# US 2004/0001546 A1 · Spatiotemporal prediction for bidirectionally predictive (B) pictures and motion vector prediction for multi-picture reference motion compensation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2004/0001546 A1 - Document ID: US20040001546A1 - Status: Published application - Kind: A1 - Application number: 10/444,511 - Filed: 2003-05-23 - Published: 2004-01-01 - Earliest claimed: 2002-06-03 - Inventors: Alexandros Tourapis; Shipeng Li; Feng Wu - Web page: https://patentgrep.com/patent/US20040001546A1 ## CPC - H04N 19/577: Motion compensation with bidirectional frame interpolation, i.e. using B-pictures - H04N 19/102: characterised by the element, parameter or selection affected or controlled by the adaptive coding - H04N 19/105: Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction - H04N 19/107: between spatial and temporal predictive coding, e.g. picture refresh - H04N 19/137: Motion inside a coding unit, e.g. average field, frame or block difference - H04N 19/142: Detection of scene cut or scene change - H04N 19/176: the region being a block, e.g. a macroblock - H04N 19/51: Motion estimation or motion compensation - H04N 19/513: Processing of motion vectors - H04N 19/52: by predictive encoding - H04N 19/56: Motion estimation with initialisation of the vector search, e.g. estimating a good candidate to initiate a search - H04N 19/573: Motion compensation with multiple frame prediction using two or more reference frames in a given prediction direction - H04N 19/58: Motion compensation with long-term prediction, i.e. the reference frame for a current frame not being the temporally closest one (H04N19/23 takes precedence) - H04N 19/593: involving spatial prediction techniques - H04N 19/61: in combination with predictive coding - H04N 19/87: involving scene cut or scene change detection in combination with video compression - H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION ## Abstract Several improvements for use with Bidirectionally Predictive (B) pictures within a video sequence are provided. In certain improvements Direct Mode encoding and/or Motion Vector Prediction are enhanced using spatial prediction techniques. In other improvements Motion Vector prediction includes temporal distance and subblock information, for example, for more accurate prediction. Such improvements and other presented herein significantly improve the performance of any applicable video coding system/logic. ## Claims ### Claim 1 A method for use in encoding video data within a sequence of video frames, the method comprising: identifying at least a portion of at least one video frame to be a Bidirectionally Predictive (B) picture; and selectively encoding said B picture using at least spatial prediction to encode at least one motion parameter associated with said B picture. ### Claim 2 (depends on 1) The method as recited in claim 1, wherein said B picture includes a macroblock. ### Claim 3 (depends on 2) The method as recited in claim 2, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter produces a Direct Macroblock. ### Claim 4 (depends on 1) The method as recited in claim 1, wherein said B picture includes a slice. ### Claim 5 (depends on 1) The method as recited in claim 1, wherein said B picture includes at least a portion of a macroblock. ### Claim 6 (depends on 1) The method as recited in claim 1, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter further includes employing linear motion vector prediction for said B picture based on at least one reference picture that is at least another portion of said video frame. ### Claim 7 (depends on 1) The method as recited in claim 1, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter further includes employing non-linear motion vector prediction for said B picture based on at least one reference picture that is at least another portion of said video frame. ### Claim 8 (depends on 1) The method as recited in claim 1, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter further includes employing median motion vector prediction for said B picture based on at least two reference pictures that are both portions of said video frame. ### Claim 9 (depends on 1) The method as recited in claim 1, wherein said at least one motion parameter includes at least one motion vector. ### Claim 10 (depends on 1) The method as recited in claim 1, wherein at least one other portion of at least one other video frame is processed to further selectively encode said B picture using temporal prediction to encode at least one temporal-based motion parameter associated with said B picture. ### Claim 11 (depends on 10) The method as recited in claim 10, wherein said temporal prediction includes bidirectional temporal prediction. ### Claim 12 (depends on 10) The method as recited in claim 10, wherein said at least one other video frame is a Predictive (P) frame. ### Claim 13 (depends on 10) The method as recited in claim 10, further comprising selectively scaling said at least one temporal-based motion parameter based at least in part on a temporal distance between said other video frame and said frame that includes said B picture. ### Claim 14 (depends on 13) The method as recited in claim 13, wherein temporal distance