JP4241227B2 - Data transmitting apparatus and data receiving apparatus, data transmitting method and data receiving method, and data communication system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a data transmission device and data reception device, a data transmission method and a data reception method, and a data communication system for communicating real-time data such as encoded moving images, music, and voice via a network.
[0002]
[Prior art]
With the expansion of the Internet environment, a great amount of data has been communicated via a network. The data to be communicated includes real-time data such as moving images, music, and voice in addition to non-real-time data such as characters and still images.
[0003]
In order to communicate these real-time data via a network, some encoding process is performed. In order to increase the encoding efficiency, these encoding processes use a method of encoding and communicating only difference data with reference data using the strength of correlation in the temporal vicinity of the data. Many. However, this method requires that the encoded data is transmitted to the receiving side in the correct order, without omission and at an appropriate timing.
[0004]
On the other hand, the Internet protocol (hereinafter abbreviated as IP) is the basis of current network communication. Features of communication using IP include packet-type communication in which data is divided and communicated, and best-effort type communication.
[0005]
In this best effort type communication, it is not guaranteed that transmitted data reaches the receiving side in the order of transmission. Furthermore, it is not guaranteed that the transmitted data reaches the receiving side. In an actual network environment, a packet that does not reach the receiving side, that is, packet loss may reach as much as 10 percent.
That is, there are many problems to perform real-time data communication on such a network.
Among these problems, (1) retransmission, (2) error correction, (3) other (resynchronization), etc. are conceivable as techniques for dealing with packet loss.
[0006]
The retransmission of (1) is the most frequently used technique, for example, a transmission control protocol (hereinafter abbreviated as TCP) or the like. The details of TCP are described in IETF technical document RFC793.
[0007]
In simple terms, this method involves data communication while handshaking between the transmitting side and the receiving side. That is, data communication proceeds by sending back ACK, which is data indicating that the data has been properly received from the receiving side to the transmitting side. The transmitting side retransmits data when ACK is not returned within a certain time. Thus, the handshake using ACK guarantees that the transmission data has arrived at the receiving side.
[0008]
Further, as an example of the error correction of the above (2), there is a method introduced as a conventional example in “Encoding method and encoder” described in Japanese Patent Laid-Open No. 2000-134619.
These methods are simply to detect and restore packet loss by interleaving data, distributing it into a plurality of packets, and attaching error correction codes to them.
[0009]
For example, k pieces of data are interleaved into h + k pieces of data, and error correction codes are attached. The transmitting side transmits this h + k pieces of data. On the receiving side, if, for example, arbitrary k pieces of data among h + k pieces of data can be received, the entire transmission data can be completely restored by using the data included therein and the error correction code.
The ratio between h and k is determined by statistical information of the network, and it is probabilistically guaranteed that transmission data is properly restored on the receiving side.
[0010]
As an example of resynchronization in (3) above, there is a “transmitting apparatus and method thereof and receiving apparatus and method thereof” described in JP-A-2002-10265. Unlike the above-mentioned techniques (1) and (2), this technique aims at minimizing the influence of packet loss on video, music, and voice communication rather than data integrity. is there.
[0011]
First, with reference to FIGS. 12 and 13, the influence of packet loss taking moving image data as an example will be briefly described. The squares F1, F2,. Called ..)
[0012]
Of these frames F1, F2,..., Thick frames F1 and F5 indicate frame data (hereinafter abbreviated as I picture) that can be decoded alone. Further, the other narrow frame rectangles (frames F2, F3, F4, F6, F7, etc.) indicate frames (hereinafter abbreviated as P pictures) made of differential data that cannot be decoded by themselves. Further, a curved arrow a connecting these squares indicates a dependency relationship between frames.
[0013]
In a normal moving picture coding system, an I picture is set to appear periodically, and a set of data from one I picture to the next I picture is called a group of pictures (hereinafter abbreviated as GOP). In the example of FIG. 12, it can be said that the frames F1 to F4 are one GOP. This GOP can be decoded independently of each other. In other words, if a packet loss occurs in a GOP, data in the same GOP cannot be decrypted properly thereafter.
[0014]
FIG. 12 schematically shows a case where all communication is performed properly. Frames F1, F2,... Corresponding to the respective packets sent from the transmission side shown in FIG. 12A are sent to the reception side in the correct order and without packet loss, as shown in FIG. Suppose you arrive at the right time. In this case, all the frames F1, F2,... Can be properly decoded, whereby the original data can be properly restored.
[0015]
On the other hand, FIG. 13 shows a case where, for example, the frame F3 is lost due to packet loss among the frames F1, F2,... Corresponding to the respective packets sent from the transmission side shown in FIG. Assume that a frame with a cross in the figure lost a packet).
[0016]
In this case, it is natural that the data of the frame F3 cannot be decoded, but even if the data of the subsequent frame F4 is properly received, the frame F4 depends on the data of the frame F3, so that it is decoded properly. The original data cannot be restored properly. For this reason, in order to properly restore the original data, it cannot be resumed until the start frame of the next GOP, that is, the frame F5 is received.
[0017]
In MPEG or the like, generally, data corresponding to about 0.5 seconds is generally a GOP. That is, there is a possibility that data of about 0.5 seconds at maximum cannot be decoded due to packet loss. This is a sufficiently long time for human perception, and causes a large deterioration in image quality. The technique described in Japanese Patent Laid-Open No. 2002-10265 described above is intended to improve such a problem, and the technique described in Japanese Patent Laid-Open No. 2002-10265 will be briefly described below.
First, a technique cited as a conventional example in Japanese Patent Laid-Open No. 2002-10265 will be briefly described with reference to FIG.
[0018]
Simply speaking, when a packet loss is detected, a NACK indicating that the data has not arrived is returned from the receiving side to the transmitting side, whereby the transmitting side transmits data that is the source of the difference (hereinafter referred to as reference data). The data with changed data is transmitted.
[0019]
First, as shown in FIG. 14, when the frame F4 is received while the frame F3 has not arrived, it is detected that the frame F3 has lost a packet. In this state, as in the example of FIG. 13, neither the frame F4 nor the subsequent frame F5 can be properly decoded, and the original data cannot be restored properly.
[0020]
In the technique disclosed in Japanese Patent Laid-Open No. 2002-10265 as a conventional example, a NACK is returned from the reception side to the transmission side in this case. The transmitting side that has received the NACK transmits the data corresponding to the next frame F5 as the difference data corresponding to the frame properly received by the receiving side in the vicinity of the missing frame. For example, in the case of FIG. 14, the frame F2, which is data immediately before the missing frame F3, is used as new reference data.
On the receiving side, since the frame F2 is properly received, the differential data of the frame F5 corresponding thereto can be properly decoded.
[0021]
As described above, according to this technique, the frame F3 and the frame F4 cannot be properly decoded, and the restoration of the original data corresponding thereto fails, but the restoration can be resumed in the subsequent frames. .
[0022]
However, in this technique, when communication is performed between multiple points, in order to cope with a case where a loss occurs in a different frame on each of the receiving sides, reference data corresponding to each receiving side on a one-to-one basis is used. There is a problem that the sender needs to have. The technique described in Japanese Patent Application Laid-Open No. 2002-10265 solves such a problem.
[0023]
In simple terms, the technique described in Japanese Patent Application Laid-Open No. 2002-10265 is such that reference data to be switched when a packet loss occurs is determined in advance as a rule on the transmission side and all reception sides. As a result, the transmitting side only needs to notify all receiving sides that reference data is to be switched, and it is not necessary to have reference data corresponding to each receiving side on a one-to-one basis.
[0024]
FIG. 15 is a diagram for explaining this, and shows a case where the reference frame when the packet loss occurs is designated as the frame F1.
In FIG. 15, for example, when a packet loss corresponding to the frame F3 occurs, the frame F4 is encoded using the frame F1 as reference data.
