TW200840368A - Techniques for content adaptive video frame slicing and non-uniform access unit coding - Google Patents

Abstract

Techniques for content adaptive video frame slicing and non-uniform access unit coding for improved coding efficiency are provided. An encoder and decoder are disclosed to process (encode or decode) a single non-uniform video access unit (VAU) employing flexible macroblock ordering (FMO) in conjunction with different slice coding types in response to global motion detection of a camera pan or a scroll within the single VAU.

Description

200840368 IX. Description of the invention: [Technical field to which the invention pertains] 'and more specifically regarding the use of the encoding of the access unit to achieve the generalization of the present invention regarding the video editing of the content of the video frame cutting and non- A good coding efficiency technology. The U.S. Provisional Application is hereby incorporated by reference. This application claims the benefit of the benefit of the benefit of the benefit of the benefit of the benefit of

U [Prior Art] In all current video reduction standards, the coded representation of the video or video access unit (VAU) is included as a slice of the next lower layer in the coding level. The slice layer allows the grouping of functions of integer macroblocks (data) in the video frame. The packet is functionally used as a resynchronization unit in the coded representation of the frame. To use as a positive resynchronization point, all predictive coding schemes/dependent schemes such as intraframe prediction (based on neighboring pixels) and motion vector prediction are discontinued across all slice boundaries. Prior to Η·264 (and excluding optional, attachment κ: Η·263 + slice structuring mode 'rectangular slice submode'), previously such as 261.261, MPECM, MPEG-2/H.262, H Video compression standards such as .263 and MPEG-4 support a slice structure consisting essentially of an integer number of consecutive macroblocks (in raster scan order), where the limited slice size is slightly different. Knife—The =264 standard introduces the concept of a "slice group" that enables the division of a macroblock of a frame into multiple slice groups and multiple slice groups in a completely arbitrary manner. The slices are, and therefore are not constrained by, having to be contiguous in raster scan order. The arbitrary decomposition is described by the so-called "slice ^ 127986.doc 200840368 group mapping" to describe the data transmitted to the solution (FMO) 除 'decoding frame. The code is called flexible. Slice group mapping macro block sorting

L therefore there is a need for techniques for adapting the content of video frame splicing and coding of non-access units to achieve improved coding efficiency. SUMMARY OF THE INVENTION A number of techniques are provided for content-enhanced video frame splicing and non- _ access unit coding to achieve improved coding efficiency. Provided herein is a device including a processor that operates to divide into a plurality of slice groups and slices, the content is tunable, and the one or more slice coding types are implemented in an early VAU. The encoding of the Video Access Unit (VAU). In one embodiment, a memory is coupled to the processor. In a bad example, the present invention provides an encoding device including an encoding engine operable to combine a single video access unit (VAu) in response to panning of a camera or global motion detection of a scroll. Different slice coding types are used to apply flexible macro block ordering (FM〇). In another example, an encoding device includes an encoding engine operative to combine in response to one or more changes in a composite scene - different slice encoding types in a single video access unit (VAU) Apply Flexible Macro Block Sorting (FM0) 'where the one or more changes affect one or more portions of the video frame instead of the entire video frame. The one or more changes may include cropping scene changes, intersecting and fades, fade in or fade out, zooming in or out, and global motion diversification such as panning or scrolling. In another aspect, a decoding device including a decoding engine is provided. The decoding engine is operative to apply a flexible macroblock ordering (FMO) to decode the single non-coherently encoded VAU in conjunction with a slice encoding type of 127986.doc -6 - 200840368 in a single video access unit (vau). . In another configuration, a computer program product is provided comprising a computer readable medium comprising instructions for processing multimedia material. These instructions enable a computer to use a flexible macroblock ordering (FM0) implementation to divide the frame of content into a plurality of slice groups and slices. The instructions also cause the computer to perform non-conforming VAU encoding of the divided frames using one or more slice encoding types.