information is encoded within a header associated with said encoded B picture. ### Claim 15 (depends on 10) The method as recited in claim 10, wherein said at least one other portion includes at least a portion of a macroblock within said at least one other video frame. ### Claim 16 A computer-readable medium having computer implementable instructions for configuring at least one processing unit to perform acts comprising: accessing data for a sequence of video frames; identifying at least a portion of at least one video frame to be a Bidirectionally Predictive (B) picture; and selectively encoding said B picture using at least spatial prediction to encode at least one motion parameter associated with said B picture. ### Claim 17 (depends on 16) The computer-readable medium as recited in claim 16, wherein said B picture includes a macroblock. ### Claim 18 (depends on 17) The computer-readable medium as recited in claim 17, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter produces a Direct Macroblock. ### Claim 19 (depends on 16) The computer-readable medium as recited in claim 16, wherein said B picture includes a slice. ### Claim 20 (depends on 16) The computer-readable medium as recited in claim 16, wherein said B picture includes at least a portion of a macroblock. ### Claim 21 (depends on 16) The computer-readable medium as recited in claim 16, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter further includes employing linear motion vector prediction for said B picture based on at least one reference picture that is at least another portion of said video frame. ### Claim 22 (depends on 16) The computer-readable medium as recited in claim 16, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter further includes employing non-linear motion vector prediction for said B picture based on at least one reference picture that is at least another portion of said video frame. ### Claim 23 (depends on 16) The computer-readable medium as recited in claim 16, wherein selectively encoding said B picture using at least spatial prediction to encode said at least one motion parameter further includes employing median motion vector prediction for said B picture based on at least two reference pictures that are both portions of said video frame. ### Claim 24 (depends on 16) The computer-readable medium as recited in claim 16, wherein said at least one motion parameter includes at least one motion vector. ### Claim 25 (depends on 1) The computer-readable medium as recited in claim 1, wherein at least one other portion of at least one other video frame is processed to further selectively encode said B picture using temporal prediction to encode at least one temporal-based motion parameter associated with said B picture. ### Claim 26 (depends on 25) The computer-readable medium as recited in claim 25, wherein said temporal prediction includes bidirectional temporal prediction. ### Claim 27 (depends on 25) The computer-readable medium as recited in claim 25, wherein said at least one other video frame is a Predictive (P) frame. ### Claim 28 (depends on 25) The computer-readable medium as recited in claim 25, having computer implementable instructions for configuring said at least one processing unit to perform acts comprising: selectively scaling said at least one temporal-based motion parameter based at least in part on a temporal distance between said other video frame and said frame that includes said B picture. ### Claim 29 (depends on 28) The computer-readable medium as recited in claim 28, wherein temporal distance information is encoded within a header associated with said encoded B picture. ### Claim 30 (depends on 25) The computer-readable medium as recited in claim 25, wherein said at least one other portion includes at least a portion of a macroblock within said at least one other video frame. ### Claim 31 An apparatus for use in encoding video data within a sequence of video frames, the apparatus comprising: logic operatively configured to access video data for a sequence of video frames, identify at least a portion of at least one video frame to be a Bidirectionally Predictive (B) picture, and selectively encode said B picture using at least spatial prediction to encode at least one motion parameter associated with said B picture. ### Claim 32 (depends on 31) The apparatus as recited in claim 31, wherein said B picture includes a macroblock. ### Claim 33 (depends on 32) The apparatus as recited in claim 32, wherein said logic selectively encodes said B picture using at least spatial prediction to encode said at least one motion parameter to produce a Direct Macroblock. ### Claim 34 (depends on 31) The apparatus as recited in claim 31, wherein said B picture includes a slice. ### Claim 35 (depends on 31) The apparatus as recited in claim 31, wherein said B picture includes at least a portion of a macroblock. ### Claim 36 (depends on 31) The apparatus as recited in claim 31, wherein said logic is further configured to employ linear motion vector prediction for said B picture based on at least one reference picture that is at least another portion of said video frame. ### Claim 37 (depends on 31) The apparatus as recited in claim 31, wherein said logic is further