JPH0514409A - Picture packet processing unit - Google Patents

Abstract

PURPOSE:To prevent fluctuation of picture quality without any timewise fluctuation of distortion of a picture. CONSTITUTION:A moving picture coding section 11 sends a moving picture coding signal to a high order packet composition means 14 based on a moving picture signal and sends a coding information quantity signal to a generated information quantity integration means 12. The generated information quantity integration means integrates the information quantity produced in the unit of a picture frame for each prescribed time and a packet transmission control means 13 sends a control signal to the high order packet composition means based on the integrated value. The high order packet composition means 14 generates a packet of a moving picture use high order packet signal based on the control signal and the generated information quantity in the unit of the picture frame and a low order packet composition means 15 composes a low order packet signal based on the high order packet signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image packet processing apparatus for performing image packet transfer using a packet switching network for real-time communication such as an asynchronous transfer mode network.

[0002]

2. Description of the Related Art In the field of image communication, when image information is transmitted, the amount of generated image information is enormous and the redundancy thereof is large. Therefore, various band compression transmission systems or high efficiency encoding systems have been used. Is being considered. 11
FIG. 2 shows the configuration of such an image coding transmission side device. The apparatus shown in the figure includes a moving image coding unit 1, a smoothing buffer unit 2,
The encoding control unit 3 is included.

The moving picture coding section 1 is an apparatus for inputting and coding a moving picture signal of a normal television broadcast or the like, and the smoothing buffer section 2 is outputted from the moving picture coding section 1. An apparatus for smoothing a moving image coded signal and outputting an overflow / underflow signal of its buffer memory. The coding control unit 3 is a device that receives the overflow / underflow signal output from the smoothing buffer unit 2 and generates a coding control signal in accordance with this signal. In the image coding transmission side device configured as described above, when the moving image signal is input to the moving image coding unit 1, the moving image coding unit 1 is based on the coding control signal from the coding control unit 3. Outputs a moving image coded signal. Then, the moving image coded signal is smoothed by the smoothing buffer unit 2 and output to the fixed rate transmission line as an image coded transmission signal. In addition, the encoding control unit 3 performs feedback control on the moving image encoding unit 1 according to the overflow / and-flow signal of the buffer memory output from the smoothing buffer unit 2, whereby the moving image encoding unit 1 operates in the buffer memory. Overflow / underflow regulations are regulated.

[0004]

However, since the above-mentioned image coding transmission side device supports fixed rate transmission, for example, when the amount of generated information of a moving image coded signal temporarily increases, An overflow signal is generated in the buffer memory of the smoothing buffer unit 2, and the coding control by feedback control works. As this control method, for example, there is a method of limiting the amount of information generated from the moving picture coding unit 1 by roughening the quantization step. However, the coarse quantization increases the distortion of the image and deteriorates the image quality. That is, since the generated information amount of the moving image coded signal depending on the activity of the screen of the image is limited to a fixed rate, the distortion of the image temporally fluctuates, and there is a problem that the image quality fluctuates. ..

On the other hand, in recent years, high-speed packet switching and Asynchronous Transfer Mode (ATM) have been attracting attention as a method for solving such a problem. This high-speed packet network, asynchronous transfer mode (AT
M) Network means a packet transmission rate of 1.5Mbps to 150Mbps
Since it is faster and the communication protocol in the network is simplified, it is possible to perform communication of voice, image, etc., which requires real time. Further, since all kinds of information including data are handled in the same packet in a unified manner, it is possible to flexibly support multimedia communication. The asynchronous transfer mode is
This is a high-speed packet switching transfer mode named by CCITT (International Telegraph and Telephone Consultative Committee). Hereinafter, a packet network including this asynchronous transfer mode network will be defined and described as a high-speed packet network.

Further, as image transmission using such a high-speed packet network, for example, packet transmission of Yasuhiko Yasuda video, Television Society of Japan, Vol. 42, No. 6, June 1988, pp. 533-537. There is one described in. This publication describes, as packet transmission of video, burstiness of video signals, statistical multiplexing, and packet discard measures. However, even with such a technique, image packet transmission has not yet been put into practical use, and there has been a demand for a technique that solves the above-mentioned problem of image transmission at a fixed rate. The present invention has been made to solve the above problems, and an object of the present invention is to provide an image packet processing device capable of preventing a change in image quality due to a fixed rate.

