JP2001078182A - Image transmitter and image transmission method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image transmitter that can cope with even a high bit rate by providing a means that transmits only pictures of a specific kind to suppress disturbance in a received image caused by collision of network packets and realizing discrimination of an IBP picture (discrimination of a kind of a frame) used for that means by means of a simple discrimination method using initial set information without retrieving a header of a stream so as to relieve a load on a CPU. SOLUTION: An A/D converter 11 applies A/D conversion to a received video signal and an MPEG2 encoder LSI 12 compresses the A/D-converted signal. The compressed data are outputted to an FIFO memory 13 after the number of bytes of the data stored in the FIFO memory 13 is written in an internal register 12a of the encoder LSI 12. The data read from the FIFO memory 13 are given via a data bus 6 to an Ethernet circuit 7, where header information is given to the data, and the resulting data are transmitted to an Ethernet 2. Then the encoder LSI 12 outputs IPB pictures in a predetermined sequence, and a frame counter 16 counts the number of the outputted frames to discriminate the picture type. In the case that the Ethernet 2 is in congestion, the Ethernet circuit 7 transmits the picture to the Ethernet 2 while interleaving the B pictures.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for converting continuous stream data such as moving images into Ethernet or Fast Ethernet.
TECHNICAL FIELD The present invention relates to an image transmission device and an image transmission method for performing real-time transmission via a network such as a network device.

[0002]

2. Description of the Related Art A conventional image transmission apparatus for transmitting continuous stream data such as moving images captured by a camera in real time using a LAN such as Ethernet or Fast Ethernet will be described below. A conventional image transmission apparatus 1 converts an input video signal into an MP as shown in FIG.
An MPEG compression circuit 5 for compressing data according to the EG1 standard, and an Ethernet circuit 7 for receiving compressed data (elementary stream) from the MPEG compression circuit 5 via a data bus 6 and transmitting the data to an Ethernet (network) 2. . The MPEG compression circuit 5 mainly includes an A / D converter 11 for converting an analog input video signal into a digital signal, and a digital video signal output from the A / D converter 11
An MPEG1 encoder LSI 12 that compresses according to the PEG1 standard,
It comprises a FIFO memory 13 for temporarily storing moving image data compressed by the EG1 encoder LSI 12, a CPU 14 for controlling the entire MPEG compression circuit 5, and a DRAM 15 which is a memory used for operating the CPU 14. The compressed moving image data transmitted to the Ethernet 2 is received by the image receiving device 3, decompressed by MPEG1, and output to the display monitor 4 to display the moving image.

In the image transmission apparatus 1 having such a configuration, an input video signal is converted into an A / D signal by an A / D converter 11.
After the D conversion, the data is compressed by the MPEG1 encoder LSI 12 and output in the form of an elementary stream. Once this is FI
The data is held in the FO memory 13 and read out sequentially in a first-in first-out manner in accordance with a request from the CPU 14. FIFO memory 13
Is sent to the Ethernet circuit 7 via the data bus 6, where the Ethernet / IP
After being added with header information such as TCP, the packet is transmitted over the Ethernet 2.

When moving image data output from the image transmission apparatus 1 is transmitted to the image reception apparatus 3 via a network 2 to which a plurality of PCs such as a LAN are connected, the network 2 is congested and traffic is If the number of packets increases, data collision may occur and data may be lost. Therefore, there is a problem that the image displayed on the display monitor 4 is disturbed even when the data received by the image receiving device 3 is decoded.

[0005] In order to reduce such a phenomenon, a method of avoiding packet collision by reducing the amount of data to be transmitted has been considered. Generally, to reduce the data amount, there are methods such as lowering the compression rate of MPEG1 or thinning out frames. As an example of a method of thinning out a frame, a method of discarding a B picture or a P picture and transmitting only an I picture can be considered. In order to realize this, in the image transmission apparatus 1, the MPEG1 encoder LSI 12
It is necessary to have a function of retrieving picture information from the header of the elementary stream output from the PC and determining the picture coding type.

[0006] The operation of the image transmission apparatus 1 when such a picture coding type determination (determination of an IBP picture by header search) is processed by software will be described with reference to a flowchart shown in FIG.

In the MAIN routine shown in FIG.
Initialize the PEG1 encoder LSI (step 21),
Next, it is checked whether it was called from the Frame ISR (step 22). If there is no call (step 22 → N),
Loop and wait until there is a call. Where Frame
ISR is DRAM called by hardware interrupt
15 is a program stored in the program. When called by the Frame ISR (step 22 → Y), the data is read from the I picture buffer in which the data of the I picture is stored, and a header such as Ethernet, IP, or TCP is added (step 23), and each packet is To the Ethernet 2 (step 24). Thereafter, the flow returns to step 22 until there is no more data to be transmitted,
To 24 are repeated.