[0025]
Further, when a packet loss corresponding to the subsequent frame F5 occurs, in this case as well, the frame F1 is used as the reference data, not the frame F4 that is the closest data properly received.
[0026]
In any of the above cases, when the NACK returned by any receiving side is received on the transmitting side, the reference data is switched to data predetermined by the rule, and all that is notified to the receiving side. The data can be properly restored on the receiving side. However, a means for storing the frame data with the possibility of switching is required on all receiving sides.
[0027]
FIG. 16 is a diagram corresponding to FIG. 4 described in Japanese Patent Application Laid-Open No. 2002-10265 and showing the configuration of the transmission side (server device). In Japanese Patent Laid-Open No. 2002-10265, the server apparatus has main encoding means (first encoding module) and second encoding means (second encoding module) that supplements the main encoding means (first encoding module). There may be a plurality of second encoding modules.
[0028]
However, in Japanese Patent Application Laid-Open No. 2002-10265, it is not essential that the second encoding module exists. That is, the presence of the second encoding module has nothing more than the meaning of preparing in advance encoded data corresponding to reference data to be switched when any receiving side detects a packet loss. That is, switching of reference data and re-encoding may be sequentially performed using main encoding means (first encoding module).
[0029]
[Patent Document 1]
JP 2000-134619 A
[Patent Document 2]
JP 2002-10265 A
[0030]
[Problems to be solved by the invention]
In the data communication technique using retransmission as described above, the order and completeness of data are guaranteed. Therefore, it is suitable for non-real time communication. However, in this data communication method, handshaking is always required and a large delay occurs. Therefore, it is not suitable for real-time data communication.
[0031]
In addition, although handshaking is not performed in the error correction technique, in order to interleave the data, the data is dispersed on the time axis, which also causes a large delay. Therefore, it is not suitable for real-time data communication.
Further, the resynchronization process described in Patent Document 2 described above includes a handshake process such as NACK communication. Therefore, a large delay occurs in this part.
[0032]
As described above, most of the conventional techniques employ a data communication method with a delay. However, in bi-directional real-time data communication, delay is a big problem, and communication that does not require a cause of delay such as handshake is required.
[0033]
Moreover, although the example applied to the multipoint communication system is shown in the above-mentioned patent document 2, this technology assumes communication between a very small number of terminals even if it is multipoint communication. It is a thing. For this reason, as the present invention assumes, in the situation where the number of receiving sides becomes extremely large and any receiving side detects packet loss in almost all frame data, the above-mentioned patent document There is a high possibility that the technology of 2 will substantially fail.
[0034]
Therefore, the present invention can minimize data loss and minimize the occurrence of delay, and can be applied to real-time data communication, particularly bidirectional real-time data communication. An object is to provide a data transmission method, a data reception method, and a data communication system.
[0035]
[Means for Solving the Problems]
(1) A data transmission apparatus according to the present invention is a data transmission apparatus that encodes encoding target data input in time series and outputs the encoded data. The data transmission apparatus distributes the encoding target data into a plurality of data. The distribution means and the respective encoding target data distributed by the data distribution means are provided correspondingly, and each of the input encoding target data is independently encoded and obtained And a plurality of encoding means capable of transmitting each encoded data.
[0036]
As described above, the data transmission device described in (1) distributes encoding target data (for example, data for one frame of moving image data) to a plurality of pieces, and corresponds to each distributed encoding target data. The encoding means encodes each independently, and a plurality of encoded data obtained thereby are transmitted as encoded data of a plurality of channels.
[0037]
That is, the encoding means on the data transmission side described in (1) are independent from each other, and, for example, the technique described in the above-mentioned Patent Document 2 is different from that described as a conventional example therein. It is not a thing. That is, conventional techniques can be used as they are as individual encoding means. This is an important factor for maintaining compatibility with the prior art.
[0038]
In addition, the operation on the data transmission side described in (1) simply continues to output data and does not depend on the number on the data reception side. Therefore, it is possible to easily perform broadcast-type communication in which a large number of data receiving sides exist. In addition, since the data transmitting device and the data receiving device do not perform handshake processing, the data delay caused thereby can be minimized. As is obvious, there is no load for handshaking with a large number of data receiving apparatuses.
[0039]
This also has the effect of not degrading the encoding efficiency. For example, the technique used in the technique described in Patent Document 2 described above, when receiving a NACK signal from the receiving side at the time of packet loss, encodes data using temporally separated data as reference data. This reduces the correlation between the data and degrades the encoding efficiency. On the other hand, in the present invention, when encoding, only temporally neighboring data is used as reference data. Therefore, the correlation between data is expected to be high, and the encoding efficiency is not deteriorated.
[0040]
(2) It is preferable that the plurality of encoding units perform encoding processing on each encoding target data by changing encoding parameters between the respective encoding units.
[0041]
This means, for example, that the position of the above-mentioned I picture is made different by a plurality of encoding means, which causes a data error such as packet loss in the encoded data of a certain channel on the receiving side. However, the probability that decoding can be performed using the encoded data of other channels is increased, and the influence of packet loss can be reduced.
[0042]
(3) A data receiving device that decodes the encoded data transmitted from the data transmitting device described in (1) or (2) and outputs the decoded data, and is transmitted from the data transmitting device. Each of the encoded data can be received and decoded, and a plurality of decoding means capable of detecting a data error in the received encoded data and any one of the plurality of decoding means is prioritized. A decoding means management means that can be set as a decoding means and that can select the decoded data of any one of the plurality of decoding means; and the decoding means management means The data error detection state in the decoding means set as the priority decoding means is monitored, and if there is no data error, the decoded data in the priority decoding means is selected and output, and the data If there is any other decoding means that can properly decode the encoded data, it is checked whether there is another decoding means that can properly decode the encoded data. It is characterized by selecting and outputting the decoded data.
[0043]
As described above, in the data receiving device described in (3), the decoding means configured to receive the encoded data of a plurality of channels transmitted from the transmission side by the corresponding decoding means and set as the priority decoding means. When a data error is detected during the decoding process, the decoded data of the other decoding means that can properly decode the encoded data having the data error is selected and output, so that the packet loss The original data can be properly restored without being affected.
[0044]
Further, in the data receiving device described in (3), as in the data transmitting device, as a decoding unit of each of the plurality of decoding units, the conventional example (for example, the technique described in Patent Document 2 described above is used). Among them, those equivalent to those described as conventional examples) can be used.
[0045]
That is, the encoded data transmitted by the data transmitting apparatus of the present invention can be received and decoded by a data receiving apparatus not corresponding to the present invention. However, in this case, the operation performed by the present invention, that is, when there is a data error, the operation of dynamically switching the decoding means that can be decoded and selecting the appropriate decoded data cannot be performed. The significance of being able to use this data receiving device is great.
[0046]
(4) When the decoding means management means determines that there is another decoding means capable of properly decoding the encoded data having the data error, the priority decoding means is assigned to the decoding means. It is preferable to update.
[0047]
In this way, when it is determined that there is another decoding means that can properly decode the encoded data having the data error, by performing processing for updating the priority decoding means to the decoding means, Depending on the time, it is possible to dynamically switch the decoding means capable of decoding properly and select the appropriate decoded data, thereby properly restoring the original data.
[0048]
(5) The data error detected by the data error detection means is a lack of encoded data for each of the encoding target data.
[0049]
By detecting the lack of encoded data with respect to the encoding target data (expressed as packet loss in the embodiment described later) and performing processing corresponding thereto, high quality that is not affected by packet loss Data can be restored.
[0050]
(6) In a data transmitting apparatus that encodes encoding target data input along a time series and outputs the encoded data, a data distribution unit that distributes the encoding target data into a plurality of data, and the data distribution unit Is provided for each encoding target data distributed in the above, and each input encoding target data can be independently encoded, and each encoded data obtained thereby can be transmitted. A plurality of encoding means, and a data reduction means provided on the input side of at least one of the plurality of encoding means and for reducing the data to be encoded distributed by the data distribution means. It is characterized by having.