In a further configuration, a computer program product comprising a computer readable medium is provided, the computer readable medium comprising a fingerprint for processing multimedia material. The instructions cause a computer to apply the flexible macroblock ordering (FM〇) to a different non-coherently encoded VAU in conjunction with a different slice encoding type in a single video access unit (VAU). The techniques described herein provide a method of encoding video access units using multiple slice types to achieve enhanced coding efficiency. In the detailed description, especially when combined with the accompanying drawings, it will be easier to understand the additional aspects. [Embodiment] As an example, an example is considered to be more than other configurations "engine", "machine", and in this article, the term "institutive" is used to mean, or to clarify ". The configurations or designs described herein are not necessarily or preferably not preferred, and the terms "core,", "processor" and "processing unit" are used interchangeably. The definitions of the scope and in this description refer to the use of the phase sample configuration. However, the disclosure covers many different forms. 127986.doc 200840368 can be used in a series of pictures, frames and / or domain to characterize the video signal, any of the pictures, frames and / or domains can be further included - or multiple slices as used herein, the term 'frame, one can include one or more A broad term for frames, fields, pictures, and/or slices. Groups include systems and methods for facilitating channel switching in multimedia transmission systems. Multimedia data can include - one motion video image, motion video image, still image, text Or any other suitable audiovisual material. Multimedia processing systems such as video encoders are based on standards such as Motion Picture Experts Group (MPEG)-1, _2 and _4, International Telecommunication Union (Ιτυ) _τ h 263 standard, and the ITU-T H.264 standard and its similar standard, IS〇 / IEc MpEG_ 4 $ 10 knife, that Ge-level video coding (AVC) and other international standards using coding

The law encodes multimedia material, and each of the above-mentioned standards is entirely for reference. The encoding and extension are generally related to compressing multimedia material for transmission and/or storage. Compression can be widely seen as the process of removing redundancy from multimedia material. Let t -, 5 now 'the pictures include frames (a positive picture) or fields (for example - the interlaced video stream includes - the domain of the odd or even lines of the picture interval). Step-by-step, each frame or field can be two-negative = or multiple slices, or sub-parts of the frame or field. The term "frame" as used herein, when used alone or in combination with a word, may refer to a picture frame, a field, or a slice thereof. The video coding method uses a compressed or lossy compression algorithm. The method compresses each frame to compress the video signal.: 127986.doc 200840368 In-frame coding (also referred to as inner (four)) pure-frame-to-frame coding (also referred to as inter-frame coding) refers to other (Reference) frame to encode a frame. Redundancy, wherein the frames that are close to each other in the time series of the frame when the door number 1 is normally displayed have at least some matching or at least partially matching each other. Part of it.