Here, the timing of occurrence of a hardware interrupt in step 22 will be described with reference to FIGS. The frame pulse shown in FIG. 7A is a signal generated by the MPEG1 encoder LSI 12 and switched between high and low in the frame period of the compressed data output from the MPEG1 encoder LSI 12. Then, at the edge of the frame pulse, a hardware interrupt of the CPU 14 is generated, and the Frame ISR operating on the CPU 14 is executed.

FIG. 6 shows the operation of the Frame ISR.
This will be described with reference to the flowchart shown in FIG.

First, in the Frame ISR, the elementary stream held in the FIFO memory 13 is read out and written into the data buffer A located on the DRAM 15 (step 31). Next, 32 bits of data are read from the data buffer A (step 32).
2-bit data is 100 hex (picture start code:
Picture start code) is checked (step 33). If the 32-bit data does not match 100 hex (step 33 → N), the process returns to step 32 and reads the 32-bit data held at the next address of the data buffer A (step 32). The processes of steps 32 and 33 are repeated until a picture start code is found.

If the 32-bit data matches 100 hex (step 33 → Y), it indicates that the data thereafter is a picture area, so the next 32-bit data is read out (step 34) and 3 bit Picture coding type (Picture coding typ) used for discrimination of I, B and P pictures contained in the bit data
e) is extracted (step 35). If the picture coding type is 001b (step 36 →
Y), indicating that the next data is an I picture, in this case, the data stored in the data buffer A is written to a dedicated buffer (I picture buffer: located on the DRAM 15) for storing only the I picture ( Step 3
7).

Here, the picture coding type is
In the case of 010b (step 36 → N → step 38 → Y),
Since the next data is a P picture, the contents of the data buffer A are not copied to the I picture buffer. Similarly, when the picture coding type is 011b (step 36 → N → step 38 → N → step 39 →
Y), since the next data becomes a B picture, it is not copied to the I picture buffer. In this way, only the I picture is stored in the I picture buffer. Then, MAIN
The routine is called (step 40), and the interrupt processing ends.

Conventionally, by performing such processing, data composed of only I pictures could be transmitted. However, when determining the picture by searching the header included in the elementary stream,
It is necessary to read data until a picture coding type is found. In the worst case, the processing of the current Frame ISR does not end until the next hardware interrupt occurs, and the data is stored in the FIFO memory 13. There is a risk that the data to be read will be overwritten and the data to be read will be lost. In this case, since data that is not transmitted even if it is an I-picture appears, the image displayed on the display monitor 4 of the image receiving device 3 is disturbed or interrupted.

[0014]

In the image transmission apparatus described in the above conventional example, a plurality of applications (an application for extracting MPEG data from an MPEG encoder LSI, an application for adding an Ethernet and TCP / IP header, a receiving apparatus) Applications that receive control commands from the CPU 14) must be executed in parallel, and the load on the CPU 14 becomes large. Further, if it takes time to search the header as described above, the load on the CPU 14 is further increased, and if the bit rate is increased to obtain high image quality, there is a problem that the processing cannot be performed in time.

In addition, since the header search requires buffering of the data to some extent, the delay of the transmission data with respect to the input data (captured moving image) increases by that much, which causes the real-time property to be impaired. Become.

Accordingly, the present invention provides an image transmission apparatus for transmitting image data compressed by MPEG1 via a network such as Ethernet in order to suppress a disturbance of a received image caused by a collision of network packets, in order to suppress only a specific type of picture. IB with means for transmitting, used for this means
P-picture discrimination (frame type discrimination) is realized by a simple discrimination method based on the initial setting information without searching the stream header, thereby reducing the load on the CPU.
The purpose is to support high bit rates.

[0017]

As a means for achieving the above object, there is provided an image transmitting apparatus for transmitting information data including moving image data via a network in order to reproduce a moving image in the image receiving apparatus. And encoding means for compressing and encoding the moving image data, and transmitting means for transmitting the compressed moving image data to the network, wherein the encoding means determines a plurality of types of frames in advance. Output in the same order, and counting the number of frames of the moving image data output from the encoding means, to determine the type of frame of the moving image data output from the encoding means, An image characterized in that when a network is congested, frames of a specific type of the compressed moving image data are thinned out and transmitted. It is intended to provide a feeding apparatus.

