 12 at each time of encoding. A generated information volume 7 for t-number of seconds is compared with a target generated information volume for t seconds by a target value/generated value comparing means 8 to output a controlled variable 9. An encoding control part 10 controls the generated information volume in an encoding part 2 in accordance with the obtained controlled variable 9. Consequently, an influence of increase or reduction of the information volume in a short period upon the control is reduced to improve the picture quality, and the line is efficiently operated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a variable transmission rate image encoding device, and particularly to a variable transmission rate image encoding device that performs feedback control of the amount of generated information.

[Prior art] Figure 7 shows, for example, the Confucian technique IT85-61.1-8 (IL
85) It is a block diagram showing the coding control system of the conventional image coding device shown in Atsumi Murakami's "Image efficiency coding by vector quantization", and in the figure, (1) is the target of coding. (2) is an encoding unit that encodes the input video signal.

(3) is the encoded data output from the encoder (2),
(4) is a one-frame generated information amount calculating means for calculating the information amount of encoded data for one encoded frame.

(5) is the amount of information generated in one frame outputted from the amount of information calculated in one frame (4), and (8) is the amount of information generated in one frame that is output from the calculation means (4). This is a target value/occurrence value comparison means that outputs a control amount (9).

(10) is an encoding control unit that controls the amount of information generated by encoding the next frame in the encoding unit (2) according to the control amount.

In addition, FIGS. 8(a) and (l) are diagrams showing an example of the above-mentioned "1 standard value/occurrence value comparison means (8), and in the figure (18
) is a target value of 1, (19) is a subtracter, and (20) is a control table that calculates control M (9) from the generated information amount (5) and target shift 1 (18) and outputs 1]J1.

Furthermore, FIG. 9 is a block diagram showing an example of the encoding section (2), in which (22) calculates the difference between the input video clavicle (1) and the preliminary Δ111 signal (35) and calculates the difference between the input video clavicle (1) and the pre-Δ111 signal (35). This is a subtracter that obtains the interframe signal.

(23) is a valid/invalid block identification/judgment unit that determines whether or not to quantize a focused area based on the magnitude of the inter-frame difference signal; (24) is a quantization encoder; and (25) is a quantization decoder.

(26) is an adder that adds the quantized decoded signal (33) and the prefllll signal (35), and (27) is a frame memory that stores the decoded image and generates the prefllll signal (35).

(28) is a multiplexing unit that multiplexes quantized data (32) from the quantized encoder (24), valid/invalid block identification information (30), encoding control information, etc. to obtain encoded data (3); (29) is a valid/invalid block identification threshold that controls the amount of generated information.

(31) is a quantization control signal that controls the amount of information generated in the quantization encoder (24), and (34) is an interframe decoded signal.

Next, the operation will be explained.

In FIG. 7, an input video signal (1
) is encoded for each frame (video frame) in the encoder (2), and the encoded data (3) is output, and the encoded data (3) is used for use or recorded on a fixed transmission rate line. Ru.

Note that a moving image can be decoded from the encoded data (3) by a decoding unit corresponding to the encoding algorithm of the encoding unit (2).

Since the encoded data of the image encoding device is usually constant, encoding control is required to keep the amount of information generated in the encoder (2) constant. Therefore, when encoding one predetermined frame, the amount of generated information (occurrence value) of the encoded data of one frame encoded one frame before that frame.
and the target amount of information (target value), and based on the result, control is performed to bring the amount of information generated by encoding the current frame as close to the target value as possible.

On the other hand, the one-frame generated information amount calculation means (4) calculates the generated information amount of one frame of encoded data (3) that has already been encoded.

Then, the determined one-frame generation information amount (5) is compared with a target value (18) by a target value/occurrence value comparing means (8).

Here, the target value is V (bit sec) for the line speed and N (1'ralie/se) for the target number of frames per second in normal fixed transmission rate video encoding.
e ), then the target prey bioinformation amount P (bit/fr
a river e) is for P = v / N (bit/fra
me).

= 4 - In the configuration shown in FIG. 8(a, ), the control amount (9) is simply determined by the difference between the generated value (7) and the target value (18).

In addition, in the configuration of FIG. 8(b), the control table (20)
Since it is constructed from a ROM or the like, the control characteristics can be improved more than the configuration shown in FIG. 8(a).

Then, the encoding control section (10) controls the amount of information generated in the encoding section (2) using a predetermined information amount control method (for example, controlling the increase/decrease of the valid/invalid block identification threshold).

