Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
1999-09-02
2003-12-02
Chung, Phung M. (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
C714S710000
Reexamination Certificate
active
06658618
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This application relates to the field of digital data coding and compression and more particularly to the field of coding digital data for video conferencing in a manner that reduces the amount of data provided in response to errors in received data.
2. Description of Related Art
A video signal includes data representing a temporal succession of still images which are referred to as “frames”. If the time between each frame is small enough and the frames are displayed in temporal sequence, the video signal simulate motion. For example, the NTSC television standard provides for transmitting a video signal with thirty frames per second. On a conventional U.S. television set which receives NTSC video signals, the display is thus updated thirty times per second to provide a simulated moving image.
Transmitting video signals over data channels having relatively low band width provides challenges. Generally, there exists a tradeoff between the number of frames per second which may be provided and the resolution of each frame. Increasing the resolution of each frame increases the amount of data corresponding to each frame. Thus, for relatively high resolution frames, the number of frames per second which may be transmitted in a low band width channel is smaller than the number of frames which may be transmitted per second in connection with low resolution frames.
One solution for improving the performance of video signals transmitted over relatively low bandwidth channels is to eliminate the need to transmit each frame completely. Instead, what is transmitted is other information that allows reconstruction of each frame at the receiving end. For example, the H.263 ITU standard provides for transmitting a first frame in its entirety and then transmitting one or more motion vectors and prediction error information for subsequent frames. The motion vector(s) represent translation of portion(s) of the image from a reference frame to a current frame while the prediction error information represents the difference between the current frame and the reference frame with the motion vector information applied thereto. Thus, each video frame may be determined by applying the motion compensation vector to the reference frame and then applying the prediction error information. The video signal can then be transmitted by sending the first frame followed by motion vectors and prediction error information for each of the subsequent frames.
The H.263 standard also allows for relative coding of a portion of a frame, which is called a “macroblock.” In the case of using a macroblock, the vector and error information may be provided relative to the macroblock, as opposed to an entire frame. In addition, instead of transmitting an entire new reference frame, it may only be necessary to transmit a new macroblock, which may then be used for subsequent relative encoding with motion vector(s), error compensation, etc.
A drawback to the H.263 technique for encoding video information is that, since each macroblock is based on the motion vectors and prediction error information applied to the previous macroblock, then a transmission error or data loss causes all subsequent macroblocks to be suspect. Recovery from such an error may be performed by the transmitter resending a new macroblock. For transmission channels with a relatively low error rate, this does not pose a significant problem. An enhanced version of the H.263 standard, deemed H.263+ or H.263 version 2, provides a mechanism to indicate to the transmitter that the received signal contains errors. However, the H.263+ standard has limitations, some of which are described in annex N.5 of the H.263+ standard specification.
Thus, when the receiver receives motion vector information and prediction error information that is incorrect, the receiver sends a signal back to the transmitter to request that an entire reference frame be sent so that error recovery is possible.
SUMMARY OF THE INVENTION
According to the present invention, transmitting and receiving video data includes an encoder storing macroblocks in a plurality of reference buffers, the encoder transmitting macroblocks that are encoded relative to each of the reference buffers, a decoder receiving the macroblocks, and in response to the decoder indicating an error with respect to a particular one of the macroblocks encoded relative to a first one of the buffers, the encoder using a second one of the buffers for error recovery if there is no error associated with the second buffer. In response to the decoder indicating an error with respect to a particular one of the macroblocks and all of the buffers having errors associated therewith, the encoder may transmit at least one macroblock that is encoded independent of any other macroblocks. The decoder may receive a macroblock having an error associated therewith and, in response to all buffers of the decoder having errors associated therewith, the decoder may transmit a request to the encoder to transmit at least one macroblock that is encoded independent of any other macroblocks. In response to the request, the encoder may transmit at least one macroblock that is encoded independent of any other macroblocks.
The encoder using a second one of the buffers for error recovery may include coding a macroblock using a second one of the buffers as a reference and marking the first one of the buffers as having an error. In response to marking the first one of the buffers as having an error, a timer may be set to indicate an amount of time since the error occurred and, in response to the amount of time since the error occurred being greater than a predetermined threshold, the error associated with the first one of the buffers may be cleared. The predetermined threshold may correspond to a round trip transmission and processing time between the encoder and the decoder.
A number of buffers used may be increased in response to errors requiring the encoder to transmit at least one macroblock that is encoded independent of any other macroblocks and decreased if, after a predetermined amount of time, no errors requiring the encoder to transmit at least one macroblock that is encoded independent of any other macroblocks occur. Each macroblock may be associated with a temporal reference and a table may be provided that maps each temporal reference number with one of: a particular one of the buffers and a value indicating that a macroblock corresponding to a temporal reference number is not being stored.
According further to the present invention, transmitting video data includes an encoder storing macroblocks in a plurality of reference buffers, the encoder transmitting macroblocks that are encoded relative to each of the reference buffers, and, in response to an error with respect to a macroblock corresponding to a first one of the buffers, the encoder using a second one of the buffers for error recovery if there is no error associated with the second buffer. In response to an error with respect to a macroblock and all of the buffers having errors associated therewith, the encoder may transmit at least one macroblock that is encoded independent of any other macroblocks. In response to a receiving a request, the encoder may transmit at least one macroblock that is encoded independent of any other macroblocks. The encoder using a second one of the buffers for error recovery may includes coding a macroblock using a second one of the buffers as a reference and marking the first one of the buffers as having an error. In response to marking the first one of the buffers as having an error, a timer may be set to indicate an amount of time since the error occurred and, in response to the amount of time since the error occurred being greater than a predetermined threshold, the error associated with the first one of the buffers may be cleared. The predetermined threshold may correspond to a round trip transmission and processing time between the encoder and the decoder. number of buffers used may be increased
Gu Qunshan
Root Timothy
Chung Phung M.
Polycom, Inc.
Wong, Cabello, Lutsch, Rutherford & Brucculeri L.L.P.
LandOfFree
Error recovery method for video compression coding using... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Error recovery method for video compression coding using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Error recovery method for video compression coding using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3143035