An image transmission method for transmitting information data including moving image data via a network in order to reproduce the moving image in the image receiving apparatus, the method comprising compression encoding the moving image data, When transmitting the moving image data to the network, while encoding and outputting a plurality of types of frames in a predetermined order and counting the number of frames of the moving image data to be encoded and output, The frame type of the moving image data to be encoded and output is determined, and when the network is congested, frames of a specific type of the compressed moving image data are thinned out and transmitted. It is an object of the present invention to provide a characteristic image transmission method.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, a schematic configuration of an embodiment of an image transmission apparatus according to the present invention will be described with reference to an example of a network image transmission system shown in FIG.

In FIG. 1, an image transmission apparatus 1 includes an MPEG compression circuit (encoding means) 5 for compressing an input video signal according to the MPEG1 standard, and a compressed data (elementary stream) transmitted from the MPEG compression circuit 5 to a data bus 6. And an Ethernet circuit (transmission means) 7 for transmission to the Ethernet (network) 2. This MPEG
The compression circuit 5 mainly includes an A / D converter 11 for converting an analog input video signal into a digital signal, and the A / D converter 1
An MPEG1 encoder LSI 12 for compressing the digital video signal output from the MPEG1 encoder according to the MPEG1 standard, and an FIF for temporarily storing moving image data compressed by the MPEG1 encoder LSI12
O memory 13 and CPU 14 for controlling the entire MPEG compression circuit 5
And D which is a memory used to operate the CPU 14
And a frame counter 16 for counting frame pulses by software.
15. The compressed moving image data transmitted to the Ethernet 2 is received by the image receiving device 3, decompressed by MPEG1, and output to the display monitor 4 to display the moving image.

In the image transmission apparatus 1 having such a configuration, an input video signal is converted into an A / D signal by the A / D converter 11.
After the D conversion, the data is compressed by the MPEG1 encoder LSI 12 and output in the form of an elementary stream. At this time, FI
The number of bytes of data to be written to the FO memory 13 is stored in the internal register (FIFO Data register) 1 of the MPEG1 encoder LSI 12.
Writing in 2a. This is temporarily stored in the FIFO memory 13 and is sequentially read out on a first-in first-out basis in accordance with a request from the CPU 14. Note that the MPEG1 encoder LSI 12 determines the number of bytes of data to be written to the FIFO memory 13 simultaneously with the output of the elementary stream.
The data is written in the internal register (FIFO Data register) 12a of the SI 12. The data read from the FIFO memory 13 is sent to the Ethernet circuit 7 via the data bus 6, where the header information such as Ethernet / IP / TCP is added, and then sent to the Ethernet 2.

Here, as a first embodiment, a case where only an I picture is transmitted will be described with reference to the drawings.
In this case, the software configuration includes a MAIN routine for initializing the MPEG1 encoder LSI and sending data to the Ethernet (FIG. 2A) and a hardware interrupt routine (Frame IS) as shown in the flowchart of FIG.
R) ((B) in the figure). Hereinafter, the operation of these programs will be described with reference to the flowchart of FIG.

In the Main routine shown in FIG.
First, the MPEG1 encoder LSI 12 is initialized. Here, microcode is loaded into the MPEG1 encoder LSI 12, and necessary parameters are set. The parameters include bit rate, screen size, I-picture distance (I-picture distance), B-picture insertion number (B-picture number), and the like. Where I
The discussion proceeds assuming that the picture interval is set to 9 and the number of inserted B pictures is set to 2 (step 51). Next, a frame counter 16 described later is set to 0 (step 52).

Further, a command to start encoding is issued to the MPEG1 encoder LSI 12 (step 53). MPEG1
When encoding (encoding) in the encoder LSI 12 starts, compressed data (elementary stream) is output from the MPEG1 encoder LSI 12 and temporarily stored in the FIFO memory 13. The picture configuration of the elementary stream is as set in the parameters set in step 51. In this case, for example, I, B, B, P, B, B, P, B,
As shown in B, IB, B, P, B, B, P, B, B, I..., The interval between I pictures is 9 frames, and the number of B pictures is 2 frames. Then, it is checked whether the call was made from FrameISR (step 54), and if there was no call (step 54 →
N), loop and wait until there is a call. Where F
rame ISR is called by hardware interrupt
This is a program stored in the DRAM 15. Frame IS
When called by R (step 54 → Y), the data is read from the I picture buffer in which the data of the I picture is stored, and a header such as Ethernet, IP, or TCP is added (step 55). The data is transmitted to the Ethernet 2 (step 56). Thereafter, the flow returns to step 54 and the processing of steps 54 to 56 is repeated until there is no more data to be sent.