Multimedia A multimedia processor, such as a video encoder, can encode a frame by f-pixel subsets of frames. The subset of pixels may be referred to as a block or inter-block_) and may include, for example, 16 x 16 pixels. The code Η further divides each 16χ16 macroblock into a plurality of sub-blocks. Each can further include additional sub-blocks. For example, a sub-block of a (5) 6 macroblock may include 16X8 and 8X16 sub-blocks. Each of the 16 Χ 8 and 8 Χ 16 sub-blocks may include, for example, a plurality of (four) sub-blocks, which may themselves include, for example, 4 χ 4, 4 χ 8 and 8 Χ 4 sub-blocks, and the like. The term "block" as used herein may refer to a macroblock or sub-block of any size. The encoder uses an algorithm based on inter-frame coding motion compensation to exploit temporal redundancy between successive frames. The motion compensation algorithm identifies at least a portion of the heartbeat-block-reference frame. The block can be shifted in the pure relative to the matching portion of the (equal) reference frame. The shift is characterized by - or multiple motion vectors. The difference between the block and the partially matched portion of the (equal) reference frame can be characterized as - or a plurality of remaining portions. The encoder can encode a frame to include data for one or more motion vectors and the remainder of a particular partition of the frame. A particular block partition can be selected for encoding a frame by approximately minimizing the cost function, the cost function (for example) balancing the 127986.doc for the frame content formed by a code - 9- 200840368 Code size and distortion or perceived distortion. By. Inter-hole coding gives more reduction efficiency than intra-frame coding. However, inter-frame coding can be problematic if the field reference material (such as a reference frame or reference field) is lost due to channel errors and the like. In addition to the loss of reference material due to errors, the reference material may also be unavailable due to the initial acquisition or re-acquisition of the video signal at the inter-frame coded frame. In such cases, it may not be possible to decode the inter-frame coded material or to create undesirable artifacts and errors that can be propagated. Such situations may result in the most common frame form in which the undesired user of the extended time period experiences an encoded frame within an independently decodable frame to enable the video signal to be synchronized/resynchronized. The MpEG_x& H 26x standard uses a content called a picture group (GOP), and the GOP includes an intra-frame coded frame (also referred to as an I-frame) and a temporary prediction p-frame or bi-directional prediction of the reference I frame. The box and/or other P and /*B frames within the GOP. A longer G〇p is desirable for increased compression, but a shorter GOP achieves faster acquisition and/or synchronization/resynchronization. Increasing the number of I-frames will allow faster acquisition and/or synchronization/re-synchronization, but at the expense of lower compression. 1 illustrates a block diagram of an exemplary multimedia communication system 1 根据 in accordance with certain configurations. System 100 includes an encoder device 110 that communicates with a decoder device 150 via a network. In one example, encoder device 11() receives a multimedia signal from an external source 102 and encodes the signal for transmission over network 140. In this example, encoder device 110 includes a processor 112 coupled to a memory 114 and transceiver 116. Processor 112 encodes material from the multimedia 127986.doc -10- 200840368 data source and provides it to transceiver 116 for communication over network i4. In this example, decoder device 150 includes a processor 152 coupled to a memory 154 and transceiver 156. Processor 152 can include one or more general purpose processors and/or a digital signal processor. The memory may include a memory or a disk-based memory. Transceiver 156 is responsive to receive multimedia material over network 14 and provide the multimedia data to processing 152 for decoding. In the example, the transceiver (5) includes a wireless transceiver. Network 140 may include one or more wired or wireless communication systems including one or more of an Ethernet network, a telephone (eg, p〇TS), a cable, a power line, and a fiber optic system and/or a wireless system The wireless system includes a code division multiple access (cdma or CDMA2000) communication system, a frequency division multiple access (FDMA) system, an orthogonal frequency division multiple (OFDM) access system, and a time division multiple access. (TDMA) systems such as GSM/GPRS (General Packet Radio Service) / EDGE (Enhanced Data gSM Environment), a TETRA (Terrestrial Relay Radio) Mobile Phone System, a Wideband Code Division Multiple Access (WCDMA) System, a High One or more of a data transmission rate (1xEV_d〇 or 1xEV-DO Gold Multicast) system, an IEEE 802.11 system, a MediaFLO system, a DMB system, a DVB-Η system, and the like. 2A illustrates a block diagram of an exemplary encoder device 110 that may be used in the system of FIG. 1 in accordance with certain configurations. In this configuration, the code cry 110 includes an inter-frame code encoder component i 18, an intra-frame codec component 120, a reference generator component 122, and a transmitter component 124. Inter-frame code encoder 118 encodes the inter-frame coding portion of the video referenced to other portions of the video material in other time frames 127986.doc -11 · 200840368 for temporal prediction (e.g., using motion compensated prediction). The intra-frame code encoder 12" code can encode the intra-frame code portion of the video that is decoded separately without reference to other video data that is temporally located. In some configurations, intra-frame code encoder 120 may use spatial prediction to utilize redundancy of other video data located in the same time frame. . In the - state, the reference data generator 122 generates information indicating the location of the intra-frame encoded video data and the inter-frame encoded video data generated by the encoders 120 and 118, respectively. For example, the reference material can include identifiers for sub-blocks and/or macroblocks - the decoder uses the identifiers to locate - position in the frame. The reference material can also include a frame number for locating a frame in a sequence of video frames. The transmitter 124 transmits the inter-frame encoded data, the intra-frame encoded data, and the transmitted data in some configurations by a network such as the network 14 of FIG. The (4) can be transmitted on - or a plurality of communication links. The term communication key is used in a generic sense and may include any communication channel including, but not limited to, a wired or wireless network, a virtual channel, an optical link, and the like. In some configurations, the information encoded within the frame is transmitted on the base layer communication link and the inter-frame encoded data is transmitted on the enhancement layer communication link. In some configurations, the data encoded in the frame and the data encoded between the frames are transmitted on the same link.纟 In some configurations, one or more of the inter-frame coded poor material, intra-frame coded data, and reference material may be transmitted on the sideband communication link. For example, t can be used such as η·264 supplementary enhancement information () message or MPEG_2 user_data message side 127986.doc -12- 200840368 ▼ communication key. In the sheep this day, the information of the mother and the husband's test: the data in the two boxes, the frame = and one or more of the reference materials can include multiple data packets by one input. The data package includes a packet that belongs to the virtual channel and is identifiable to the virtual channel. Other forms of the virtual channel are known, such as frequency division, time division, code spreading, and the like.