[0051]
In addition to the effect of the data transmission device of (1) described above, the data transmission device described in (6) inputs reduced data to at least one of the plurality of encoding units. For example, it is possible to easily realize a sample display or a moving image thumbnail display for an image.
[0052]
Also, in the data transmitting apparatus described in (6), as in the data transmitting apparatus in (1), the plurality of encoding units are configured to perform encoding between the respective encoding units for each encoding target data. It is preferable to perform the encoding process by changing the encoding parameter.
[0053]
(7) A data reception device that decodes the encoded data transmitted from the data transmission device according to (6) and outputs the decoded data, and each of the data transmission devices transmitted from the data transmission device A plurality of decoding means capable of receiving and decoding the encoded data and detecting a data error of the received encoded data, and an output size acquisition for acquiring the output size of the decoded data to be output And a decoding means capable of decoding with an output size adapted to the output size acquired by the output size acquisition means can be set as a priority decoding means, and among the plurality of decoding means, A decoding means management means capable of selecting the decoded data of any of the decoding means, and the decoding means management means is a decoding means set as the priority decoding means. The data error detection state is monitored, and if there is no data error, the decoded data in the priority decoding means is selected and output. If there is a data error, the encoded data can be decoded properly. It is characterized by checking whether there is another decoding means, and if there is another decoding means that can be properly decoded, the decoded data of the decoding means is selected and output.
[0054]
Thereby, in addition to the effect of the data receiving apparatus described in (3), for example, sample display or moving image thumbnail display can be easily realized in the image. In order to realize this, the data receiving apparatus only needs to receive the reduced encoded data sent from the data transmitting apparatus side and decode it. In the case of a personal computer, the processing load on the user side device can be reduced and the load on the network can be reduced. In addition, thumbnail images and the like can be viewed on the receiving side.
[0055]
In the data receiving method described in (7), the data error detected by the data error detecting means is a lack of encoded data for the respective encoding target data.
[0056]
(8) Decoding data expansion / reduction means capable of expanding / reducing the decoded data selected by the decoding means management means is provided, and the decoded data expansion / reduction means When the data size is different from the output size acquired by the output size acquisition unit, a data size adjustment process is performed on the decoded data so that the data size matches the output size acquired by the output size acquisition unit. Preferably it is done.
[0057]
As a result, a data error is detected by a decoding means capable of decoding with an output size adapted to the output size specified by the output size specifying means, and decoding is performed by another decoding means having a different output size. Even in this case, the decoded output size can be converted into an output size suitable for the output size specified by the output size specifying means and output.
[0058]
(9) When the decoding means management means determines that there is another decoding means capable of properly decoding the encoded data having the data error, the priority decoding means is assigned to the decoding means. If the output data size of the updated priority decoding means is different from the output size acquired by the output size acquisition means, the decoded data of the priority decoding means is selected and output. Preferably, the priority decoding means is returned to a decoding means that matches the output size acquired by the output size acquisition means.
[0059]
As described above, when it is determined that there is another decoding unit that can appropriately decode the encoded data having the data error, by performing processing for updating the priority decoding unit to the decoding unit, Depending on the time, it is possible to dynamically switch the decoding means capable of decoding properly and select the appropriate decoded data, thereby properly restoring the original data.
[0060]
If the output data size of the updated priority decoding unit is different from the output size acquired by the output size acquisition unit, the decoded data of the priority decoding unit is selected and output, and then the priority Since the decoding means is returned to the decoding means that matches the output size acquired by the output size acquisition means, the output of decoded data different from the output size acquired by the output size acquisition means is minimized. You can stay. As a result, it is possible to minimize the time for enlarging or reducing the decoded data and outputting it, and to suppress degradation of the quality of the output data.
[0061]
(10) In a data transmission method for encoding data to be encoded input along a time series and outputting the encoded data, the data to be encoded is distributed into a plurality of data, and each of the distributed data The encoding target data is input to a plurality of data encoding means corresponding to the respective encoding target data, encoded by the respective encoding means, and each encoded data obtained thereby is encoded. It is characterized by transmitting.
[0062]
By adopting such a data transmission method, it is possible to obtain the same effect as that of the above-described data transmission apparatus (1). Also in the data transmission method described in (10), the plurality of encoding units perform encoding processing on each encoding target data by changing encoding parameters between the encoding units. Is preferred.
[0063]
(11) A data receiving method for decoding encoded data transmitted by the data transmitting method according to (10) and outputting the decoded data, wherein each transmitted encoded data is One of a plurality of decoding means that can receive and decode the received encoded data and can detect a data error of the received encoded data is set as a priority decoding means, and the priority decoding is performed. Monitoring the data error detection state in the decoding means set as the conversion means, if there is no data error, select and output the decoded data in the priority decoding means, if there is a data error, It is checked whether or not there is another decoding means that can properly decode the encoded data. If there is another decoding means that can properly decode the decoded data, the decoded data of the decoding means is selected. It is characterized in that to the output.
[0064]
By adopting such a data receiving method, the same effect as the above-described data receiving apparatus (3) can be obtained.
[0065]
(12) In a data transmission method for encoding data to be encoded input along a time series and outputting the encoded data, the data to be encoded is distributed into a plurality of codes, and each of the distributed codes The data to be encoded is input in correspondence with a plurality of data encoding means provided for the respective encoding target data, and at least one of the plurality of data encoding means In other words, the encoding target data subjected to the data reduction processing by the data reduction means is inputted, encoded by the respective encoding means, and the respective encoded data obtained thereby are transmitted.
[0066]
By adopting such a data transmission method, an effect similar to that of the data transmission apparatus of (6) described above can be obtained. Also, in the data transmission method described in (12), as in the data transmission device of (1), the plurality of encoding units may perform the respective encoding target data between the respective encoding units. It is preferable to perform the encoding process by changing the encoding parameter.
[0067]
(13) A data receiving method for decoding each encoded data transmitted by the data transmitting method described in (12) and outputting the decoded data, and an output size of the decoded data to be output Among the plurality of decoding means capable of receiving and decoding each transmitted encoded data and detecting a data error in the received encoded data. A decoding means capable of outputting decoded data with an output size suitable for the output size set is set as a priority decoding means, and a data error detection state in the decoding means set as the priority decoding means is monitored. If there is no data error, the data decoded by the priority decoding means is selected and output. If there is a data error, the encoded data is appropriately set. It is characterized by checking whether there is other decoding means that can be encoded, and if there is other decoding means that can be properly decoded, the decoded data of the decoding means is selected and output. Yes.
By adopting such a data receiving method, the same effect as that of the data receiving apparatus described in (7) above can be obtained.
[0068]
Also in the data receiving method described in (13), the data error detected by the data error detecting means is a lack of encoded data for the respective encoding target data. Further, when the decoded data size of the selected decoding means is different from the required output size, the data size adjustment processing is performed on the decoded data so that the data size matches the required output size. It is preferable to carry out.
[0069]
Further, when it is determined that there is another decoding means capable of properly decoding the encoded data having the data error, the priority decoding means is updated to the decoding means, and the updated priority decoding is performed. If the output data size of the converting means is different from the required output size, the decoded data of the preferential decoding means is selected and output, and then the preferential decoding means is decoded to match the required output size. It is preferable to return to the converting means.
[0070]
(14) A data transmission apparatus that encodes encoding target data input along a time series and outputs the encoded data, and receives and decodes the encoding target data transmitted from the data transmission apparatus In the data communication system having the data receiving device that outputs the decoded data, the data transmitting device is the data transmitting device according to any one of (1), (2), and (6), A data communication system, wherein the data receiving device is the data receiving device according to any one of (3), (4), (5), (7), (8), and (9).