configured to employ non-linear motion vector prediction for said B picture based on at least one reference picture that is at least another portion of said video frame. ### Claim 38 (depends on 31) The apparatus as recited in claim 31, wherein said logic is further configured to employ median motion vector prediction for said B picture based on at least two reference pictures that are both portions of said video frame. ### Claim 39 (depends on 31) The apparatus as recited in claim 31, wherein said at least one motion parameter includes at least one motion vector. ### Claim 40 (depends on 31) The apparatus as recited in claim 31, wherein said logic is further configured to process at least one other portion of at least one other video frame is and selectively encode said B picture using temporal prediction to encode at least one temporal-based motion parameter associated with said B picture. ### Claim 41 (depends on 40) The apparatus as recited in claim 40, wherein said temporal prediction includes bidirectional temporal prediction. ### Claim 42 (depends on 40) The apparatus as recited in claim 40, wherein said at least one other video frame is a Predictive (P) frame. ### Claim 43 (depends on 40) The apparatus as recited in claim 40, wherein said logic is further configured to selectively scale said at least one temporal-based motion parameter based at least in part on a temporal distance between said other video frame and said frame that includes said B picture. ### Claim 44 (depends on 43) The apparatus as recited in claim 43, whereinsaid logic is further configured to include temporal distance information within a header associated with said encoded B picture. ### Claim 45 (depends on 40) The apparatus as recited in claim 40, wherein said at least one other portion includes at least a portion of a macroblock within said at least one other video frame. ### Claim 46 A method for encoding video data, the method comprising: identifying at least a portion of at least one video frame to be coded in an enhanced direct mode; and encoding said portion in said enhanced direct mode using at least spatial information associated with said portion within said at least one video frame. ### Claim 47 (depends on 46) The method as recited in claim 46 wherein encoding said portion in said enhanced direct mode further includes using temporal information associated with said portion and at least one other portion of at least one other video frame. ### Claim 48 (depends on 46) The method as recited in claim 46, wherein encoding said portion in said enhanced direct mode further includes using motion vector prediction based on at least one other portion within said at least one video frame. ### Claim 49 The method as recited in Clam 48, wherein said motion vector prediction includes median prediction. ### Claim 50 (depends on 46) The method as recited in claim 46, wherein said enhance direct mode includes using spatial prediction to calculate said spatial information based on at least one linear function that considers motion information of at least one other portion of said at least one video frame. ### Claim 51 (depends on 46) The method as recited in claim 46, wherein said enhance direct mode includes using spatial prediction to calculate said spatial information based on at least one non-linear function that considers motion information of at least one other portion of said at least one video frame. ### Claim 52 A computer-readable medium having computer implementable instructions for configuring at least one processing unit to perform acts comprising encoding video data by identifying at least a portion of at least one video frame to be coded in an enhanced direct mode, and encoding said portion in said enhanced direct mode using at least spatial information associated with said portion within said at least one video frame. ### Claim 53 (depends on 52) The computer-readable medium as recited in claim 52 wherein encoding said portion in said enhanced direct mode further includes using temporal information associated with said portion and at least one other portion of at least one other video frame. ### Claim 54 (depends on 52) The computer-readable medium as recited in claim 52, wherein encoding said portion in said enhanced direct mode further includes using motion vector prediction based on at least one other portion within said at least one video frame. ### Claim 55 The computer-readable medium as recited in Clam 54, wherein said motion vector prediction includes median prediction. ### Claim 56 (depends on 52) The computer-readable medium as recited in claim 52, wherein said enhance direct mode includes using spatial prediction to calculate said spatial information based on at least one linear function that considers motion information of at least one other portion of said at least one video frame. ### Claim 57 (depends on 52) The computer-readable medium as recited in claim 52, wherein said enhance direct mode includes using spatial prediction to calculate said spatial information based on at least one non-linear function that considers motion information of at least one other portion of said at least one video frame. ### Claim 58 An apparatus comprising: logic operatively configured to encode