That is, the input video signal (1) is subtracted by the predicted value (35) which is the decoding result of the previous frame in the subtracter (22), and becomes an interframe difference signal.

Further, the inter-frame difference signal is subjected to valid/invalid block identification judgment by the valid/invalid block identification judgment unit (23), that is, in the inter-frame difference signal that has been divided into blocks of a predetermined size, a large amount of signal remains in the inter-frame difference. It is determined that only the blocks that are present are valid and the quantization encoder (24
) and outputs quantized data (32), but for blocks determined to be invalid, only valid/invalid block identification determination information (30) indicating invalidity is output as encoded data.

Note that the validity/invalidity determination is made by comparing the valid/invalid block identification threshold (29) from the encoding control unit (10) with the absolute value average, standard deviation, average value, etc. of the signal of the inter-frame difference signal block. It will be done.

Then, the effective block is quantized by a quantization encoder (24), and is converted into quantized data (32) by a multiplexer (28).
) will be sent to. In addition, in the quantization encoder (24), the amount of information generated in the quantization encoder (24) is changed by changing the quantization characteristics such as the quantization step width using the quantization control signal (31) from the encoding control unit (10). can be controlled.

Then, valid/invalid block identification judgment information (30)
, quantization control signal (31) and other information is sent to the multiplexing unit (28).
The data is multiplexed by the encoder (2) and output as encoded data (3) to the fixed transmission rate line.

On the other hand, the next predicted signal (35) in interframe coding
In order to obtain quantization data (
32), and an adder (26) adds the quantized decoded signal (33) and the predicted signal (35) to convert it into an interframe decoded signal (34).

Further, the inter-frame decoded signal (34) is stored in the frame memory (27) and then referred to as the next prediction signal (35).

Furthermore, as shown in FIG. 10, the encoding control unit (10) controls the increase/decrease of the valid/invalid block identification threshold ΔTh in accordance with the control amount.

For example, if "25" is input as the control amount, a value obtained by increasing the previous threshold by "2" is set as the valid/invalid block identification threshold.

In this way, if the amount of information generated in the previous frame is low, the amount of information generated is increased in the next frame, and if the amount of information generated in the previous frame is high, the amount of information generated is decreased in the next frame.

In addition to the above-described control, the amount of generated information is often controlled by using a method of dropping frames.

Further, the above-described control can also be realized by software.

[Problems to be Solved by the Invention] In recent years, variable transmission rate transmission is about to be put into practical use in the field of moving image transmission.

The amount of information in a video signal depends on the amount of motion in the screen as long as currently commonly used interframe predictive coding is used. For this reason, with fixed transmission rate encoding, when the amount of movement in the screen increases, it is necessary to raise the valid/invalid block identification threshold, widen the quantization step width, add frame dropping, etc. There is no choice but to equalize the amount of information by degrading the image.

However, variable transmission rate encoding is known to be advantageous for video transmission because the line speed increases as video information increases, so there is no need to degrade image quality.

However, even if the encoding control shown in FIG. 7 is applied to the variable transmission rate, the amount of generated information will be as shown in FIG. 3 ( There is a problem in that the advantages of improving image quality due to variable transmission speeds are no longer realized.

In addition, in the configuration shown in FIG. 7, if the control period is made longer than every frame to an n-frame period or a t-second period, the fluctuation in the amount of generated information will increase, but the control will be delayed,
There was a problem that not only did this not lead to improvement in image quality, but also the ability to converge to the target amount of information was lost.

Furthermore, an example shown in FIG. 11 can be considered as control to maintain constant image quality using the effect of variable transmission speed.

However, the amount of generated information in this case is as shown in FIG. 3(b), and there is a problem in that the amount of generated information depends only on the movement of the image, and there is no feedback control of the amount of information. Note that (36) in FIG. 11 indicates a fixed control section.

This invention was made to solve the above problems, and by using a line with a variable transmission rate, the image quality is improved compared to conventional control using a line with a fixed transmission rate. The purpose is to obtain an image coding control method that enables efficient operation of the line by determining the target amount of generated information.

[Means for Solving the Problems] A variable transmission rate image encoding device according to the present invention includes an encoding section that encodes a signal, and one frame 13 of the encoded data encoded by the encoding section. a calculation means for calculating the amount of generated information for a given period of time; an integration means for calculating the cumulative amount of information generated within a predetermined period of time calculated by the calculation means; The apparatus includes a comparison means for determining a control amount by comparing the amount with a target amount of generated information, and a control section for controlling the encoding section so as to increase or decrease the amount of generated information of a frame to be encoded based on the control amount.

