Pulse or digital communications – Bandwidth reduction or expansion
Reexamination Certificate
1998-05-27
2002-05-21
Kelley, Chris (Department: 2213)
Pulse or digital communications
Bandwidth reduction or expansion
C375S240000, C370S487000, C370S498000, C370S527000
Reexamination Certificate
active
06393055
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a method of switching from a first video sequence to a second one, both old and new sequences being composed of pictures of type T=I, P or B according to the fact that said pictures are independently coded, or predicted from earlier I or P pictures, or bidirectionally predicted from earlier and later P pictures and/or I pictures. It also relates to a device for implementing said switching method and to a decoding system including such a device. This invention may be used for example for replacing a national television programme by a local one, and is particularly useful in relation with the MPEG-2 standard.
The principles of said standard, adopted by the Moving Pictures Experts Group (MPEG) of the International Standard Organization (ISO) and described in numerous documents, are recalled in detailed manner for instance in “Switching facilities in MPEG-2 necessary but not sufficient”, SMPTE Journal, Vol.104, n
o
12, Dec. 1995, pp. 788-802.
As explained in that document, switching means the splicing of programme segments or complete programmes (including video, audio and related data), i.e in fact the concatenation of different elementary streams. In the following, for sake of simplicity, the description will be limited to the case of only two different bitstreams, that may have been generated at different locations and/or at different times, and were not a priori intended to be spliced together when they were generated.
Concerning the structure itself of these bitstreams, it must be recalled that a compressed video sequence consists of groups of pictures (GOPs) comprising different types of coded pictures I, P and B. The I pictures Cintra coded) are independently coded (by intra-frame techniques only, with no need for previous information, i.e without reference to other pictures), while the P pictures (predictive-coded) are predicted from earlier I pictures or from earlier P ones, and the B pictures (bidirectionally predictive coded) are bidirectionally predicted from earlier and later P pictures and/or I pictures.
According to these different types of prediction, pictures are coded with more or less efficiency (I pictures are the least efficiently coded ones, P pictures are more efficient, and B pictures are the most efficient in their coding). Since the number of bits necessary to code the pictures is not constant, as the decoding is done at a regular period T(V), the bits of the coded pictures do not spend the same time in the decoder's buffer.
The MPEG standard specifies an abstract model of decoding called
V
ideo
B
uffer
V
erifier (VBV) and used to verify that an MPEG bitstream is decodable with reasonable buffering and delay requirement: by way of the buffer occupancy, constraints are imposed to the bitstream, so that decoding may occur without buffer underflow or overflow. The so-called vbv_delay then represents the time spent in said decoder's buffer by the first bit of a coded picture. Performing a switch with a picture regularly decoded at said period T(V) is not possible if the vbv_delays of the first “old” picture suppressed and the first “new” picture transmitted are different (one calls the old pictures the ones which will be replaced by the new ones from a given point: this point, called the splice, is the boundary between data belonging to the old bitstream and data belonging to the new one). Moreover, since the sizes of the pictures replaced and the pictures inserted are different, a decoder's buffer underflow or overflow may occur. A simple switch between two coded video sequences is therefore generally not possible.
In order to avoid any presentation discontinuity at the display side (no apparent interruption in the flow of the data keeps the viewers absorbed on a continuous basis), it is necessary to have no decoding discontinuity, which is possible only if the splice is seamless, i.e. if a smoothly flowing bitstream is obtained. This implies that the decoding time of the first access unit of the new stream is consistent with respect to the decoding time of the access unit of the old stream preceding the splice, which is generally not the case except at favourable locations that are called splicing points and are indicated in particular fields of the bitstreams. Such an access point is located between two coded pictures and defines a fixed standardized vbv_delay value of the first bit following such a point, also called splice decoding delay (the construction of a splicing point also guarantees that no overflow is generated at the transition between old and new sequences). A switching operation can then be performed only if the same splicing points are present on both new and old video sequences. However, currently these points are purely formal, only few splice decoding delay values are defined, and no coding systems are able to perform them. Moreover, in case of video sequence replacement (i.e. of two successive video switchings, with a first forward transition from the old sequence to the new one, and a second backward transition from the new sequence back to the old one), splicing points can be used only if the temporal position of the old and new splicing points corresponding to the backward switching are predictable and compatible. This implies regular and standard GOP structure for the two video sequences, but this requirement is generally not fulfilled.
SUMMARY OF THE INVENTION
A first object of the invention is therefore to propose a method that allows video sequence switching without the constraints of splicing point creation.
To this end the invention relates to a method such as described in the preamble of the description and moreover characterized in that an additional sequence of k pictures is inserted at the switching point between the two sequences, k having a value sufficient in order to have compatible sequences and said additional pictures being coded with a few number of bits.
According to said method, the new video sequence is delayed by k periods T(V) until vbv_delays are compatible. This delay creates for the decoder a lack of k pictures between the old and the new video sequence, and the additional sequence of k pictures takes this place. The decoder will now continuously and successfully display the pictures of the old sequence, the k minimal pictures, and the pictures of the new sequence.
Another object of the invention is to propose a device for implementing said switching method.
To this end the invention relates, in a decoding system for decoding n parallel video bitstreams corresponding to video sequences of pictures said intra if they have been coded without any reference to any other picture or said predicted or interpolated according to the fact that they are predicted by motion-compensated prediction from an earlier intra or predicted picture or bidirectionally interpolated from an earlier and a later picture, or in association with such a system, to a switching device comprising selecting means provided for a selective switchover of the bitstream which has to be decoded according to a switching method such as hereinabove described.
Another object of the invention is also to propose a decoding system including such a switching device.
To this end the invention relates to a decoding system for decoding n parallel video bitstreams corresponding to video sequences of pictures said intra if they have been coded without any reference to any other picture or said predicted or interpolated according to the fact that they are predicted by motion-compensated prediction from an earlier intra or predicted picture or bidirectionally interpolated from an earlier and a later picture, said system comprising a switching device that includes in cascade selecting means, for the selection of the bitstream which has to be decoded, and decoding means for decoding said selected bitstream, wherein said selecting means comprise, in view of the implementation of a method as hereinabove described, control means for a selective switchover of the bitstream which has
An Shawn S.
Gross Russell
Kelley Chris
Koninklijke Philips Electronics , N.V.
LandOfFree
Method of switching video sequences and corresponding... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of switching video sequences and corresponding..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of switching video sequences and corresponding... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2905878