[0007]

An image packet processing apparatus of the present invention integrates the amount of information generated per frame unit of an image in an image packet processing apparatus using a packet switching network for performing real-time communication, A generated information amount integrating means for outputting an information amount integrating signal corresponding to this integrated value, a predetermined operation is performed based on the information amount integrating signal, and an upper packet control signal and a lower packet control signal are output based on the operation result. The packet transmission control means and the moving image coded signal are input to generate a timing upper packet signal with reference to an image frame synchronization signal, and the upper packet signal subsequent to the timing upper packet signal is used as the upper packet signal. Upper packet for outputting a moving image upper packet signal according to the amount of generated information of the image based on the upper packet control signal A standing section, enter the upper packet signal, is characterized in that a lower packet assembling means for outputting a lower packet signal based on the low-order packet control signal.

[0008]

In the image packet processing apparatus of the present invention, the moving picture coding section sends the moving picture coded signal to the upper packet assembling means based on the moving picture signal, and the coded information quantity signal is the generated information quantity integrating means. To send to. The generated information amount integrating means integrates the amount of information generated per frame unit of the image at predetermined time intervals, and the packet transmission control means sends a control signal to the upper packet assembling means based on the integrated value. The upper packet assembling unit generates a packet of the moving image upper packet signal based on the generated information amount of the image frame unit by the control signal, and the lower packet assembling unit,
A lower packet signal is constructed based on this upper packet signal. Therefore, the image distortion does not change with time, and the image quality can be prevented from changing.

[0009]

Embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of an image packet processing device of the present invention. The apparatus shown in the figure comprises a moving picture coding unit 11, a generated information amount accumulating unit 12, a packet transmission control unit 13, an upper packet assembling unit 14, and a lower packet assembling unit 15.

The moving picture coding unit 11 has a function of receiving a moving picture signal and outputting a moving picture coding signal in which the redundancy of the picture is band-compressed by using the high efficiency coding method of the moving picture. There is. Examples of this high-efficiency coding method include intra-frame coding and inter-frame coding. In addition, here
In the case of a moving picture television signal,
It corresponds to a still screen that is continuous at regular intervals along the time axis.
The intra-frame encoding is a process of performing an encoding process on a pixel block in an image frame, and is a process of estimating a target pixel from adjacent pixels, for example. In addition, the inter-frame encoding is a process of performing an encoding process on a pixel block between adjacent image frames. For example, a process of analogizing a certain pixel block from the pixel blocks of the image frames before and after it is performed. It is a thing. In general, inter-frame coding has a greater reduction rate of the coding rate due to compression than intra-frame coding.

The generated information amount accumulating means 12 receives the coded information amount signal from the moving picture coding unit 11, accumulates the information amount generated per image frame at every predetermined time T seconds, and the resulting information. It has a function of outputting a quantity integration signal. The packet transmission control means 13 inputs the output information amount integrated signal, performs a predetermined calculation based on the information amount integrated value, and outputs the upper packet control signal and the lower packet control signal, respectively, to the upper packet assembly means 14. It also has a function of outputting to the lower packet assembling means 15.
The upper packet assembling unit 14 receives the moving image coded signal output from the moving image coding unit 11, generates a timing upper packet signal with the image synchronization signal as a reference, and outputs the timing upper packet signal as the timing upper packet signal. As a succeeding upper packet signal, it has a function of outputting a moving image upper packet signal based on the upper packet control signal output from the packet transmission control means 13. The lower packet assembling means 15 inputs the upper packet signal from the upper packet assembling means 14, and the packet transmission control means 1
3 has a function of outputting the lower packet signal as an image packet transmission signal to the variable rate transmission line based on the lower packet control signal.

FIG. 2 shows a specific structure of the upper packet assembling means 14. That is, the upper packet assembling unit 14 is composed of the moving image information buffer memory 14a, the header information buffer memory 14b, the timing control unit 14c, and the multiplexing unit 14d. Moving image information buffer memory 14a
Is a FIFO (first-in first-fast
Out) Memory for writing according to the rules. The header information buffer memory 14b includes the timing control unit 1
Frame synchronization number (FS) output from 4c, which will be described later.
And a byte length (LB) as the information length of the moving image information. Further, the timing control unit 14c sends the moving image information buffer memory 14a and the header information buffer memory 14b to the moving image information buffer memory 14a and the header information buffer memory 14b based on the moving image encoded signal from the moving image encoding unit 11 and the upper packet control signal from the packet transmission control unit 13. , A write signal (W), a read signal (R), and a header information signal including a frame synchronization number (FS) and a byte length (LB) to the header information memory 14b. There is. Furthermore, the multiplexing unit 14d
Has a function of generating an upper packet signal from the moving image information of the moving image information buffer memory 14a and the header information of the header information buffer memory 14b.