[0071]
As a result, the data transmitting apparatus side can take over the effects of the data transmitting apparatus described in (1), (2), and (6) as it is, and the data receiving apparatus side can perform the above described (3), ( 4), (5), (7), (8), (9), the effects of the data receiving apparatus can be taken over as they are. For this reason, normal data restoration is possible on the data receiving device side without being affected by packet loss, and in the data communication system of the present invention, no handshake with the data receiving device side is required. Delay can be minimized, and a data communication system optimal for real-time data communication, particularly bidirectional real-time communication can be obtained.
[0072]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below. The contents described in this embodiment include descriptions of the data transmission apparatus and data reception apparatus, data transmission method and data reception method, and data communication system of the present invention. In this embodiment, a moving image is taken as an example of real-time data to be processed. However, the present invention is not limited to a moving image, and can be similarly applied to real-time data such as music and voice.
[0073]
[Embodiment 1]
FIG. 1 is a block diagram showing a configuration example of Embodiment 1 of the data transmission apparatus of the present invention. The data transmission apparatus according to the first embodiment encodes data for each encoding unit (for example, one frame of moving image data) along a time series from real-time data to be transmitted (in this embodiment, a moving image). Data input means 1 for inputting the data to be encoded, data distribution means 2 for distributing the encoding target data input to the data input means 1 into N pieces (N is a natural number), N distributed by the data distribution means 2 N encoding units 301 to 30N are provided corresponding to the respective encoding target data.
[0074]
In the embodiment of the present invention (including Embodiment 1 and Embodiment 2 described later), for the sake of simplicity, it is assumed that the encoding unit is a frame, and one frame is one packet. Will be described as being transmitted.
[0075]
The encoding units 301 to 30N include a data encoding unit 31 that encodes image data corresponding to each of them, and an encoded data transmission unit 32 that transmits the encoded data. In addition, the encoding by each data encoding means 31 is performed by changing the encoding parameter for each of the encoding means 301 to 30N (this will be described later).
[0076]
The encoded data encoded by the data encoding means 31 are output as N-channel encoded data on the data communication path 10 by the corresponding encoded data transmitting means 32, respectively. Not sent to the receiver via the network.
[0077]
FIG. 2 is a block diagram showing a configuration example of the first embodiment of the data receiving apparatus according to the present invention. The data receiving apparatus according to the first embodiment includes a plurality (N) of decoding means 401 to 40N provided independently of each other, decoding means management means 5 (priority decoding means storage means 51 and decoding). A data output means 6 for outputting the decoded data.
[0078]
Note that N of the decoding units 401 to 40N does not necessarily match the N of the encoding units 301 to 30N illustrated in FIG. That is, the number of decoding units 401 to 40N and the number of encoding units 301 to 30N are not necessarily the same.
[0079]
Decoding means 401 to 40N are encoded data receiving means 41 for respectively receiving N channel encoded data sent from the data transmitting apparatus shown in FIG. 1, and when a data error occurs in the received encoded data The data error detecting means 42 for detecting the data error and the encoded data decoding means 43 for decoding the encoded data are provided.
[0080]
In the embodiment of the present invention (including Embodiment 1 and Embodiment 2 described later), the data error to be detected is data loss (hereinafter referred to as packet loss). Therefore, in the following description, the data error detection unit 42 is referred to as a packet loss detection unit 42.
[0081]
Further, as described above, the decoding means management means 5 includes the priority decoding means storage means 51 and the decoding means selection means 52, and the decoding means selection means 52 selects one of the decoding means 401 to 40N. The priority decoding means storage means 51 can be set as the priority decoding means, and the decoding of any of the decoding means 401 to 40N based on the setting contents of the priority decoding means storage means 51 Data can be selected.
[0082]
Specifically, the decoding means selection means 52 sets an arbitrary decoding means as an initial value for the priority decoding means storage means 51 as the priority decoding means, and this priority decoding means storage means 51. The decrypted data by the priority decrypting means is selected from the stored contents of the data, and the selected decrypted data is output to the data output means 6. In the present invention, the encoded data decoding means included in the decoding means not selected by the priority decoding means storage means 51 can be configured to be paused.
[0083]
The decoding means selection means 52 monitors the data error detection state in the decoding means set as the priority decoding means, and when the packet loss is detected by the priority decoding means, the packet lost frame is appropriately It is checked whether or not there are other decoding means that can be decoded. If there is a decoding means that can be properly decoded, the decoding means becomes a new priority decoding means in the priority decoding means storage means 51. It has a function of storing (updating the stored contents), selecting decoded data by the decoding means capable of decoding, and outputting the selected decoded data to the data output means 6. This series of processing will be described later.
[0084]
The data output means 6 outputs the decoded data selected by the decoding means selection means 52 as a data output.
[0085]
In such a configuration, the operation will be described in detail with reference to the flowchart of FIG.
[0086]
First, when the decoding process is started, the priority decoding unit storage unit 51 is initialized (step S1), and any one of the decoding units 401 to 40N has priority as the priority decoding unit. It is stored in the decryption means storage means 51.
[0087]
After that, the N-channel encoded data transmitted from the data transmitting apparatus shown in FIG. 1 is received by the respective encoded data receiving means 41 of the decoding means 401 to 40N corresponding to the respective channels (step S2). . Thereby, the decoding process is started in each of the encoded data decoding units 43 of the decoding units 401 to 40N.
[0088]
At this time, it is determined whether or not a packet loss has been detected in the priority decoding means (priority decoding means initialized by the decoding means selection means 52) stored in the priority decoding means storage means 51 (step S3). If no packet loss is detected, the decoded data decoded by the priority decoding means becomes the data output (step S4).
[0089]
At this time, if a packet loss is detected by the priority decoding means (step S3), it is possible to properly decode the frame corresponding to the encoding target data (in this case, one frame of data) with the packet loss. It is checked whether there is another decoding means (step S5), and if there is a decoding means that can be properly decoded, the decoding means stores the priority decoding means storage means 51 as the priority decoding means. It is set (step S6). Then, the decoded data of the decoding means newly set as the priority decoding means becomes the data output (step S4).
[0090]
Here, being able to decrypt properly means that there is no data error. In the embodiment of the present invention (including Embodiment 1 and Embodiment 2 described later), the data error is a packet loss. However, in addition to the packet loss, this data error includes ▲ 1) Due to jitter, delay, etc., the packet (data) has not arrived at the timing required for decoding, or [2] some error has been detected in the arrived packet (data). In other words, the fact that decoding can be properly performed means that there is no packet loss and the state does not correspond to the above (1) and (2).
[0091]
If there is no decoding means that can properly decode in step S5, some processing for packet loss, for example, processing of continuously using the last decoded data is performed (step S7), and thereafter Then, it waits until any one of the decoding means can be properly decoded (step S8). When any one of the decoding means can be properly decoded, the decoding means is stored as the priority decoding means in the priority decoding means storage means 51 (step S6), and the new priority decoding is performed. The decrypted data of the decrypting means used as the means becomes the data output (step S4).
[0092]
Then, returning again to step S2, each decoding unit 401 to 40N corresponding to the N channel receives data for one frame to be encoded next, and performs the processing after step S3. In this way, the decoding process of the encoded data is performed. Hereinafter, the operation described above will be described in more detail with reference to FIGS.
[0093]
FIG. 4 shows that in the data transmission apparatus according to the first embodiment of the present invention shown in FIG. 2, N of N coding means 301 to 30N is 2 (N = 2), that is, two coding means ( The encoding processing operation in the case where the encoding unit 301 and the encoding unit 302 are assumed to be described will be described.