video data by identifying at least a portion of at least one video frame to be coded in an enhanced direct mode, and encode said portion in said enhanced direct mode using at least spatial information Is associated with said portion within said at least one video frame. ### Claim 59 (depends on 58) The apparatus as recited in claim 58 wherein said logic is further operatively configured to use temporal information associated with said portion and at least one other portion of at least one other video frame to encode said portion in said enhanced direct mode. ### Claim 60 (depends on 58) The apparatus as recited in claim 58, wherein said logic is further operatively configured to encoding said portion in said enhanced direct mode further using motion vector prediction information based on at least one other portion within said at least one video frame. ### Claim 61 The apparatus as recited in Clam 60, wherein said motion vector prediction includes median prediction. ### Claim 62 (depends on 56) The apparatus as recited in claim 56, wherein said logic is further operatively configured to use spatial prediction to calculate said spatial information based on at least one linear function that considers motion information of at least one other portion of said at least one video frame. ### Claim 63 (depends on 56) The apparatus as recited in claim 56, wherein said logic is further operatively configured to use spatial prediction to calculate said spatial information based on at least one non-linear function that considers motion information of at least one other portion of said at least one video frame. ### Claim 64 A method to predict a reference picture in direct mode video encoding, the method comprising: selecting a reference picture from a group comprising a minimum reference picture for a plurality of predictions related to at least a portion of a video frame to be encoded, a median reference picture for said plurality of predictions, and a current reference picture based on a single direction prediction; and encoding said at least one portion of said video frame based on selected reference picture. ### Claim 65 (depends on 64) The method as recited in claim 64, wherein selecting said reference picture further includes selecting at least one spatially related prediction. ### Claim 66 (depends on 64) The method as recited in claim 64, wherein selecting said reference picture further includes selecting at least one temporally related prediction. ### Claim 67 A computer-readable medium having computer implementable instructions for configuring at least one processing unit to perform acts comprising: selecting a reference picture from a group comprising a minimum reference picture for a plurality of predictions related to at least a portion of a video frame to be encoded, a median reference picture for said plurality of predictions, and a current reference picture based on a single direction prediction; and encoding said at least one portion of said video frame based on selected reference picture. ### Claim 68 (depends on 67) The computer-readable medium as recited in claim 67, wherein selecting said reference picture further includes selecting at least one spatially related prediction. ### Claim 69 (depends on 67) The computer-readable medium as recited in claim 67, wherein selecting said reference picture further includes selecting at least one temporally related prediction. ### Claim 70 An apparatus comprising: logic that is operatively configured to select a reference picture from a group comprising a minimum reference picture for a plurality of predictions related to at least a portion of a video frame to be encoded, a median reference picture for said plurality of predictions, and a current reference picture based on a single direction prediction, and encode said at least one portion of said video frame based on selected reference picture. ### Claim 71 (depends on 70) The apparatus as recited in claim 70, wherein said logic is operatively configured to select at least one spatially related prediction. ### Claim 72 (depends on 70) The apparatus as recited in claim 70, wherein said logic is operatively configured to select at least one temporally related prediction. ### Claim 73 A method for use in selecting between temporal prediction, spatial prediction, or both temporal and spatial prediction for encoding at least a portion of at least one video frame in an enhanced direct mode, the method comprising: selecting temporal prediction if at least one motion vector of a collocated portion of said video frame is zero; if surrounding portions within said video frame use different reference pictures than a collocated reference picture, then select spatial prediction only; if a motion flow associated with said portion of said video frame is substantially different than a motion flow associated with a reference picture, then select spatial prediction; if temporal prediction of direct mode is signaled inside an image header, then selecting temporal prediction; and if spatial prediction of direct mode is signaled inside said image header, then selecting spatial prediction. ### Claim 74 (depends on 73) The method as recited in claim 73, further comprising: correcting at least one