Next, the operation of the packet processing device having the above configuration will be described. First, when a moving picture signal is input to the moving picture coding unit 11, the moving picture coding unit 11 sends the moving picture coding signal to the higher-order packet assembling means 14 and also the generated information amount integrating means 12 A coded information amount signal is sent to. The generated information amount integration means 12 integrates the information amount generated per image frame from the input coded information amount signal every T seconds, and sends the obtained information amount integration signal to the packet transmission control means 13. .. Then, the packet transmission control means 13 sends out a predetermined upper packet control signal and lower predetermined packet control signal based on the input information amount integration signal.

FIG. 3 shows an example of the change over time in the amount of coded information generated. As shown in the figure, the amount of information generated per frame changes with time. Here, the generated information amount integration means 12 outputs an integrated value, which is the accumulation of the information amount from the first frame to the nth frame corresponding to T seconds, as an information amount integration signal at the end of the nth frame. FIG. 4 shows an example of the change over time of the information amount integrated value.

Next, the packet transmission control means 13 obtains the average information amount A (ave) n per frame by dividing the information amount integrated value by the number of frames n. On the other hand, the amount of generated information per frame has a temporal change as shown in FIG. FIG. 5 shows the generated information amount and the average information amount A for each frame.
It is a figure which shows the relationship with (ave) n. Now, the m-th 1
If the generated information amount per frame is Gm, this generated information amount changes with time as shown in FIG. Then, the packet transmission control means 13 uses the average information amount A (ave) n as a threshold value and determines the generated information amount Gm as the average information amount A (ave) n.
The following time is the normal state, the generated information amount Gm is the average information amount A (a
ve) When n or more, the overload state is set, and the control signal according to this load state is transmitted. Regarding the average information amount A (ave) n, when the value of T seconds is decreased, it shows an average for a short time, and when the value of T seconds is increased, it shows an average for a long time. And, the long-time average shows a gradual time variation.

On the other hand, in the upper packet assembling means 14, the moving image coded signal is transferred to the moving image information buffer memory 14 thereof.
Written to a. That is, the moving image coded signal generates different amount of information for each frame of the image, but the generated information is written in the moving image information buffer memory 14a for each frame according to the FIFO rule. At the same time, the timing control unit 14c writes the frame synchronization number (FS) obtained by converting the frame synchronization signal (FSYNC) as the image timing signal for the header into the header information buffer memory 14b. Here, the frame synchronization number (FS) is 3 image frames per second.
The numbers 0 to 29 corresponding to 0 are cyclically repeated. Further, the byte length (LB) as the information length of the moving image information amount is written in the header information memory 14b.

FIG. 6 shows the format structure of the upper packet. The upper packet for timing is a header (HR)
Has a frame synchronization number (FS) and is regularly transmitted in frame cycle units, and is set to have the same length as the payload length of a lower packet which is a fixed length packet. The moving image upper packet is a variable length packet that has a frame synchronization number (FS) and a byte length (LB) as a header (HR) and follows the timing upper packet, and has a byte length (LB). The valid data of the moving picture coded information corresponding to

FIG. 7 shows the relationship between the amount of generated information and the generation of the upper packet signal. Further, FIG. 8 shows a flowchart of packet transmission control. First, the packet transmission control means 1
3 compares the generated information amount Gm and the average information amount A (ave) n with each other (step S1). If Gm <A (ave) n, the normal packet state is determined and the upper packet control in the normal state is performed. A signal is transmitted (step S2). Also, Gm ≧ A (ave)
If it is n, it is determined to be in an overloaded state, and the upper packet control signal in the overloaded state is transmitted (step S3). The timing control unit 14c receives the upper packet control signal from the packet transmission control unit 13, and if this is the control signal in the normal state, it matches the read timing of the moving image information buffer memory 14a with the synchronization signal of the next frame. Control (step S4). On the other hand, in the overload state, the read timing is controlled to a predetermined timing position earlier than the frame synchronization signal of the next frame (however, this position depends on the threshold value of the average information amount A (ave) n). Yes (step S5). In this way, by executing the upper packet control, that is, the read timing control of the buffer memory, the image packet signal sent to the transmission path is smoothed.