Each time the encoder encodes a signal frame by frame, the controller performs feedback control so that the amount of generated information within a predetermined time approaches the target amount of generated information.

"Operation" The variable transmission rate image encoding device according to the present invention integrates the amount of information generated by the encoding of each frame by the encoding unit over a predetermined period of time using the integrating means, and compares the accumulated amount with a target value using the comparing means. Based on the control amount obtained by doing this, the control unit applies feedback control to control the amount of information generated in the next frame.

[Example] An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an image encoding device in an embodiment of the present invention.

In FIG. 1, (1) is an input video signal of an image to be encoded, and (2) is an encoding unit-C that encodes the input video signal (1).

(3) is dog encoded data output from the encoder (2),
(4) is encoded data for one frame (3
) is a calculation means for calculating the amount of information generated in one frame.

(5) is the amount of information generated in one frame outputted from the amount of information generated in one frame calculation means (4), and (6) is the past predetermined time period (
This is an information amount accumulating means for accumulating the amount of information generated for t seconds (7).

(8) is an Ll target value/occurrence value comparison means that compares the amount of information generated for one second (7) and the target amount of information generated for t seconds and outputs the control amount (9) according to the result. (10) is an encoding control unit as a control unit that controls the amount of information generated in the next frame in the encoding unit (2) according to the control amount.

FIG. 2 is a diagram showing the configuration of the means (6) for accumulating the amount of information generated for t seconds in the present invention, in which (11) is a shift register that stores the amount of information generated for each frame, (12)
is an adder that adds the amount of information generated in one frame stored in the shift register for one second.

Next, the operation of this embodiment will be explained.

In Figure 1, a human-powered video signal (1
) is encoded for each frame (video frame) in the encoder (2), encoded data (3) is output, and the encoded data (3) is transmitted or recorded.

On the other hand, the 1-frame generated information amount calculation means (4) calculates the generated information amount (5) of the encoded data (3) of 1 frame that has already been encoded.

In the second generated information amount accumulating means (6), the one frame generated information amount (5) is manually inputted to a shift register (11), and the contents are shifted every time one frame is encoded.

Furthermore, the content of the shift register (11) is calculated by an adder (12) each time it is encoded.

The capacity of the shift I register (11) is set to the number of frames encoded per second, and the data before that is sequentially forgotten.

Then, the determined amount of information generated per second (7) is compared with the target amount of information generated per second in the target value/occurrence value comparing means (8), and the control amount (9) is set at 141.

Furthermore, the encoding control unit (10
) controls the amount of information generated in the encoder (2).

For example, when encoding a video signal input at time T, it is encoded using a frame that accumulates the amount of information generated per second, and an encoding control determined by the control amount obtained from the accumulated amount of information generated per second. As shown in FIG. 4, which shows the relationship between frames to be controlled, the information amount of encoded data generated between time T-t and time T is accumulated and controlled i31.
] Find 1n.

Then, by controlling the sign according to the obtained control amount,
The video signal input at time T is encoded. The change in the amount of information generated at this time is shown in FIG. 3(c).

Next, another embodiment of the present invention will be explained with reference to FIG.

FIG. 5 is a block diagram showing the configuration of the unit (6) for accumulating the amount of information generated per second. In the figure, (5) indicates the amount of information generated per frame, and (11) indicates the number of frames encoded per second. It is a shift register with capacity.

(1B) is the 1 read from the nft register (11)
Frame generation information amount (14) is a subtracter.

(15) is an adder, and (16) is a register.

The amount of information generated in one frame (5) has a capacity corresponding to the number of frames encoded in one second.
is memorized. Then, the amount of information generated in one frame (13) of the encoded frame t seconds ago is read as the oldest data from the register (11), and the subtracter 0
In step 4), it is subtracted from the latest amount of information generated in one frame (5).

Furthermore, the subtracted result is sent to the adder (15) as the previous frame? ] is added to the amount of information generated for t seconds obtained at the time of encoding to obtain the current amount of information generated for t seconds (7). The obtained amount of generated information per second (7) is sent to the 1-1 regular cruise/generated cruise comparison means (8), and is also stored in the register (16), and when converting the next frame into η, the previous frame It is used as one-second occurrence information ff1 (1-7) obtained at the time of encoding.