Figure 2B illustrates a block diagram of an exemplary decoding device 15G in a system (10) configured in accordance with some configurations. In this configuration, the decoder 15 includes a receiver component 158, a selective decoder component (10), a reference beauties determiner component 162, and a channel switching predictor component 164 and - error: One or more reference materials such as device element 166 may be used by the detector. Receiver 158 receives the encoded video material (e.g., the data encoded by encoder 11 in the figure (1)). Receiver 158 can receive the encoded material via a wired or wireless network, such as network 140 of FIG. This data can be received by - or multiple communication links. In some configurations, the encoded data within the frame is received on a base layer communication link and the inter-frame encoded data is received on an enhancement layer communication link. In some configurations, the information in the frame, the flat code > and the data encoded between the frames are received on the same communication link. In some configurations, one or more of the inter-frame coded data, the in-frame coded material, and the reference material may be received on one side of the communication link. For example, a side of the communication link may be used, such as 补充·264 supplemental enhancement information (sei) information or MPEG-2 user-data message. In some groups, one or more of the information encoded in the frame, the information encoded between the frames, and the reference material are transmitted by a virtual channel. A virtual channel can be packaged 127986.doc -13- 200840368 = a poor packet, the data packet contains - the data packet identification 2 belongs to the virtual channel - the identification packet header. Other forms of identifying a virtual channel are known in the art. The selective decompressor 160 decodes the received inter-frame coding and video data in the frame. In some configurations, the received data includes the form of the inter-frame coding of % & μ at a portion of the sfL^ material and the intra-frame coding of one of the video data. After decoding the data encoded by the inter-frame, ▲ multi-tested data, the information encoded between the frames can be decoded. Examples of data encoded using motion correction prediction include identifying the position of the reference material - the motion vector and the frame identifier. If the portion of the frame identified by the motion vector and frame identifier of the inter-frame coding format is available (10) if it has been decoded, the selective decoder 16 can decode the form of the inter-frame code. If the reference material is not available, the selective decoder 160 can decode the form of the intra-frame coding. In the L-sample, the reference breeding determiner 162 identifies the received reference material indicating the location of the encoded and inter-frame encoded material in the received encoded video material. For example, the reference data can include a selective decoder 160 for locating a sub-block of a location and/or an identifier of a macroblock in a frame. The reference material may also include a frame serial number for locating a frame in the video frame sequence. Using the received reference material enables a decoder to determine if the reference material to which the inter-frame encoded data is dependent is available. The availability of reference material can be affected by the user switching one of the channels of a multi-channel communication system. For example, multiple video broadcasts may use one or more of the 127986.doc • 14-200840368 communication links for the receiver 158 to change to a different broadcast channel using the 4❹ command receiver 158 for inter-frame coding. The reference to f material may not be available immediately on the new channel, " 51 51 ★ , , KM (four) (four) 164 6 〇, 'member switch command has been issued, and the selective decoder is sent to Beixun to identify the message. The reference material for the inter-frame coding format is