[0094]
4A shows an example of encoded data encoded by the encoding unit 301, and FIG. 4B shows an example of encoded data encoded by the encoding unit 302. As described in the prior art, the squares F1, F2,... Schematically show data for each frame of the moving image, and the squares F1, F2,. It is called F2, ... Also, thick frames F1 and F5 indicate frame data (I picture) that can be decoded alone. Further, the other narrow frame rectangles (frames F2, F3, F4, F6, F7, etc.) indicate frames (P pictures) made of differential data that cannot be decoded by themselves. In addition, a curved arrow a connecting these squares indicates a dependency relationship between the data.
[0095]
In the example shown in FIGS. 4A and 4B, the encoding unit 301 and the encoding unit 302 encode the same encoding target data, but have different encoding parameters. Here, as shown in FIGS. 4A and 4B, the encoding parameters are set so that the position of a frame that can be decoded independently, that is, the position of the I picture is different. In the example of FIG. 4, in the encoding means 301, the I picture is frames F1, F5,..., And in the encoding means 302, the I picture is frames F3, F7,. The setting is such that the position of the I picture is different.
[0096]
FIG. 5 explains the decoding operation on the data receiving apparatus side when the encoded data as shown in FIG. 4 is input. FIG. 5 shows a case where N in the N decoding means 401 to 40N is 2 (N = 2), that is, there are two decoding means (decoding means 401 and decoding means 402). Will be described with respect to the decryption processing in the decryption means 401 and the decryption means 402.
[0097]
Here, the encoded data encoded by the encoding unit 301 (FIG. 4A) is input to the decoding unit 401 and the encoded data encoded by the encoding unit 302 (FIG. 4B). )) Is input to the decoding means 402. In the example of FIG. 5, it is assumed that a frame with a cross in the figure has lost a packet. At this time, it is assumed that the decoding unit 401 is initially set as the priority decoding unit in the priority decoding unit storage unit 51.
[0098]
First, the frame F1 is received by the decoding means 401 and 402. As can be seen from FIG. 5A, the decoding means 401, which is the priority decoding means at this time, detects the packet loss of the frame F1. Not. Therefore, in this case, the decoded data of the decoding unit 401 which is the priority decoding unit is selected, and the decoded data becomes the data output (corresponding to steps S1, S2, S3 and S4 in FIG. 3). .
[0099]
Subsequently, the frame F 2 is received by the decoding means 401 and 402. In this case, as can be seen from FIG. 5 (a), the frame F2 is not detected by the decoding means 401, which is the priority decoding means at this time, in the same way as the frame F1. Therefore, also for this frame F2, the decoded data of the decoding means 401 which is the priority decoding means is selected, and the decoded data becomes the data output (in steps S1, S2, S3 and S4 in FIG. 3). Corresponding to processing).
[0100]
Subsequently, the frame F3 is received by the decoding means 401 and 402. As can be seen from FIG. 5A, the frame F3 is lost in the decoding means 401, which is the priority decoding means at this time. This is detected by the data error detecting means 42.
[0101]
As a result, the decoding means selection means 52 checks whether there is another decoding means that can properly decode the frame F3, and as a result, the decoding means 402 can perform appropriate decoding. Is detected, the content stored in the priority decoding means storage means 51 is updated by the decoding means 402 as the priority decoding means, and the decoded data by the decoding means 402 is selected for this frame F3. Thus, the decoded data becomes a data output (corresponding to the processes of steps S5, S6, and S4).
[0102]
Subsequently, the frame F4 is received by the decoding means 401, 402. As can be seen from FIG. 5B, the packet loss of the frame F4 is detected in the decoding means 402, which is the priority decoding means at the present time. Not detected. Therefore, in this case, the decoded data of the decoding unit 402, which is the priority decoding unit, is selected, and the decoded data becomes the data output (corresponding to the processes of steps S2, S3, and S4).
[0103]
Subsequently, the frame F5 is received by the decoding means 401 and 402, and the decoding means 402 which is the current priority decoding means, as can be seen from FIG. Since no packet loss is detected by the decoding means 402, the frame F5 is decoded by the decoding means 402, the decoded data that has been decoded is selected, and the decoded data becomes the data output. (Corresponding to steps S2, S3 and S4).
[0104]
Subsequently, the frame F6 is received by the decoding means 401, 402. In this case, the decoding means 402, which is the priority decoding means at the present time, detects that the frame F6 is lost as can be seen from FIG. 5B. At this time, if the decoding means selection means 52 detects that the decoding means 401 can properly decode the frame F6, the contents stored in the priority decoding means storage means 51 are prioritized by the decoding means 401. Updated as the decoding means, the decoded data by the decoding means 401 is selected for this frame F6, and the decoded data becomes the data output (corresponding to the processing of steps S5, S6, S4).
[0105]
Subsequently, the frame F7 is received by the decoding means 401, 402. As can be seen from FIG. 5 (a), the decoding means 401, which is the priority decoding means at the present time, receives the packet in the decoding means 401. Since no loss is detected, the decoded data by the decoding unit 401 is output as data for the frame F7 (corresponding to the processing of steps S2, S3, and S4).
[0106]
The decoding process is performed in the above procedure. Eventually, in this example, the decoded data is output in the order shown by the broken line A in FIGS. 5 (a) and 5 (b). That is, the decoded data of the decoding unit 401 is output for the frames F1 and F2, the decoded data of the decoding unit 42 is output for the frames F3, F4, and F5, and the decoding data of the decoding unit 401 is output for the frames F6 and F7. As the data is output, depending on the packet loss status of each frame sent sequentially from the data transmitting device side, the decoding data of the decoding means that can properly decode is dynamically switched and selected And output it.
[0107]
As a result, in this example, all the frames are properly decoded without packet loss, whereby the original data (herein, moving image data) can be properly restored.
As is clear from the above description, the first embodiment has the following effects.
[0108]
The data transmission apparatus according to the first embodiment distributes encoding target data (for example, data for one frame of moving image data) to a plurality of pieces, and the distributed encoding target data is respectively encoded by corresponding encoding means. Encoding is performed independently, and a plurality of encoded data obtained thereby are transmitted as encoded data of a plurality of channels.
[0109]
That is, the encoding means on the data transmission side are independent from each other, and, for example, the technique described in Patent Document 2 described above is not different from what is cited as a conventional example. That is, conventional techniques can be used as they are as individual encoding means. This is an important factor for maintaining compatibility with the prior art.
[0110]
The operation on the data transmission side simply continues to output data and does not depend on the number of data reception sides. Therefore, it is possible to easily perform broadcast-type communication in which a large number of data receiving sides exist. In addition, since the data transmitting device and the data receiving device do not perform handshake processing, the data delay caused thereby can be minimized. As is obvious, there is no load for handshaking with a large number of data receiving apparatuses.
[0111]
This also has the effect of not degrading the encoding efficiency. For example, in the technique described in Patent Document 2 that receives a NACK signal from the receiving side at the time of packet loss, encoding is performed using data that is temporally separated as reference data. However, in this method, the correlation between the data is small. Thus, the encoding efficiency is deteriorated. On the other hand, in the present invention, when encoding is performed, only neighboring data is used as reference data. Therefore, the correlation between data is expected to be high, and the encoding efficiency is not deteriorated.
[0112]
Next, the operation on the data transmitting apparatus side of the first embodiment simply outputs the encoded data and does not depend on the number on the data receiving apparatus side. Accordingly, it is possible to easily perform broadcast-type communication in which a large number of data receiving apparatuses exist.
[0113]
Next, in the data receiving apparatus, as in the data transmitting apparatus, as the encoded data decoding means 43 used in the respective decoding means 401 to 40N, the same one as in the conventional example can be used. This means that data transmitted by the data transmitting apparatus of the present invention can be received by a data receiving apparatus not corresponding to the present invention.
[0114]
That is, if the data receiving apparatus (see FIG. 2) shown in the first embodiment of the present invention is used, the processing as described in the flowchart of FIG. 3 and FIG. 4 and FIG. 5 can be performed. It is possible to receive real-time data without generating it.