temporally predicted parameter based on spatial information. ### Claim 75 (depends on 73) The method as recited in claim 73, further comprising: correcting at least one spatially predicted parameter based on temporal information. ### Claim 76 A computer-readable medium having computer implementable instructions for configuring at least one processing unit to perform acts comprising: selecting between temporal prediction, spatial prediction, or both temporal and spatial prediction for encoding at least a portion of at least one video frame in an enhanced direct mode, such that: temporal prediction is selected if at least one motion vector of a collocated portion of said video frame is zero, only spatial prediction is selected if surrounding portions within said video frame use different reference pictures than a collocated reference picture, spatial prediction is selected if a motion flow associated with said portion of said video frame is substantially different than a motion flow associated with a reference picture, temporal prediction is selected if temporal prediction of direct mode is signaled inside an image header, and spatial prediction is selected if spatial prediction of direct mode is signaled inside said image header. ### Claim 77 (depends on 76) The computer-readable medium as recited in claim 76, further comprising: correcting at least one temporally predicted parameter based on spatial information. ### Claim 78 (depends on 76) The computer-readable medium as recited in claim 76, further comprising: correcting at least one spatially predicted parameter based on temporal information. ### Claim 79 An apparatus comprising: logic operatively configured to select between and employ temporal prediction, spatial prediction, or both temporal and spatial prediction for encoding at least a portion of at least one video frame in an enhanced direct mode, wherein said logic: selects temporal prediction if at least one motion vector of a collocated portion of said video frame is zero, selects only spatial prediction if surrounding portions within said video frame use different reference pictures than a collocated reference picture, selects spatial prediction if a motion flow associated with said portion of said video frame is substantially different than a motion flow associated with a reference picture, selects temporal prediction if temporal prediction of direct mode is signaled inside an image header, and selects spatial prediction if spatial prediction of direct mode is signaled inside said image header. ### Claim 80 (depends on 79) The apparatus as recited in claim 79, wherein said logic is further operatively configured to correct at least one temporally predicted parameter based on spatial information. ### Claim 81 (depends on 79) The apparatus as recited in claim 79, wherein said logic is further operatively configured to correct at least one spatially predicted parameter based on temporal information. ### Claim 82 A method for use in encoding video data, the method comprising: selecting a reference portion of a future video frame to serve as a B picture to at least one portion of an earlier video frame; using motion vectors associated with said reference frame to calculate motion vectors associated with said at least one portion; and encoding said at least one portion based on said calculated motion vectors associated with said at least one portion. ### Claim 83 (depends on 82) A method as recited in claim 82, wherein using said motion vectors associated with said reference frame to calculate said motion vectors associated with said at least one portion further includes estimating at least one possible prediction for use in direct mode coding by projecting and inverting backward and forward motion vectors of the reference portion. ### Claim 84 (depends on 83) The method as recited in claim 83, wherein encoding said at least one portion based on said calculated motion vectors associated with said at least one portion further includes applying selective projection and inversion based on at least one temporal parameter associated with said reference portion with respect to said at least one portion. ### Claim 85 (depends on 82) The method as recited in claim 82, wherein only one reference portion is used for B pictures when encoding in direct mode. ### Claim 86 (depends on 82) The method as recited in claim 82, wherein encoding said at least one portion based on said calculated motion vectors associated with said at least one portion further includes encoding in a direct mode wherein at least one of said calculated motion vectors is based on at least one projected motion vector that refers to at least two reference portions in two different reference pictures. ### Claim 87 (depends on 82) The method as recited in claim 82, wherein encoding said at least one portion based on said calculated motion vectors associated with said at least one portion further includes encoding in a direct mode wherein at least one of said calculated motion vectors is based on spatial prediction associated with said reference portion. ### Claim 88 A computer-readable medium having computer implementable instructions for configuring at least one processing unit to perform acts comprising: selecting