C, and then identifying the location of the intra-frame coding format, and selectively decoding the identified intra-frame coding format. The availability of the test materials can also be affected by the errors in the received data. Error ❹j 16 6 can use error detection technology (such as forward correction = identify errors that cannot be corrected in the bit stream. If there is an uncorrectable error in the reference data according to the frame, the error is measured. The 166 can send a selective decoder (10) to identify which video = ... selective decoder 16. It can then determine whether to decode the = code (for example, if the parameter is available) or decode the frame. , ^ S (for example, if the reference is not available). In the configuration, the encoder 110 of Figure 2A can be rearranged and/or combined: two or more components. In addition, the components of the encoder 11 can be hardware, Soft, medium intermediate, microcode or any combination thereof. In some configurations, one or more of the decoders 150 of Figure 2B can be rearranged and/or combined. The components of the horse 150 can be constructed by hardware, software, blade, middle, microcode or any combination thereof. Lu Hai reveals the internal communication ^ empty intermediary surface 2 d (for example) for the coffee ^ Forward Link Only [FLO] Air Interface 127986.doc • 15- 200840368

Specification for Terrestrial Mobile Multimedia Multicast*' is built on the MediaFL〇TM video encoding that delivers instant video services in the TM3 system. The FLO empty mediation specification was published as a technical standard in TIA_1〇99 in August 2006, and will be cited for all purposes. It is fully incorporated herein that the raster scan order inevitably imposes a horizontal property on the slice partition. Two separate slices for the slices are illustrated in Figures 3A and 3B. Figure 3A illustrates a first exemplary sample slice partitioning according to a frame 200 using a certain configuration of a slice structure according to the standard of ^1>]. With different cross hatching Representing different slice partitions. In this example, some of the macroblocks of the slices occupy two adjacent horizontal columns. In frame 200, a slice structure includes a macroblock located on a first horizontal column. Block 202 and macroblocks 2〇4 on a second horizontal column. In these structures, not all of the macroblocks in a slice group need to be directly adjacent. Figure 3B illustrates - according to A second exemplary sample slice is used in accordance with one of the configuration of the MpEG_2 standard slice structure in Fig. 3B. The slice structures are individually represented by a_q. The slice structures are arranged by column. Horizontally disordered arrangement. For example, the slice structure A extends over the entire first horizontal column R1. Similarly, the slice structure B extends the entire second horizontal column R2. However, in this example, the third level 歹(8) is listed. Shared by slice structures C and D. Horizontal column, R2 ,, μ, R 6, R 7, R 8, R 9, R i square, ρ 1 u, ο

The arrangement of the slice structures of R11 and R12 is 127986.doc -16- 200840368 is intended to be exemplary. However, each slice can occupy up to only one horizontal column and the right edge of the frame always marks the boundary of a slice.