[0115]
On the other hand, when the conventional data receiving apparatus that is not the data receiving apparatus of the present invention as shown in FIG. 2 is used, the processing as described in the flowchart of FIG. 3 and FIG. 4 and FIG. When packet loss is detected by one decoding means, it is impossible to dynamically select and output the decoded data from other decoding means that can be properly decoded, but the conventional decoding process remains as it is It can be carried out. Thus, the present invention can use conventional decoding means as the individual decoding means 401 to 40N. This is an important factor for maintaining compatibility with the prior art.
[0116]
As described above, in the first embodiment, a very large effect can be obtained by making a slight change to the conventional technique.
[0117]
[Embodiment 2]
FIG. 6 is a block diagram showing a configuration example of Embodiment 2 of the data transmission apparatus of the present invention. FIG. 6 is obtained by adding data reduction means 11 to the configuration of the data transmission apparatus of FIG. 1 used for the description of the first embodiment. The other components are the same as those in FIG. The same reference numerals are given. The data reduction unit 11 is provided for one encoding unit in the second embodiment, but may be provided for a plurality of encoding units.
[0118]
In FIG. 1 used for the description in the first embodiment, an example in which the number of encoding means is N of the encoding means 301 to 30N is shown. In FIG. In this example, three encoding units 301, 302, and 303 are provided, and the data reduction unit 11 is provided on the input side of the encoding unit 303.
[0119]
That is, the data to be encoded (for example, one frame of moving image data) input to the data input means 1 is distributed to N (three in the case of FIG. 6) by the data distribution means 2, and each of the distributed data is distributed. Is input to the encoding means 301 and 302 as it is, and is input to the encoding means 303 as reduced data through the data reduction means 11. The other components in FIG. 6 are the same as those in FIG. 2, and thus the same parts are denoted by the same reference numerals.
[0120]
The data reduction unit 11 may be one having a function of reducing the size of image data such as moving image data as processing target data. In addition, music data, audio data, and the like may have a function of converting a sampling rate low, or may have a function of converting multi-channel data to monaural data.
[0121]
FIG. 7 is a block diagram showing a configuration example of Embodiment 2 of the data receiving apparatus of the present invention. FIG. 7 is obtained by adding an output size acquisition unit 12 and a decoded data enlargement / reduction unit 13 to the configuration of the data receiving apparatus of FIG. 2 used for the description of the first embodiment. The same parts are denoted by the same reference numerals.
[0122]
In FIG. 2 used for the description in the first embodiment, an example in which the number of decoding means is N among the decoding means 401 to 40N is shown. In FIG. An example in which the number of converting means includes three decoding means 401, 402, and 403 is shown.
[0123]
The data processing procedure in the data receiving apparatus shown in FIG. 7 is basically the same as the decoding procedure (the flowchart in FIG. 3) of the data receiving apparatus shown in FIG. 2 in the first embodiment. One point is different.
[0124]
The first difference is the part related to the initialization processing of the priority decoding means storage means 51 in the processing shown in the flowchart of FIG. 3 used in the description of the first embodiment.
FIG. 8 is a flowchart for explaining step S1 in the flowchart shown in FIG. 3, that is, the initialization process of the priority decoding means storage means 51.
[0125]
In the initialization processing of the priority decoding means storage means 51 in the second embodiment, according to FIG. 8, it is first determined whether or not there is an output size designation (step S11), and if there is an output size designation. Decoding means having an output size that is the same as or closest to the specified output size (output size acquired by the output size acquisition means 12) is stored in the priority decoding means storage means 51. (Step S12). If no output size is specified, an arbitrary decoding unit is set as the priority decoding unit in the priority decoding unit storage unit 51 (step S13).
[0126]
That is, in the first embodiment, an arbitrary decoding unit is set as the initial value, but in the second embodiment, the designated output size, that is, the output size acquisition unit 12 acquires the specified output size. A decoding unit having an output size that is the same as or closest to the output size is set. If no output size is specified, an arbitrary decoding unit is set as in the first embodiment.
[0127]
The second difference is the part related to data output (step S4 in FIG. 3) in the processing shown in the flowchart of FIG. 3 used in the description of the first embodiment. In the second embodiment, the processing relating to the data output is performed as shown in FIG.
[0128]
In FIG. 9, first, it is determined whether or not there is an output size designation (step S21). If there is an output designation, the output size of the priority decoding means stored in the priority decoding means storage means 51 is designated output. It is determined whether or not it is the same as the size (data size designated by the output size acquisition means 12) (step S22). If not, the decoded data enlargement / reduction means 13 adjusts the size of the decoded data. (Step S23), the decoded data after the adjustment is output (Step S24).
[0129]
If the output size is not specified in step S21, or the output size of the priority decoding unit stored in the priority decoding unit storage unit 51 in step S22 is the specified output size (the data acquired by the output size acquisition unit 12). If it is equal to (size), it is output without performing the process of adjusting the output size of the decoded data (step S24).
[0130]
Thus, in the second embodiment, when data output is performed, if the output size of the decoded data selected by the decoding means selection means 52 and the output size obtained by the output size acquisition means 12 are different, The decoded data is adjusted in size so as to be the output size acquired by the decoded data enlargement / reduction means 13.
[0131]
A third difference is processing related to updating of the priority decoding means. That is, in the case of the second embodiment, as the priority decoding means, a decoding means having an output size suitable for the output size obtained from the output size acquisition means 12 is set as an initial value. When packet loss occurs in the means, decoded data with an output size different from the output size obtained from the output size acquisition means 12 may be selected as an exceptional decoded data selection process.
[0132]
Even when such exceptional decoded data selection processing is performed, the decoding means selection means 52 outputs the decoded data having different output sizes as in the case where the exceptional decoded data selection processing is not performed. The priority decoding means is set as the priority decoding means in the priority decoding means storage means 51 (step S6 in FIG. 3). When this exceptional decoded data selection processing is performed, priority decoding with different output sizes is performed. After selecting the decoded data from the means, the priority decoding means set in the priority decoding means storage means 51 is changed to a decoding means having an output size acquired by the output size acquisition means 12 (as an initial value). A process of returning to the set decoding means) is performed. Note that this processing procedure is not shown.
[0133]
10 and 11 are for explaining the decoding process of the second embodiment. In the second embodiment, as described in FIG. 6, the data reduced by the data reduction unit 11 is given to the encoding unit 303, and the encoded data from the encoding unit 303 is The encoded data corresponds to the reduced data, and the encoded data is input to the decoding unit 403.
[0134]
In FIGS. 10 and 11, the reduced data is encoded so that it can be decoded independently in all frames (all frames are I pictures) in consideration of the size of the data. It is supposed to be. Each reduced frame is represented as frames s1, s2, s3,.
[0135]
First, FIG. 10 will be described. FIG. 10 shows an example in which the decoding unit 401 is set as an initial value in the priority decoding unit storage unit 51. That is, in this case, the decoding means having the same output size (or the closest output size) as the output size acquired by the output size acquisition means 12 are the decoding means 401 and the decoding means 42, Of the decoding units 401 and 402, the decoding unit 401 is set as a priority decoding unit.
[0136]
Here, as shown in FIG. 10A, in the encoded data input to the decoding means 401, packet loss (marked with x in the figure) occurs in the frames F3 and F6. As shown in FIG. 6B, in the encoded data input to the decoding means 42, it is assumed that packet loss (marked with x in the figure) occurs in the frames F2 and F6. . As shown in FIG. 10C, it is assumed that no packet loss has occurred in the encoded data input to the decoding means 403. Since the processing procedure for each frame has been described in detail in the first embodiment, it will be described here in a simplified manner.