Next, the Frame ISR will be described. Frame I
Since SR is a program called by a hardware interrupt, it is called according to the timing of occurrence of a hardware interrupt shown in FIG. 3, the frame pulse shown in FIG. 3A is a signal generated by the MPEG1 encoder LSI 12 and switched between high and low in the frame period of the compressed data output from the MPEG1 encoder LSI 12. Then, at the edge of this frame pulse, a hardware interrupt of the CPU 14 is generated, and C
The frame ISR operating on the PU 14 is executed ((C) in the figure).

In the frame ISR of this embodiment, the number of frames is counted and the value of the frame counter 16 is incremented at the edge of a frame pulse, that is, when a hardware interrupt occurs. Then, as shown in FIG. 3D, the picture coding type (frame type) of the data read from the FIFO memory 13 is known in advance by the parameters of the initial setting, so that the frame counter shown in FIG. It is possible to easily determine the data of the I picture to be transmitted from the value of 16 (FIG. 11F).

The operation of the Frame ISR is roughly shown in FIG.
This will be described with reference to the flowchart shown in FIG.

First, the maximum value of the frame counter 16 (Co
unt max.), the value (= 9) of the interval between I pictures set in step 51 is set. As a result, the frame counter 51 performs a count-up operation from 1 to 9, and then returns to 1 to perform a loop-type count-up operation. As described in the timing chart of FIG. 3, the actual count operation updates the count value of the frame counter 51 every time a hardware interrupt occurs. Since the timing at which the I picture is read from the FIFO memory 13 is when the count value is 1, 1 is set as the I-picture position (step 61).

Then, every time a hardware interrupt occurs, it is determined whether or not the value of the frame counter 16 is the maximum value (= 9) (step 62), and the maximum value (= 9).
(Step 62 → Y), the frame counter 16 is set to 1 (step 63), and the maximum value (= 9)
Otherwise (step 62 → N), the frame counter 16 is incremented by one (step 64).
At this time, the value of the frame counter 16 is I-picture posi
If the value matches the option (= 1) (step 65 →
Y) Since the I picture is held in the FIFO memory 13, the FIFO data register in the MPEG1 encoder LSI 12 is read, the data is read from the FIFO memory 13 by the number of bytes, and the I picture buffer (FIG. (Not shown) (step 66). When the value of the frame counter 16 is I-picture position (= 1)
If not (step 65 → N), B is stored in the FIFO memory 13.
Or because a P picture is held,
A read operation is performed to update the read address of the IFO memory 13, but is not stored in an I picture buffer on the DRAM 15. Finally, the MAIN routine is called to terminate the interrupt processing (step 67).

The MPEG1 encoder LSI 12 has a CPU
The encoded data is temporarily stored in the FIFO memory 13 so that the data can be taken out in response to the request in step 14.
The data is written in the internal register (FIFO Data register) 12a of the EG1 encoder LSI 12. Then, the CPU 14 reads the value of the FIFO data register 12a, thereby
The amount of data held in the memory 13 can be known, and data can be read by this number of bytes.

As described above, in the first embodiment, the picture is determined by using the frame count function based on the initial setting of the MPEG1 encoder LSI 12 without performing the header search of the stream unlike the related art. Therefore, the load on the CPU 14 can be greatly reduced.

In the first embodiment, only I pictures are transmitted by using this method.
We are reducing the amount of data. In the above example, the I-picture distance (I-picture distance) is set to 9, but if it is increased to 12, 15, 18,..., The number of I-pictures per time can be reduced. The data amount can be further reduced without lowering the image quality. As a result, although the frame-feeding state is conspicuous at the time of reproducing the moving image, the data collision occurring on the network is improved.

Next, as a second embodiment of the present invention,
FIG. 4 shows a case where an I picture and a P picture are transmitted.
This will be described with reference to the flowchart shown in FIG. In addition,
In the present embodiment, the operation in the Main routine shown in FIG. 4A is the same as the operation in the first embodiment described with reference to FIG. The same step numbers are assigned to parts performing the same operation, and the description thereof will be omitted.

Hereinafter, the Frame ISR shown in FIG. 4B will be described. Note that parts performing the same operations as those described in FIG. 2B are given the same step numbers as in FIG. 2B. First, the maximum value of the frame counter 16 (Count ma
x.), the value (= 9) of the I-picture interval set in step 51 is set. As a result, the frame counter 51 performs a count-up operation from 1 to 9, and then returns to 1 to perform a loop-type count-up operation. As described in the timing chart of FIG. 3, the actual count operation updates the count value of the frame counter 51 every time a hardware interrupt occurs. The timing at which the I picture is read from the FIFO memory 13 is when the count value is 1, and the timing at which the P picture is read from the FIFO memory 13 is when the count value is 4 or 7. 1, 4, and 7 are set as IP-picture positions (step 61b).