Figure 4A illustrates a sample frame 25B having a partition based on type: flexible macroblock ordering (FM0) according to the H264/Avc standard. In FIG. 4A, frame 250 includes - the following table is identified by the cut group #2 and two (7) are identified by the slice group (4) and #1 for the interest area (r〇i) encoding before the stage partitions 256 and 258. . As can be seen in Figure 10, the foreground partition has an area of interest in the frame as a frame structure, which includes a subset of the columns and a macroblock adjacent to the subset of the rows. However, the macroblocks in partition 256 are adjacent. Thus, foreground partition 258 also includes a plurality of macroblocks arranged to include a subset of macroblocks that are adjacent in the vertical and horizontal directions. Partition 256 and partition 258 are labeled as rectangular areas to indicate the R 〇 I of the particular partition. Therefore, the left and lower right coordinates above the rectangles are necessary and are transmitted from the encoder device 11 to the decoder device 15A. 4B illustrates a sample frame 3〇〇 having a partition based on type i FMO according to the H.264/AVC standard. In Figure 4B, frame 3A includes an improved error resilience and a hidden checkerboard pattern. For illustrative purposes, the macroblocks that are depicted as white are associated with the slice group #〇. They are shown as black macroblocks for the slice group #丨. Therefore, there is an alternating pattern that can be influenced by the use of FMO. This is allowed because FMO no longer needs these slices to form adjacent macroblocks. Thus, the checkerboard pattern essentially provides a hashed slice. The FMO of the H.264/AVC standard includes seven different types labeled as Type 〇 - Type 6. However, for purposes of illustration, only the types 127986.doc -17-200840368 1 and type 2 are described herein to provide examples of slice structures. The FMO for error resiliency allows the macroblocks to be ordered in the form of a macroblock-free block surrounded by any other macroblock from the same slice group. Thus, in the event of an error (e. g., a slice is lost during transmission), the reconstruction of the lost block may depend on the information of the available blocks. Type 6 FM is the most important random type. Type 6 FM〇 allows full flexibility for the user. Other FM0 types, such as type 〇-type 5, are limited to having to follow a particular pattern. Although 嶋 is assigned to support different uses, it is currently regarded as an error resilience tool and is promoted as an error resilience tool.

In the compression standard prior to the H.264 video compression standard, the coding type of the VAU must be uniform over the entire range of each video frame. This necessitates that the slices that make up the frame must be encoded using the same coding type 1 (inner), P (predictive), or (bi-predictive or bi-predictive). With the introduction of the 264 standard, this limitation was eliminated. The H.264 standard allows for different coding types to be used in a VAU. Such a slice can be roughly of a different (encoding) type, resulting in a non-uniformly encoded vau. In addition, H264 can also generate a VAU by using a coding type that is consistent with an integer video frame, such as a Type 1 VAU, a P type VAU, or (D). The current configuration provides an encoding engine 5〇〇 (Fig. 7) that combines the possibility of using different slice (coding) types in a VAU I to utilize the 〇264 fm〇 rule 2 to achieve panning or volume in the camera. Moving (common and therefore most important) ^ In the case of a local motion and in the case where the scene consists of segments that are semantically different, an improved coding efficiency is achieved. Figure 1A illustrates a composite (i.e., multiple regions) scene VAU 900 having 127986.doc • 18-200840368 having a plurality of semantically different segments 9〇2, 904, 906, 9〇8, and 91〇, Example: Business News. In the upper left part of the scene identified by the segment number 9〇2, there is live video (new_advertising clip). In the upper right, there are financial indicators for text and graphics rendered in large font sizes, as indicated by fragment numbers 904 and 906. At the bottom of the multi-region scene VAU 900, there is a market display symbol and a quote from the right-to-left color, which is indicated by the segment number 9〇8, and a text that is rendered by the segment number 910 and rendered as text. Courier. In the multi-region scene synthesis, different scene segments will undergo unsynchronized changes such as scene change, cross fade, fade in and fade out, zoom in and zoom out, etc. due to their semantics and content differences. For example, if When the content of the segment 902 suddenly changes due to a "cut" scene change, only the macroblock in the segment 902 is encoded as the intra-frame coded macroblock, but the remaining macroblocks are encoded as The inter-frame coding macroblocks utilize the continuous temporal correlation in segments 904, 906, 908, and 910, and the contents of the segments 94, 4, 9, 6, 908, and 910 are unchanged at the moment. Therefore, the content is adjustable. The frame dividing unit 5 10 is operated to detect one of the NVC scene change, cross fade, fade in or fade out, zoom in or out, and global motion diversity. FIG. 5A illustrates an I-coded VAU (frame # 0) 350. Figure 5B illustrates a two Type B VAU (frames #1 and #2, not shown) between the Type I and Type P VAUs and when the camera begins to acquire the video signal with the frame #〇 P-coded VAU in the case of panning to the left (frame # 3) 370. The content of the video frame is substantially the same as that illustrated in FIG. 5A and FIG. 5B, except that in FIG. 5B, the scene is panned to the left as indicated by the virtual vertical line on the left side of the VAU 370, the virtual 127986.doc -19- 200840368 The vertical line marks the boundary between the new scene and the old scene details. The details of the new scene are not visible in the frame #〇, and the details of the old scene are visible in the frame #〇 and are therefore available for prediction. For the purpose of illustration of the sample, it can be assumed that a camera pointing to a particular scene with significant detail is undergoing an approximate meta-panning to the left.

