Patent Application
20050163275
The disclosure of Japanese Patent Application No. 2004-18322 filed on Jan. 27, 2004 including specification, drawings and claims are incorporated herein by reference in its entity.
The present invention relates to a stream decoding system for decoding a compress-coded stream.
In recent years, due to progress of multimedia technology, various types of compress-coding schemes have been proposed. For example, as audio compress-coding schemes, there are MPEG-1/2, DolbyAC-3, DTS, MP3 and the like. Now, the market demands sound reproduction systems with the multiple reproduction function of decoding, when a compress-coded sound stream is reproduced, the sound stream whatever the type of compress-coding scheme is used in compress-coding the sound stream.
To achieve the multiple reproduction function in a sound reproduction system, whether the type of a compress-coding scheme of a compress-coded sound stream can be determined and how smooth switching to corresponding decoding to the compress-coding method of each of the compress-coded sound streams can be performed become keys.
Many compress-coded sound streams have frame structures and many frames of such sound streams are frames with part in which stream information is stored. This part is called a “header”. Moreover, in the beginning of each header, a synchronization word indicating the beginning of a frame is present. This synchronization word takes a various different value depending on the type of an audio compress-coding scheme. For example, as for a sound stream in the MP3 format, a synchronization word having a value of 0xFFF exists. In this manner, a synchronization word indicates the beginning of a frame of a stream and also shows that main information is contained in the subsequent part of the frame. However, in a MP3 format sound stream, the value of 0xFFF might be sometimes contained in main information and thus 0xFFF is not always the synchronization word.
For this reason, normally, if only one synchronization word in a stream has been searched, it can not be determined that synchronization has been achieved. For example, for a MP3 format stream, many reproduction systems use the following procedures to determine that synchronization has been achieved.
First, when a value of 0xFFF, i.e., a value considered to be a synchronization word is detected from an input stream, header information following 0xFFF is analyzed and a frame size is calculated. Then, the location of a subsequent synchronization word is calculated from the frame size and whether or not 0xFFF is present in the location is examined. Only after the presence of 0xFFF has been confirmed, it is determined that synchronization is achieved. Herein, “synchronization is achieved” means that the value of 0xFFF is assured to be a synchronization word and the location of beginning of a frame is determined. After synchronization of a stream has been assured by such a method, the stream is decoded (see Japanese Laid-Open Publication No. 8-204691).
Hereinafter, a known stream decoding system will be described with reference to the accompanying drawings.
Next, the location of a synchronization word of a stream is searched by a synchronization word analysis section 102. If 0xFFF is detected, header information following 0xFFF is analyzed by a header analysis section 103, a frame size is calculated, and the location of a subsequent synchronization word is estimated. Then, if the same synchronization word is present in the estimated location, it is determined that this 0xFFF is a synchronization word and the streams is decoded by a decoding section 104 based on a result of the analysis of the header information. If no synchronization word is present at the estimated location, it is determined that this 0xFFF is not a synchronization word and the process returns to the synchronization word analysis section 102. After synchronization is achieved and a resultant frame from decoding is output from a signal output section 105, the process returns to the header analysis section 103 and the same processing is repeated until decoding is performed to a final frame.
As described above, decoding is performed to a stream by repeating the processing in which a frame size is calculated for each frame, a location in which a subsequent synchronization word is present is estimated, the presence of the same synchronization word in the estimated location is confirmed, and then decoding is performed.
However, with the known stream decoding system, when a stream including a mixture of multiple types of streams is to be decoded, at a portion of the stream in which a type of the stream is changed to another, a subsequent synchronization word is different to the previous synchronization word to the subsequent synchronization word. Thus, in this portion, an error tends to occur and the stream can not be decoded in a normal state. Accordingly, reproduction is forcefully terminated or wrong reproduction is performed.
To solve the above-described problem, according to the present invention, a stream decoding system for decoding a compress-coded stream includes: a signal input section for inputting the stream having a header including a synchronization word; a synchronization word analysis section for detecting the synchronization word from the input stream and determining a type of the synchronization word; a plurality of header analysis sections, connected in parallel to one another, for analyzing header information following the synchronization word; a decoding system switching section, connected between the synchronization word analysis section and the plurality of header analysis sections, for selectively outputting the stream to one of the header analysis sections corresponding to the type of the synchronization word; a plurality of decoding sections for decoding the stream based on the header information analyzed in said one of the header analysis sections; and a signal output section for outputting the decoded stream.
Moreover, according to the present invention, the synchronization word analysis section may further have the function of determining the type of a compress-coding scheme by which the input stream is coded.
Furthermore, according to the present invention, the stream decoding system may further include a subsequent synchronization word analysis section for calculating a location of a synchronization word in a subsequent frame based on the information analyzed by one of the header analysis sections and analyzing a subsequent synchronization word in the location.
With the stream decoding system of this embodiment, even a stream including multiple types of streams can be reproduced without causing forced termination of reproduction or wrong reproduction. Therefore, reliability of the system can be improved.
Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
A signal input section 301 inputs a stream having a header including a synchronization word indicating the beginning of a frame to a synchronization word analysis section 302.
The synchronization word analysis section 302 detects a synchronization word from one of frames of the input stream and analyzes the type of the synchronization word. In this case, the type of the synchronization word is determined by confirming that the detected synchronization word has a pattern matching that of a synchronization word of a sound compression scheme by which a decodable stream by the stream decoding system of this embodiment is coded. Then, a result of synchronization word determination and the frame are output to a decoding system switching section 306.
The decoding system switching section 306 is connected to three header analysis sections 303 connected in parallel to one another and selectively outputs the frame to one of the header analysis sections 303 corresponding to the type of the synchronization word. Moreover, the header analysis sections 303 are for analyzing header information following the synchronization word.