[0137]
First, the frame F1 is received by the decoding means 401, 402, 403, and the decoding process is started. Then, since the decoding unit 401 initially set as the priority decoding unit does not detect the packet loss of the frame F1, the decoded data of the decoding unit 401 is selected. Similarly, in the frame F2, since the packet loss of the frame F1 is not detected in the decoding unit 401 initially set as the priority decoding unit, the decoded data of the decoding unit 401 is selected.
[0138]
Subsequently, the frame F3 is received by the decoding means 401, 402, 403. As can be seen from FIG. 10A, the packet 401 of the frame F3 is received by the decoding means 401, which is the priority decoding means at the present time. Loss is detected.
[0139]
As a result, the decoding means selection means 52 checks whether there is another decoding means that can properly decode the frame F3, and as a result, the decoding means 402 can perform appropriate decoding. Is detected, the content stored in the priority decoding means storage means 51 is updated by the decoding means 402 as the priority decoding means, and the decoded data of the decoding means 402 is selected for this frame F3. Thus, the decoded data becomes the data output. The decoding means 403 can also decode the frame F3. In this case, the decoding means 402 having the same output size as the requested output size (the output size acquired by the output size acquisition means 12) is used. Prioritize.
[0140]
Subsequently, the frame F4 is received by the decoding means 401, 402, and 403. As can be seen from FIG. 10B, the packet of the frame F4 is received by the decoding means 402, which is the priority decoding means at the present time. Loss is not detected. Therefore, in this case, the decoded data of the decoding unit 402 which is the priority decoding unit is selected, and the decoded data becomes the data output. Similarly, since the packet loss of the frame F5 is not detected in the decoding unit 402 set as the priority decoding unit at that time, the decoded data in the decoding unit 402 is selected.
[0141]
In the processing so far, the required output size (the output size acquired by the output size acquisition unit 12) is the same as the output size of the decoding unit 401 and the decoding unit 42. Size adjustment (enlargement or reduction) is not performed on the encoded data output from 401 and 402.
[0142]
Subsequently, the frame F6 is received by the decoding means 401, 402, and 403. The decoding means 402, which is the priority decoding means at the present time, as shown in FIG. The packet loss is detected by the decoding means 402 which is the converting means.
[0143]
Thereby, the decoding means selection means 52 checks whether there is another decoding means that can properly decode the frame F6. As a result, this frame F6 is also detected by the decoding means 401, and in this case, the decoding means 403 detects that proper decoding is possible. As a result, the content stored in the priority decoding means storage means 51 is updated by the decoding means 403 as the priority decoding means, and the decoded data of the frame s6 of the decoding means 403 is selected for this frame F6. The decoded data becomes the data output.
[0144]
At this time, since the encoded data input to the decoding means 403 is encoded data of reduced size, the decoded data obtained by decoding the data is of course the required data size. Since the output size is smaller than the output size (the output size acquired by the output size acquisition unit 12), the output size is adjusted (in this case, the expansion process) by the decoded data expansion / reduction unit 13, and the expansion process is performed. As a result, the decoded data having the same output size as the requested output size (the output size acquired by the output size acquisition means 12) is output as the data output.
[0145]
Then, the next frame F7 is received by the decoding means 401, 402, 403, and it is determined whether or not a packet loss is detected by the priority decoding means at this time. Here, at this time point, the decoding means 403 is set as the priority decoding means in the priority decoding means storage means 51. In the examples so far, the decoding means 403 is processed as the priority decoding means. However, in the case of the second embodiment, since the required output size is specified, the decoded data at the requested output size (the output size acquired by the output size acquisition means 12) is output. It is necessary to let
[0146]
That is, in the example of FIG. 10, since the output size that is the same as the requested output size is the decoded data of the decoding unit 401 and the decoding unit 402, the decoded data (reduced size) of the decoding unit 403 is reduced. In this case, it can be said that the decoded data selection process is an exceptional decoded data selection process when the decoded data of the decoding unit 401 or the decoding unit 402 cannot be used.
[0147]
Therefore, when such exceptional decoded data selection processing is performed, as described above, the decoding means selection means 52 selects the decoded data by the decoding means having different output sizes, and then A process of returning the priority decoding means stored in the priority decoding means storage means 51 to the decoding means having the required output size (the output size acquired by the output size acquisition means 12) is performed.
[0148]
That is, in the case of FIG. 10, after the decoded data by the decoding means 403 is output as an exceptional decoded data selection process, the priority decoding means stored in the priority decoding means storage means 51 is the original output. Returning to the decoding means 401 or the decoding means 402 having the size, the process for the next frame F7 is performed.
[0149]
In the case of FIG. 10, both the decoding means 401 and 402 have detected a packet loss in the frame F6. Here, since the frame F7 of the decoding unit 401 is a P picture and the frame F7 of the decoding unit 402 is an I picture, the decoded data of the decoding unit 402 is adopted, and the decoded data is output as data. The Since the decoded data of the frame F7 by the decoding unit 402 is the same as the required output size, there is no need to adjust (enlarge or reduce) the size of the output data.
[0150]
As described above, in the example of FIG. 10, the decoded data is selected and output in the order shown by the broken line A. That is, the decoded data of the decoding unit 401 is selected for the frames F1 and F2, the decoded data of the decoding unit 402 is selected for the frames F3, F4, and F5, and the frames corresponding to the frame F6 are exceptionally decoded. The decoded data of the frame s6 of the encoding unit 403 is selected, and the decoded data of the decoding unit 402 is selected for the frame F7.
[0151]
FIG. 11 shows an example in which the decoding unit 403 is selected as the initial value of the priority decoding unit storage unit 51. That is, FIG. 11 shows an example in which the decoding unit 403 is selected as the decoding unit having the same output size (or the closest output size) as the output size acquired by the output size acquisition unit 12.
[0152]
Here, as shown in FIG. 11, in the encoded data input to the decoding means 401, as in the case of FIG. 10, packet loss (x mark in the figure) is added to the frames F3 and F6. In the encoded data input to the decoding means 402, packet loss (marked with x in the figure) occurs in the frames F2 and F6, and the example of FIG. In FIG. 5, it is assumed that packet loss (marked with x in the figure) occurs in the frames s3 and s4 in the encoded data input to the decoding means 403.
[0153]
In the example of FIG. 11, as indicated by the broken line A in FIG. 11, the data up to the frames s1 and s2 is decoded by the decoding means 403 which is the priority decoding means. Up to this point, since the required output size is the same as the output size of the decoding unit 403, the output data size is not adjusted (enlarged or reduced). Thereafter, it is detected that the frame s3 has lost a packet.
[0154]
At this time, as a result of checking whether or not there is a decoding unit capable of decoding the frame corresponding to the frame s3, in this case, the frame F3 also loses packet loss in the encoded data input to the decoding unit 401. Therefore, it is determined that the frame F3 of the decoding unit 402 can be decoded, and the decoded data of the frame F3 of the decoding unit 402 is selected.
[0155]
Since the decrypted data of the decrypting means 402 is larger than the required output size in the example of FIG. 11, the output size is adjusted (in this case, reduction processing) and requested. The decoded data that has been reduced to the output size (the output size acquired by the output size acquiring unit 12, that is, the same size as the decoded data of the decoding unit 403) becomes the output data. Note that the priority decoding means at this point is updated to the decoding means 402, but as described above, in this case, the decoding of the decoding means 402 having a different output size is performed as an exceptional decoded data selection process. Data is selected.
[0156]
When such exceptional decoded data selection processing is performed, as described above, after the decoded data is selected by the decoding means having different output sizes, it is stored in the priority decoding means storage means 51. The priority decoding means is returned to the decoding means having the required output size (the output size acquired by the output size acquisition means 12).
[0157]
That is, in the case of FIG. 11, after the decoded data by the decoding unit 402 is output as an exceptional decoded data selection process, the priority decoding unit stored in the priority decoding unit storage unit 51 is the original output. Returning to the decoding means 403 having the size, the decoding means 403 becomes the current priority decoding means, and the process for the next frame s4 is performed.