Next, the relationship between the upper packet and the lower packet will be described. FIG. 9 shows a timing chart of the upper packet signal and the lower packet signal. The image frame synchronization signal (FSYNC) is generated at 33 and 33 ms cycles, assuming that 30 frames per second are used. In synchronization with this frame synchronization signal, the timing upper packet from the beginning,
Then, a moving image upper packet is generated. Here, the entire length of the upper packet for timing is, as described above,
It is set to be equal to the payload length of the lower packet. That is, by setting the length of the timing upper packet to be equal to the payload length of the lower packet having a fixed length, there is an effect that the hardware load on the lower packet assembling unit 15 is reduced. The moving image upper packet, which is a variable length packet, is decomposed in units of the payload length of the fixed length lower packet. That is, since the data length (PV) of the moving image upper packet changes in units of frames, for example, in the nth frame, there are three lower packets decomposed from the moving image upper packet, and in the (n + 1) th frame, the data ( It is two according to PV). In the figure, PT indicates the data of the upper packet for timing, and hr indicates the header of the lower packet signal.

FIG. 10 shows the format structure of the lower packet. That is, the lower packet is a fixed-length packet as described above, and is composed of a header (hr) and a payload for carrying the information of the upper packet. Also,
As a configuration example of the header, a sequence number (SN), an information type (IT), and a valid information length (LI) are used. The sequence number (SN) indicates the order of packets. For example, when 4 bits are assigned, 0 → 1 → 2 → → →
Repeat 0 to 15 cyclically, such as 14 → 15 → 0 → 1 → 2 → ... This is used on the receiving side to detect packet discards and erroneous insertions. Information type (IT)
Indicates the type of information. For example, when 2 bits are allocated, 00: timing, 01: moving image, 10: voice, 11: spare can be set. Further, the effective information length (LI) indicates the effective byte length in the payload, and when 6 bits are allocated to this, for example, an information amount of up to 64 bytes can be expressed.

As described above, in the above embodiment, the load state is judged based on the generated information amount, and the timing control of the upper packet is performed according to the load state.
There is an effect that the concentration of the number of packets can be prevented, the statistical multiplexing effect can be further utilized when transmitting packets through the transmission line of the network, and the network resources can be efficiently used.

[0022]

As described above, according to the image packet processing device of the present invention, in the packet transmission side device using the packet network for performing real-time communication, the amount of generated information for each frame of the moving image code signal is determined. Since the packet of the high-order packet signal for moving images is configured based on this, there is no temporal change in image distortion, the image quality is prevented from changing, and the characteristics of the packet network that performs real-time communication are fully exhibited. You can

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an image packet processing device of the present invention.

FIG. 2 is a diagram showing a configuration of a higher-order packet assembling unit of the image packet processing device of the present invention.

FIG. 3 is an explanatory diagram of a temporal change in the amount of coded information generated by the image packet processing device of the present invention.

FIG. 4 is an explanatory diagram of a time change of an information amount integrated value of the image packet processing device of the present invention.

FIG. 5 is a diagram showing the relationship between the amount of generated information and the average amount of information in the image packet processing device of the present invention.

FIG. 6 is a diagram showing a format configuration of an upper packet of the image packet processing device of the present invention.

FIG. 7 is an explanatory diagram of the correlation between the amount of generated information and the upper packet signal of the image packet processing device of the present invention.

FIG. 8 is a flowchart of upper packet transmission control of the image packet processing device of the present invention.

FIG. 9 is a timing chart of an upper packet signal and a lower packet signal of the image packet processing device of the present invention.

FIG. 10 is a diagram showing a format configuration of a lower packet of the image packet processing device of the present invention.

FIG. 11 is a block diagram showing a configuration of a conventional fixed rate image coding transmission side apparatus.

[Explanation of symbols]

 11 Video Coding Unit 12 Generated Information Amount Accumulating Means 13 Packet Transmission Controlling Means 14 Upper Packet Assembling Means 15 Lower Packet Assembling Means

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Hiroshi Imai 1-7-12 Toranomon, Minato-ku, Tokyo Oki Electric Industry Co., Ltd.

Claims (1)

Claim: What is claimed is: 1. An image packet processing device using a packet-switching network for real-time communication, wherein the amount of information generated per frame unit of an image is integrated, and information corresponding to the integrated value is accumulated. A generated information amount integrating means for outputting an amount integrated signal, a packet transmission control means for performing a predetermined operation based on the information amount integrated signal, and outputting an upper packet control signal and a lower packet control signal based on the operation result, A moving image coded signal is input to generate a timing upper packet signal based on a frame synchronization signal of an image, and the upper packet signal subsequent to the timing upper packet signal is generated based on the upper packet control signal. Upper packet assembling means for outputting a moving image upper packet signal according to the amount of information generated in the image, and the upper packet Receives the signal, the image packet processing device being characterized in that a lower packet assembling means for outputting a lower packet signal based on the low-order packet control signal.