A decoding section 304 is connected in series to an associated one of the header analysis sections 303 and performs to the frame decoding corresponding to a sound compression scheme of the frame based on the header information analyzed by the associated one of the header analysis sections 303.
The signal output section 305 outputs data decoded by the decoding section 304.
Next, procedures of stream decoding in a stream decoding system with the above-described configuration will be described.
First, in the synchronization word analysis section 302, a synchronization word is detected from one of frames of a stream input from the signal input section 301 and the type of the synchronization word is determined. Then, the frame is selectively output by the decoding system switching section 306 to one of the header analysis sections 303 corresponding to the type of the synchronization word and header information is analyzed in the corresponding header analysis section 303.
In the decoding section 304, based on the analyzed header information, decoding is performed to the frame and decoded data is output from the signal output section 305.
Furthermore, after decoding to one of frames of the stream and data output are completed, the process returns to the synchronization word analysis section 302 and the above-described processing is repeatedly performed to a subsequent frame. Thus, decoding is performed to each of all frames in the stream.
Accordingly, in EMBODIMENT 1, even when a subsequent synchronization word is different from a previous synchronization word in a portion of a stream including a mixture of multiple types of streams in which a type of the stream is changed to another, all synchronization words are detected and then frame decoding is performed using respective ones of the header analysis sections 303 and the decoding sections 304 which corresponding to the type of each of the detected synchronization words. Thus, the generation of a stream error in the portion in which the type of the stream is changed can be prevented, so that normal reproduction can be performed.
Note that in EMBODIMENT 1, three header analysis section 303 and three decoding section 304 are provided. However, the numbers of the header analysis sections and the decoding sections are not limited thereto but the numbers can be set to be a different number according to the types of sound compression schemes of a decodable stream. Moreover, needless to say, the stream decoding system of this embodiment is applicable not only to decoding of a sound stream but also decoding of a video image stream or a complex stream of sound and video image streams. This point is the same for EMBODIMENT 2.
Specifically, in EMBODIMENT 2, a subsequent synchronization word analysis section 407 described later is connected between header analysis sections 403 and decoding sections 404. Other than that, the stream decoding system of EMBODIMENT 2 has a similar configuration to that of the stream decoding system of EMBODIMENT 1.
The subsequent synchronization word analysis section 407 calculates a frame size from header information analyzed in an associated one of the header analysis sections 403 and estimates the location of a subsequent synchronization word by calculation from a result of the frame size calculation. Also, the subsequent synchronization word section 407 compares a value in the estimated location of the subsequent synchronization word to a value of a previous synchronization word. If these values match with each other, the subsequent synchronization word section 407 determines that synchronization is achieved and then stores the subsequent synchronization word.
Next, procedures of stream decoding in a stream decoding system with the above-described configuration will be described.
First, in a synchronization word analysis section 402, a synchronization word is detected from one of frames of a stream input from a signal input section 401 and the type of the synchronization word is determined. Then, by a decoding system switching section 406, the frame is selectively output to one of the header analysis sections 403 corresponding to the type of the synchronization word and header information is analyzed in the header analysis section 403.
Then, in the subsequent synchronization word analysis section 407, a frame size is calculated based on the analyzed header information to estimate the location of a subsequent synchronization word, and a subsequent synchronization word in the location is compared to the previous synchronization word to determine whether synchronization is achieved. If it is determined that synchronization is achieved, the subsequent synchronization word is stored.
Next, in one of the decoding sections 404, decoding is performed to the frame, based on the analyzed header information and decoded data is output from a signal output section 405.
Furthermore, after decoding of one of frames in the stream and data output are completed, the process returns to the synchronization word analysis section 406 and, by the decoding system switching section 406, a subsequent frame is selectively output to one of the header analysis sections 403 corresponding to the type of the subsequent synchronization word stored in the subsequent synchronization word analysis section 407. Thereafter, by repeating the same processing, decoding is performed to each of all frames in the stream.
Moreover, when trash data (non-relevant data) is present in last part of a frame, the trash data is present in the position in which a subsequent synchronization word is estimated to exist by the subsequent synchronization word analysis section 407 and therefore it is determine that a subsequent synchronization is not present therein. In this case, decoding is not performed to the frame.
When such a stream error occurs, the process returns to the synchronization word analysis section 402 and synchronization word search is performed in order from the location of the stream in which the trash data is present. Thus, synchronization words are continuously detected.
In EMBODIMENT 2, even when a synchronization word is different from a previous synchronization word in a portion of a stream including a mixture of multiple types of streams in which a type of the stream is changed to another, all synchronization words are detected and then decoding is performed to the frame using respective ones of the header analysis sections 403 and the decoding sections 404 corresponding to the type of each of the detected synchronization words. Thus, the occurrence of a stream error in the portion in which the type of the stream is changed can be prevented, so that normal reproduction can be performed.
Moreover, detection of a synchronization word by the synchronization word analysis section 402 is performed to only initial one of frames of each type, so that a data processing amount for synchronization word detection in frames can be reduced.
Moreover, in a general stream decoding system, processing in which a frame size is calculated from header information to estimate the location of a subsequent synchronization word and then whether or not the same synchronization word is present is confirmed is performed in many cases. Thus, the same processing as that of the subsequent synchronization word analysis section 407 can be performed only by adding the function of increasing the number of synchronization word patterns of sound compression schemes by which decodable sound streams are coded and a function of storing the subsequent synchronization word. Therefore, simple implementation becomes possible.
The present invention is to be a necessary function where a single stream includes a mixture of multiple types of streams and exhibits great industrial applicability.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2004-018322 | Jan 2004 | JP | national |