[0158]
As a result, the decoding unit 403, which is the current priority decoding unit, detects that the frame s4 is lost. In this case as well, the frame F4 of the decoding unit 402 is an exceptional decoded data selection process. Will be selected. At this time, since the encoded data input to the decoding unit 402 is larger than the required output size, a reduction process is performed, and the decoded data subjected to the reduction process becomes output data.
[0159]
Then, as described above, after the decoded data by the decoding unit 402 as an exceptional decoded data selection process is output, the priority decoding unit stored in the priority decoding unit storage unit 51 is initialized. Returning to the decoding means 403, processing for the next frame s5 is performed.
[0160]
Since no packet loss is detected in the frame s5, the frame s5 is decoded as it is. Since all the encoded data input to the decoding means 403 is an I picture that can be decoded independently, it is possible to decode only that frame without depending on the state of the immediately preceding frame.
[0161]
Therefore, in the example of FIG. 11, as shown by the broken line A, the decoding order is the frames s1 and s2 of the decoding unit 403, and the frames F3 and F4 of the decoding unit 402 (reduction processing is performed on these). And the order of frames s5, s6, and s7 of the decoding means 403.
[0162]
As is clear from the above description, the second embodiment has the following effects in addition to the effects described in the first embodiment.
First, since decoding can be performed using encoded data that is closest to the size required by the application, data quality degradation can be minimized.
[0163]
Next, when using particularly reduced data, for example, a sample display or a moving image thumbnail display can be easily realized in an image. In realizing this, the second embodiment may receive the reduced encoded data sent from the data transmitting apparatus on the data receiving apparatus side, and the data receiving apparatus only decodes it. Therefore, the processing load on the data receiving apparatus side can be reduced and the load on the network can be reduced. In addition, since a thumbnail image or the like can be viewed by the data receiving device, the data communication system is easy to use from the user side.
[0164]
The present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the gist of the present invention. For example, in the first embodiment and the second embodiment described above, moving image data has been described as an example of real-time data to be processed. However, the present invention is not limited to moving image data. Can be similarly implemented.
[0165]
In the case of music or voice, the encoding unit (encoding target data) to be processed is data of a predetermined time length set in advance. The reduced data in the second embodiment can be used for displaying thumbnail images in the case of image data such as a moving image, but can be used for outputting preview data in the case of music or voice. it can.
[0166]
Further, the present invention can create a processing program in which the processing procedure for realizing the present invention described above is described, and the processing program can be recorded on a recording medium such as a floppy disk, an optical disk, a hard disk, The present invention also includes a recording medium on which the processing program is recorded. Further, the processing program may be obtained from a network.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a first embodiment of a data transmission apparatus of the present invention.
FIG. 2 is a block diagram showing a configuration of a first embodiment of a data receiving apparatus of the present invention.
FIG. 3 is a flowchart showing a decoding processing procedure of the data receiving apparatus of the present invention.
FIG. 4 is a diagram illustrating an example of an encoding process when there are two encoding units as an example of describing an encoding process in the data transmission apparatus of the present invention.
FIG. 5 is a diagram illustrating an example of a decoding process (decoding process including a packet loss) when there are two decoding units as an example of the decoding process in the data receiving apparatus of the present invention.
FIG. 6 is a block diagram showing a configuration of a second embodiment of the data transmitting apparatus of the present invention.
FIG. 7 is a block diagram showing a configuration of a data receiving apparatus according to a second embodiment of the present invention.
FIG. 8 is a diagram in which processing procedures are added so that processing relating to initialization of preferential decoding means storage means in the flowchart of FIG. 3 can be applied to the second embodiment;
FIG. 9 is a diagram in which processing procedures are added so that processing relating to data output in the flowchart of FIG. 3 can be applied to the second embodiment;
FIG. 10 is a diagram for explaining an example of a decoding process (decoding process including packet loss) according to the second embodiment;
FIG. 11 is a diagram for explaining another example of decoding processing (decoding processing including packet loss) according to the second embodiment;
FIG. 12 is a diagram for explaining conventional data communication.
FIG. 13 is a diagram for explaining data communication including conventional packet loss.
FIG. 14 is a diagram for explaining conventional data communication including packet loss.
FIG. 15 is a diagram for explaining data communication including conventional packet loss.
FIG. 16 is a block diagram showing a configuration of a conventional data transmission apparatus.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Data input means, 2 Data distribution means, 301-30N encoding means, 401-40N decoding means, 5 Decoding means management means, 51 Preferential decoding means Storage means, 52 Decoding means selection means, 6 Data output means DESCRIPTION OF SYMBOLS 10 Data communication path, 11 Data reduction means, 12 Output size acquisition means, 13 Decoded data expansion / reduction means, 31 Data encoding means, 32 Encoded data transmission means, 41 Encoded data reception means, 42 Data error detection Means (packet loss detection means), 43 encoded data decoding means, F1, F2,... Frame, s1, s2,.

Claims (7)

In a data transmitting apparatus that encodes encoding target data input along a time series and outputs encoded data,
Data distribution means for distributing the data to be encoded into a plurality of data;
Provided for each of the encoding target data distributed by the data distribution unit, each of the input encoding target data is independently encoded, and each of the encoding target data obtained by the encoding processing A plurality of encoding means capable of transmitting encoded data;
The data transmission device, wherein the plurality of encoding units perform encoding processing on each of the encoding target data by changing an encoding parameter between the encoding units. A data receiving device that decodes the encoded data transmitted from the data transmitting device according to claim 1 and outputs decoded data,
A plurality of decoding means capable of receiving and decoding each of the encoded data transmitted from the data transmitting device, and detecting a data error of the received encoded data;
A decoding means management means capable of setting any one of the plurality of decoding means as a priority decoding means and capable of selecting the decoded data of any one of the plurality of decoding means; Have
The decoding means management means monitors the data error detection state in the decoding means set as the priority decoding means, and if there is no data error, selects and outputs the decoded data in the priority decoding means If there is a data error, it is checked whether there is any other decoding means that can properly decode the encoded data, and the decoded data of the other decoding means that can properly decode the decoded data. A data receiving apparatus that selects and outputs the data.   When the decoding means management means determines that there is another decoding means capable of properly decoding the encoded data having the data error, the decoding means management means updates the priority decoding means to the decoding means. The data receiving device according to claim 2. 4. The data receiving apparatus according to claim 2 , wherein the data error is a lack of encoded data corresponding to the encoding target data. In a data transmission method for encoding encoding target data input along a time series and outputting the encoded data,
Distributing the encoding target data to a plurality of;
The dispensing each of the coded data was by the input to the plurality of data encoding means provided for each of each of these coded data, independently among the plurality of data encoding means, and respectively The encoding target data is encoded by changing encoding parameters between the encoding means, and the encoded data obtained thereby is transmitted.
A data transmission method characterized by the above. A data receiving method for decoding the encoded data transmitted by the data transmitting method according to claim 5 and outputting decoded data,
One of a plurality of decoding means capable of receiving and decoding each of the transmitted encoded data and capable of detecting a data error of the received encoded data Is set as the preferred decoding means,
The data error detection state in the decoding means set as the priority decoding means is monitored, and if there is no data error, the decoded data in the priority decoding means is selected and output, and there is the data error. In this case, it is checked whether there is another decoding means that can properly decode the encoded data, and the decoded data of the other decoding means that can be appropriately decoded is selected and output.
A data receiving method. A data transmitting apparatus that encodes data to be encoded that is input along a time series and outputs encoded data; and the encoded data transmitted from the data transmitting apparatus is received, decoded, and decoded data In a data communication system having a data receiving device that outputs
The data transmission device is the data transmission device according to claim 1, and the data reception device is the data reception device according to any one of claims 2 to 4. Data communication system.

Images