a reference portion of a future video frame to serve as a B picture to at least one portion of an earlier video frame; using motion vectors associated with said reference frame to calculate motion vectors associated with said at least one portion; and encoding said at least one portion based on said calculated motion vectors associated with said at least one portion. ### Claim 89 (depends on 88) A computer-readable medium as recited in claim 88, wherein using said motion vectors associated with said reference frame to calculate said motion vectors associated with said at least one portion further includes estimating at least one possible prediction for use in direct mode coding by projecting and inverting backward and forward motion vectors of the reference portion. ### Claim 90 (depends on 89) The computer-readable medium as recited in claim 89, wherein encoding said at least one portion based on said calculated motion vectors associated with said at least one portion further includes applying selective projection and inversion based on at least one temporal parameter associated with said reference portion with respect to said at least one portion. ### Claim 91 (depends on 88) The computer-readable medium as recited in claim 88, wherein only one reference portion is used for B pictures when encoding in direct mode. ### Claim 92 (depends on 88) The computer-readable medium as recited in claim 88, wherein encoding said at least one portion based on said calculated motion vectors associated with said at least one portion further includes encoding in a direct mode wherein at least one of said calculated motion vectors is based on at least one projected motion vector that refers to at least two reference portions in two different reference pictures. ### Claim 93 (depends on 88) The computer-readable medium as recited in claim 88, wherein encoding said at least one portion based on said calculated motion vectors associated with said at least one portion further includes encoding in a direct mode wherein at least one of said calculated motion vectors is based on spatial prediction associated with said reference portion. ### Claim 94 An apparatus comprising: logic operatively configured to select a reference portion of a future video frame to serve as a B picture to at least one portion of an earlier video frame, use motion vectors associated with said reference frame to calculate motion vectors associated with said at least one portion, and encode said at least one portion based on said calculated motion vectors associated with said at least one portion. ### Claim 95 (depends on 94) The apparatus as recited in claim 94, wherein said logic is further operatively configured to estimate at least one possible prediction for use in direct mode coding by projecting and inverting backward and forward motion vectors of the reference portion. ### Claim 96 (depends on 95) The apparatus as recited in claim 95, wherein said logic is further operatively configured to apply selective projection and inversion based on at least one temporal parameter associated with said reference portion with respect to said at least one portion. ### Claim 97 (depends on 94) The apparatus as recited in claim 94, wherein only one reference portion is used for B pictures when encoding in direct mode. ### Claim 98 (depends on 94) The apparatus as recited in claim 94, wherein said logic is further operatively configured to encode in a direct mode wherein at least one of said calculated motion vectors is based on at least one projected motion vector that refers to at least two reference portions in two different reference pictures. ### Claim 99 (depends on 94) The apparatus as recited in claim 94, wherein said logic is further operatively configured to encode in a direct mode wherein at least one of said calculated motion vectors is based on spatial prediction associated with said reference portion. ### Claim 100 A method for use in determining motion vectors during video encoding, the method comprising: selecting at least three predictors A, B and C that each uses a different reference picture having an associated temporal distance TRA, TRB, and TRC respectively, and a motion vector MVA, MVB, and MVC; and predicting a median motion vector MVpred associated with a current reference picture that has a temporal distance equal to TR. ### Claim 101 (depends on 100) The method as recited in claim 100, wherein said median predictor MVpred is calculated as: 4 MV → pred = TR × Median ⁡ ( MV → A TR A , MV → B TR B , MV → C TR C ) . ### Claim 102 (depends on 100) The method as recited in claim 100, wherein said median predictor MVpred is calculated as: MVpred=Median(Ave({right arrow over (MV)}C1,{right arrow over (MV)}C2),Ave({right arrow over (MV)}A1, {right arrow over (MV)}A2),{right arrow over (MV)}B). ### Claim 103 (depends on 100) The method as recited in claim 100, further comprising: selecting at least a fourth predictor D having an associated temporal distance TRD and a motion vector MVD, and wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median(Median({right arrow over (MV)}C1,{right arrow over (MV)}C1,{right arrow over (MV)}D), . . . Median({right arrow over (MV)}D,{right arrow over (MV)}A1,{right arrow