Returning to FIG. 5 again, FIG. 5A and FIG. 5Β illustrate the initial I-frame VAU 350 with the structure]^61^61>... captured by the frame and frame 3, respectively. ρ frame VAU 37〇. In the frame, the inter-frame coding (ie, time-predicted) macroblock 372 (with the identified boundary 374) and its corresponding motion vector (the small header 377 in the identified square region) is displayed. . Macroblock 376 identifies a macroblock that is immediately adjacent to macroblock 372 in a horizontal plane. The boundary between horizontally adjacent macroblocks 372 and 376 is indicated by boundary 374. The remaining macroblocks of the ρ frame (the boundary is not recognized) are intra-frame coded, and most of the macroblocks are along the left boundary of the frame. Due to the nature of the motion, new details enter here. The scene. The distribution of the macroblock type and the motion vector field structure are very typical for the camera panning (to the left). It is dependent on the camera and the P-type VAU 37. The time between the reference frame and the reference frame on the left side of the main frame. The vertical bar of the block encoded in the frame can span one or more macro blocks. For the purposes of illustration, in Figure 5, the area of the frame on the left side of the frame or the area of the block represents the more complex case of the camera's panning motion in a scene. . Other situations or global translations of the above observations and scrolling can be done in a simple manner. 127986.doc -20- 200840368 Table 1. The encoding of the mb-type of MB in a frame of different slice types is not shown. The coded representation of the mb_type value mb__ value of the __4x4 macroblock in the slice-type frame, ie ue(v) codeword codeword length 2 or 7, (I) ~〇- ~ 1 1 0 Or 5, (P) 5 00110 5 1 or 6, (B) ~23 '1 ~ 000011000 9 In all video compression standards, the encoding of each MB (except for skipped MB) is sent at the front of the bit stream Type (mode) so that the decoder's decomposition and encryption decoding process can predict the correct syntax of each MB of data and correctly interpret each bit stream. In a p-type coded slice/VAU, the inter-frame coding (i.e., time-predicted) MB defines a preferred compression mode and its occurrence frequency is significantly greater than the frequency of occurrence of intra-frame coding Μβ in the P-type code slice/VAU. This will get the following observation report. C〇ntext Adaptive Variable Length Coding assumes that the content of the 264·264 adaptive variable length coding (CAVLC) is used to represent the mbβ type syntax element “mb type”, which can be summarized in the frame of different slice types as the table worker summarizes. A binary representation of the type of MB encoding MB. As can be seen, the use and signaling of a 4 χ 4 coded mb in the unintended frame in the P and B slices causes an additional 4 bits and 8 bits of overhead, respectively. This situation is similar to the external-16x16 encoded MB variable, although no relevant details are provided in this article. Therefore, the rest are equal, and it is exemplified that the MB encoded in the frame in the Z slice is most effective. The time prediction frame (i.e., the p-type and B-type coded VAUs) is used to provide the most important effect on coding efficiency, and its size should be small. Since the intra-frame code 127986.doc -21- 200840368 code is the least efficient coding type among the three types, there is an increase in the number of intra-frame coding in the p-type or B-type VAU. However, 'when this happens to the f' (for example) due to? Or the complex motion dynamics of the Type 8 or the scene where the new object enters the PSVAU*, the task of the encoder is to perform the encoding of the inner m B in the most efficient way possible. 6A illustrates a single frame 4〇〇 having a single vertical stripe structure of intra-frame coded macroblock groups 41〇 and a plurality of parallel horizontally extending inter-frame coded slices 415, wherein the message The vertical strip of the in-frame encoded macroblock begins on the left border of the frame 400. In this example, the segments 1-5 are parallel and each occupy the same number of columns. The configuration of the intra-frame coded macroblock group 410 that partially overlaps and thus is located within the inter-frame coded slice i_5 is limited to one of the VAUs indicated by the change of the macroblock block 41. Inside, and the boundary above the frame 4〇〇 extends completely to the lower boundary. Figure 6B illustrates a single frame 42 of an intra-frame coded macroblock group 425 having a single vertical stripe structure. The frame 42 further includes a plurality of parallel horizontally extending inter-frame coded slices 43A, wherein the vertical strips 425 of the encoded macroblocks within the frame begin at the right edge of the frame 420. As indicated by the change hatching of macroblock 425, the vertical strips 425 of the intra-frame coded macroblocks partially overlap and are therefore located within the inter-frame coded slice u. Figure 6C illustrates a single frame 45 of an intra-frame encoded macroblock group 460 having a single horizontal stripe structure. The frame 450 further includes a plurality of parallel horizontally extending inter-frame coding slices 455 of the 127986.doc -22-200840368, wherein the intra-frame coding macroblock group 46 of the single horizontal strip structure begins The lower boundary of block 45 0 and the left side of frame 450 extend completely to the right.