= 15 - In the above-mentioned embodiment, the amount of information generated per second is calculated by the amount of information generated per second (6) as 111, or the amount of information generated per second is divided by the time t. It is also possible to output the average generated information amount (unit time). In this case, the target value/occurrence value comparison means (8)
The 11 target value in is the target amount of generated information per second.

Furthermore, in the embodiments described above, the encoding control is realized by hardware, but the present invention is not limited to this, and it is also possible to realize it by software. The operation in this case is not shown in flowchart 1 in FIG. Note that the operation is the same as that shown in FIGS. 1 and 5.

Further, in the above-described embodiments, the target amount of information to be generated is assumed to be fixed, but it is not limited to this, and may be changed depending on the usage status of other media on the line.

[Effects of the Invention] As described above, according to the present invention, each time encoding is performed one frame at a time, the cumulative amount of information generated within a predetermined time - 16 = Based on the control amount obtained by comparing the amount of payment with a preset target amount of information generated, feedback control is performed so that the amount of information generated within a predetermined time approaches the target amount of information generated. JR+J control is less affected by short-term increases and decreases in the amount of information, and it is possible to improve image quality by taking advantage of the characteristics of the variable transmission speed line, and it is also possible to control the target amount of information generated. This enables efficient line operation.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an image encoding device according to an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of means for accumulating the amount of information generated for t seconds, and FIG. 3 is a block diagram showing the amount of information generated and time. FIG. 4 is a diagram illustrating the encoding control operation of the image encoding device in an embodiment of the present invention.
5 and 6 are diagrams showing other embodiments of the present invention, and FIG. 7 is a block diagram showing the configuration of a conventional image encoding device.
FIG. 8 is a block diagram showing the target value/generated value comparison means, FIG. 9 is a block diagram showing the configuration of the encoding section, FIG. 10 is a diagram explaining the operation of the encoding control section, and FIG. 11 is a block diagram showing the configuration of the encoding section. FIG. 8 is a block diagram showing an example of applying the image encoding device shown in FIG. 7 to a variable transmission rate line. In the figure, (2) is the encoding unit, (3) is the encoded data, and (4
) is the calculation means, (5) is the amount of information generated for one frame, (
6) is an integrating means, (7) is an integrated amount of information generated within a predetermined time, (8) is a comparing means, (9) is a control amount, and (10) is a control section. Note that the same reference numerals in the figures indicate the same or corresponding parts. Agent: Patent attorney Masuo Oiwa (2 others) Procedural amendment (voluntary) ], E'S I 10 indication Patent application Hei 1=1271
33 No. 2, Name of the invention Variable transmission speed image? ] No. Naturalization No. 3, Relationship with the case of the person making the amendment Patent applicant address 2-3 Marunouchi Sara-chome, Chiyoda-ku, Tokyo Name (601) Mitsubishi Electric Corporation Representative Moriya Shiki 4 Agent address 2-3-5 Marunouchi Sara-chome, Chiyoda-ku, Tokyo
, Claims and Detailed Description of the Invention of the specification to be amended. 6. Contents of amendment Claims An encoding section that encodes a signal, a calculation means that calculates the amount of information generated for one frame of the encoded data encoded by the encoding section, and a calculation means that an integration means for calculating the integrated amount of generated information to be calculated within a predetermined time; and a comparison that determines a control amount by comparing the integrated amount of generated information ■ with a target generated information amount set in advance according to the integrated amount. and a control unit that controls the encoding unit to increase or decrease the generated information amount of the frame subjected to η naturalization according to the control amount, and each time the encoding unit encodes the signal one frame at a time, A variable transmission rate image encoding device characterized in that a control unit performs feedback control so that the amount of generated information within a predetermined time approaches a target amount of generated information.

Claims (1)

[Claims] an encoding unit that encodes a signal; a calculation unit that calculates the amount of information generated for one frame of the encoded data encoded by the encoding unit; and an amount of information generated within a predetermined time calculated by the calculation unit. an integrating means for determining an integrated amount of information, a comparing means for determining a controlled amount by comparing the integrated amount of generated information with a target generated information amount set in advance according to the integrated amount, and encoding based on the controlled amount. and a control section that controls the encoding section to increase or decrease the amount of generated information of the frame that performs the processing, and each time the encoding section encodes the signal one frame at a time, the control section controls the amount of information generated within a predetermined time. A variable transmission rate image encoding device characterized by performing feedback control so that the amount of generated information approaches a target amount of generated information.