Each time a hardware interrupt occurs, it is determined whether the value of the frame counter 16 is the maximum value (= 9) (step 62), and the maximum value (= 9).
(Step 62 → Y), the frame counter 16 is set to 1 (step 63), and the maximum value (= 9)
Otherwise (step 62 → N), the frame counter 16 is incremented by one (step 64).
At this time, the value of the frame counter 16 is set to IP-picture pos
In the case of a match (= 1 to 4 or 7) (step 65 → Y), since the I picture or the P picture is held in the FIFO memory 13, the FIFO data register in the MPEG1 encoder LSI 12 is read, and only the number of bytes is read. Data is read from the FIFO memory 13 and written to an IP picture buffer (corresponding to the I picture buffer of the first embodiment: not shown) on the DRAM 15 (step 66). When the value of the frame counter 16 is IP-picture
If it is not position (= 1 or 4 or 7) (step 6
(5 → N) Since the B picture is held in the FIFO memory 13, a read operation is performed to update the read address of the FIFO memory 13, but the B picture is not stored in the IP picture buffer on the DRAM 15. And finally
The MAIN routine is called to end the interrupt processing (step 67).

The MPEG1 encoder LSI 12 has a CPU
The encoded data is temporarily stored in the FIFO memory 13 so that the data can be taken out in response to the request in step 14.
The data is written in the internal register (FIFO Data register) 12a of the EG1 encoder LSI 12. Then, the CPU 14 reads the value of the FIFO data register 12a, thereby
The amount of data held in the memory 13 can be known, and data can be read by this number of bytes.

As described above, even when transmitting I and P pictures, it is not necessary to search the header of the stream, so that the load on the CPU 14 can be greatly reduced, and the amount of data transmitted to the network can be easily reduced. Can be.
Also, just changing the software subroutine
You can switch between sending only and sending I and P,
An image transmission device that can perform both can be easily realized.

[0038]

According to the image transmission apparatus and the image transmission method of the present invention, a picture coding type (frame type) is used.
Can be easily determined without searching the header of the elementary stream, so that the load on the CPU is greatly reduced, and the image distortion at the receiving device caused by collision or loss of packets flowing on the network This has the effect of being able to mitigate.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing an example of a network image transmission system in which an image transmission device and an image transmission method of the present invention are used.

FIG. 2 is a first diagram illustrating an image transmission apparatus and an image transmission method according to the present invention.
It is a flowchart figure for demonstrating embodiment.

FIG. 3 is a first image transmission apparatus and an image transmission method according to the present invention;
FIG. 5 is a timing chart for explaining the operation of the embodiment.

FIG. 4 is a second view of the image transmission apparatus and the image transmission method of the present invention.
It is a flowchart figure for demonstrating embodiment.

FIG. 5 is a configuration diagram illustrating an example of a conventional network image transmission system.

FIG. 6 is a flowchart for explaining a conventional image transmission device and image transmission method.

FIG. 7 is a timing chart for explaining the operation of a conventional image transmission device.

[Explanation of symbols]

 Reference Signs List 1 image transmission device 2 Ethernet (network) 3 image reception device 4 display monitor 5 MPEG compression circuit (encoding means) 6 data bus 7 Ethernet circuit (transmission means) 11 A / D converter 12 MPEG1 encoder LSI 12a internal register (FIFO) Data register) 13 FIFO memory 14 CPU 15 DRAM 16 Frame counter

Claims (2)

[Claims] 1. An image transmission apparatus for transmitting information data including moving image data via a network in order to reproduce a moving image in an image receiving apparatus, wherein the encoding apparatus compresses and codes the moving image data. Means for transmitting the compressed moving image data to the network, the encoding means outputting a plurality of types of frames in a predetermined order, and output from the encoding means. Counting the number of frames of the moving image data to determine the type of frame of the moving image data output from the encoding means, and when the network is congested, the compressed moving image An image transmission apparatus characterized in that a specific type of frame of image data is thinned out and transmitted. 2. An image transmission method for transmitting information data including moving image data via a network in order to reproduce a moving image in an image receiving apparatus, wherein the moving image data is compressed and encoded. When transmitting the obtained moving image data to the network, by encoding and outputting a plurality of types of frames in a predetermined order and counting the number of frames of the moving image data to be encoded and output, The type of the frame of the moving image data to be encoded and output is determined, and when the network is congested, frames of a specific type of the compressed moving image data are thinned out and transmitted. Characteristic image transmission method.