over (MV)}C1),Median({right arrow over (MV)}B,{right arrow over (MV)}A1,{right arrow over (MV)}A2)) ### Claim 104 (depends on 100) The method as recited in claim 100, further comprising: selecting at least a fourth predictor D having an associated temporal distance TRD and a motion vector MVD, and wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median({right arrow over (MV)}C1,{right arrow over (MV)}C2, {right arrow over (MV)}D, {right arrow over (MV)}B,{right arrow over (MV)}A1, {right arrow over (MV)}A2). ### Claim 105 (depends on 100) The method as recited in claim 100, further comprising selectively substituting an adjacent portion of a reference frame for a selected portion of said reference frame for use in determining motion vector prediction when intra coding is used. ### Claim 106 A computer-readable medium having computer implementable instructions for configuring at least one processing unit to perform acts comprising: selecting at least three predictors A, B and C that each uses a different reference picture having an associated temporal distance TRA, TRB, and TRC respectively, and a motion vector MVA, MVB, and MVC; and predicting a median motion vector MVpred associated with a current reference picture that has a temporal distance equal to TR. ### Claim 107 (depends on 106) The computer-readable medium as recited in claim 106, wherein said median predictor MVpred is calculated as: 5 MV → pred = TR × Median ⁡ ( MV → A TR A , MV → B TR B , MV → C TR C ) . ### Claim 108 (depends on 106) The computer-readable medium as recited in claim 106, wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median(Ave({right arrow over (MV)}C1,{right arrow over (MV)}C2), Ave({right arrow over (MV)}A1, {right arrow over (MV)}A2),{right arrow over (MV)}B). ### Claim 109 (depends on 106) The computer-readable medium as recited in claim 106, further comprising: selecting at least a fourth predictor D having an associated temporal distance TRD and a motion vector MVD, and wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median(Median({right arrow over (MV)}C1,{right arrow over (MV)}C2,{right arrow over (MV)}D), . . . Median({right arrow over (MV)}D,{right arrow over (MV)}A1,{right arrow over (MV)}C2),Median({right arrow over (MV)}B,{right arrow over (MV)}A1,{right arrow over (MV)}A2)) ### Claim 110 (depends on 106) The computer-readable medium as recited in claim 106, further comprising: selecting at least a fourth predictor D having an associated temporal distance TRD and a motion vector MVD, and wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median({right arrow over (MV)}C1,{right arrow over (MV)}C2,{right arrow over (MV)}D,{right arrow over (MV)}B,{right arrow over (MV)}A1,{right arrow over (MV)}A2) ### Claim 111 (depends on 106) The computer-readable medium as recited in claim 106, further comprising selectively substituting an adjacent portion of a reference frame for a selected portion of said reference frame for use in determining motion vector prediction when intra coding is used. ### Claim 112 An apparatus comprising logic operatively configured to select at least three predictors A, B and C that each uses a different reference picture having an associated temporal distance TRA, TRB, and TRC respectively, and a motion vector MVA, MVB, and MVC, and predict a median motion vector MVpred associated with a current reference picture that has a temporal distance equal to TR. ### Claim 113 (depends on 112) The apparatus as recited in claim 112, wherein said median predictor MVpred is calculated as: 6 MV → pred = TR × Median ⁡ ( MV → A TR A , MV → B TR B , MV → C TR C ) . ### Claim 114 (depends on 112) The apparatus as recited in claim 112, wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median(Ave({fraction (MV)}C1, {right arrow over (MV)}C2), Ave(MVA1,{right arrow over (MV)}A2),{right arrow over (MV)}B). ### Claim 115 (depends on 112) The apparatus as recited in claim 112, wherein said logic is further operatively configured to select at least a fourth predictor D having an associated temporal distance TRD and a motion vector MVD, and wherein said median predictor MVpred is calculated as: {right arrow over (MV)}pred=Median(Median({right arrow over (MV)}C1,{right arrow over (MV)}C2,{right arrow over (MV)}D), . . . Median({right arrow over (MV)}D,{right arrow over (MV)}A1,{right arrow over (MV)}C2),Median({right arrow over (MV)}B,{right arrow over (MV)}A1,{right arrow over (MV)}A2)) ### Claim 116 (depends on 112) The apparatus as recited in claim 112, wherein said logic is further operatively configured to select at least a fourth predictor D having an associated temporal distance TRD and a motion vector MVD, and wherein said median predictor MVpred is calculated as: {right arrow over (MV)}Pred=Median({right arrow over (MV)}C1, {right arrow over (MV)}C2,{right arrow over (MV)}D,{right arrow over (MV)}B,{right arrow over (MV)}A1,{right arrow over (MV)}A2). ### Claim 117 (depends on 112) The apparatus as recited in claim 112, wherein said logic is further operatively configured to selectively substitute an adjacent portion of a reference frame for a selected portion of said reference frame for use in determining motion vector prediction when intra coding is used.