Richard E Goedeken

US Patent:

20080137736, Jun 12, 2008

Filed:

Jan 17, 2006

Appl. No.:

11/795417

Inventors:

John William Richardson - Crystal Lake IL, US
Richard Edwin Goedeken - Somerset NJ, US
Jens Cahnbley - Princeton JCT NJ, US

Assignee:

Joseph J. Laks, Patent Operations - Princeton NJ

International Classification:

H04N 7/26

US Classification:

37524003, 37524024, 375E07139

Abstract:

There are provided apparatus and methods for parallelizing an encoding process across multiple processors. An apparatus includes a computing resource balancer and a splitter. The computing resource balancer is for assigning encoder instances to respective ones of the multiple processors based on at least one of thread affinity and process affinity. The splitter is for temporally dividing an incoming video sequence into discrete GOPs such that each of the encoder instances is capable of encoding the incoming video sequence in parallel so that each of the discrete GOPs is encoded on a respective one of the multiple processors.

US Patent:

20090003454, Jan 1, 2009

Filed:

Jan 24, 2006

Appl. No.:

11/883209

Inventors:

John William Richardson - Crystal Lake IL, US
Richard Edwin Goedeken - Somerset NJ, US

International Classification:

H04N 7/26

US Classification:

37524024, 375E072

Abstract:

A device incorporates a software-based H.264 video encoder for providing compressed, or encoded, video data. The H.264 encoder incorporates a timer algorithm such that the time available to the H.264 encoder within the macroblock mode decision and motion estimation framework is constrained, or restricted. The particular amount of time available to the H.264 encoder then determines a subset of available encoding modes that the H.264 encoder can use to encode a macroblock.

US Patent:

20110211634, Sep 1, 2011

Filed:

Feb 18, 2011

Appl. No.:

12/932169

Inventors:

Richard Edwin Goedeken - Santa Clarita CA, US
Joan Llach - Cesson-Sevigne Cedex, FR

International Classification:

H04N 7/24
H04N 7/26

US Classification:

37524002, 375E07005, 375E07126

Abstract:

A method and apparatus are disclosed and described for providing Offset Metadata insertion in multi-view coded video. The apparatus includes an offset metadata inserter () for receiving a first bitstream compressed using a single-view based two-dimensional video compression scheme, a second bitstream compressed using a multi-view based two-dimensional video compression scheme, and disparity information, and outputting a third bitstream based on the first bitstream, the second bitstream, and the disparity information. Each of the first bitstream, the second bitstream, and the disparity information correspond to a same video sequence. The third bitstream includes group of pictures information extracted from the first bitstream and underlying content of the video sequence at least some of which is extracted from the second bitstream. The third bitstream has embedded therein one or more messages that specify the disparity information for use in overlaying information on the underlying content of the video sequence during a subsequent displaying of the overlaying information on the underlying content.

US Patent:

20110211643, Sep 1, 2011

Filed:

Feb 16, 2011

Appl. No.:

12/932062

Inventors:

Hua Yang - Plainsboro NJ, US
Richard Edwin Goedeken - Santa Clarita CA, US
Joan Llach - Cesson-Sevigne Cedex, FR

International Classification:

H04N 7/24

US Classification:

37524026, 375E07011

Abstract:

A method and apparatus are disclosed and described for providing bit rate configuration for multi-view video coding. In the video encoder, the method includes encoding image data for at least one picture for at least two joint views of multi-view video content, the at least two joint views including a base view and at least one dependent view. The bit rate configuration for encoding the image data is determined to include an average bit rate and a maximum bit rate for the base view and the average bit rate and the maximum bit rate for the at least two joint views ().

US Patent:

20110211816, Sep 1, 2011

Filed:

Feb 18, 2011

Appl. No.:

12/932156

Inventors:

Richard Edwin Goedeken - Santa Clarita CA, US
Joan Llach - Cesson-Sevigne Cedex, FR

International Classification:

H04N 5/93

US Classification:

386354, 386E05007, 386E05003

Abstract:

A method and apparatus are disclosed and described for providing a synchronized workstation with two-dimensional and three-dimensional outputs. The apparatus includes a video decoder () for decoding picture data. The video decoder includes a data manager () for receiving video production commands and managing a video playback of the picture data in at least one of a two-dimensional video output mode and a three-dimensional video output mode responsive to the video production commands. The two-dimensional video output mode and the three-dimensional video output mode are capable of being used independently and simultaneously.

US Patent:

20110216827, Sep 8, 2011

Filed:

Feb 18, 2011

Appl. No.:

12/932168

Inventors:

Jiancong Luo - West Windsor NJ, US
Wanrong Lin - Belle Meade NJ, US
Richard Edwin Goedeken - Santa Clarita CA, US

International Classification:

H04N 7/32

US Classification:

37524012, 375E07243

Abstract:

A method and apparatus for efficient encoding of multi-view coded video data is provided. The apparatus includes one or more encoders () for encoding image data for a plurality of pictures for at least two views of multi-view video content. The image data is encoded in parallel in a plurality of at least one of threads and processes using a plurality of processors in order to generate a resultant bitstream there from.

US Patent:

20110216836, Sep 8, 2011

Filed:

Feb 16, 2011

Appl. No.:

12/932055

Inventors:

Jiancong Luo - West Windsor NJ, US
Richard Edwin Goedeken - Santa Clarita CA, US
Joan LLach - Cesson-Sevigne Cedex, FR

International Classification:

H04N 7/26

US Classification:

37524025, 375E07027

Abstract:

A method and apparatus are disclosed and described for providing hypothetical reference decoder conformance error detection. The apparatus includes a multi-pass encoder () for encoding pictures in a video sequence. The pictures are encoded in a first pass to determine a bit consumption of each picture, and a bit allocation is controlled during the encoding of the pictures in at least a second pass responsive to satisfying requirements for a subsequent decoding of the bitstream. The requirements relate to preventing at least one of underflow and overflow conditions in a buffer during the subsequent decoding. The bit allocation is based on the bit consumption determined in the first pass and buffer parameters. The pictures are encoded in at least the first and second passes into a plurality of bitstreams on a scene-basis such that any of the pictures belonging to a same scene are respectively encoded in a same one of the bitstreams.

US Patent:

20110216838, Sep 8, 2011

Filed:

Feb 18, 2011

Appl. No.:

12/932179

Inventors:

Wanrong Lin - Belle Meade NJ, US
Richard Edward Goedeken - Cranbury NJ, US
Jiancong Luo - West Windsor NJ, US

International Classification:

H04N 7/26

US Classification:

37524025, 375E07027

Abstract:

A method and apparatus are provided for efficient decoding of multi-view coded video data. The apparatus includes one or more decoders () for decoding image data from a bitstream. The image data corresponds to a plurality of pictures for at least two views of multi-view video content. The image data is decoded in parallel in a plurality of at least one of threads and processes. The image data is decoded using a staggered approach such that a decoding of a particular one of the plurality of pictures in a particular one or more of the plurality of at least one of threads and process commences only when inter-view reference pictures corresponding to the particular one of the plurality of pictures are decoded using one or more other ones of the plurality of at least one of threads and processes.