U Figure 6D illustrates a single frame 470 having an intra-frame coded macroblock group 475 of a single horizontal strip structure and a plurality of parallel horizontally extending inter-frame coded slices 480, wherein the single horizontal strip The structured intra-frame coded macroblock group 475 begins at the upper boundary of the frame 470 and extends completely to the right of the left side of the frame 470. When intra-frame coding is required in a substantial portion of a frame and the slice structure is not carefully cropped with respect to the geometry of the region, the slice boundaries in the region to be intra-frame coded will be reduced by the frame. Internal coding efficiency. This is essentially due to the unavailability of the slice boundaries of the neighboring pixels that were originally used to perform the intra prediction, and the consequent inability of some of the intra prediction modes due to the unavailability of the adjacent. Figures 6A-6D illustrate the case where the slice structure further divides the regions that are to be encoded and that are inconsistent with the in-frame according to certain configurations. ^ v μ , machine/knife, (divided into slice groups and slices) and non-uniform coded coding engine _. The engine includes the coding unit 52G of the content-adjustable information box division unit 5 - non-viewing L. The content adjustable frame division unit ^ ^ includes - lens boundary price detector 512, the motion field calculator 514 and the knife & 516 ° content adjustable information (four)] sub-unit into - step including the determination of the slice group and the assignment mode Group 518. Included - Lens boundary Detector 512 detects - or multiple frames - or multiple lenses 127986.doc -23- 200840368 Boundary. In the -state, the detection-lens boundary includes the system-scenario change. The scene change and the lens boundary are important, π due to the interruption of the dark discontinuous motion field and the change of the scene composition. The sports field calculator 514 calculates a motion field of one or more frames such as a frame, a frame, a B frame, and the like. In the -state, the measured global motion operations include operations such as panning or scrolling, zooming out, or zooming in on the PD type, and the motion deformation examples in the B and p types will cause the original time prediction The use of intra-frame coding in the access unit is required. Since the playing field is determined, the panning or scrolling, reduction or enlargement of the camera can be determined so that the VAU can be encoded non-uniformly. In an embodiment, information about a distinctly different motion field segment (such as the direction and intensity of motion vectors contained in the motion field segments) may be provided to the frame segmenter unit as a benefit. A hint for its segmentation task. Frame segmenter 516 is used to segment one or more frames. The frame segmenter 516 segments or divides the frames into one or more macroblock groups such as the set of macroblocks associated with the slice group #〇 and the slice group #1, As shown in any of Figures 9A-9D. A determination and assignment module 518 of the slice group to associate the identified one or more macroblock groups with one or more slice groups and one or more slices within each of the slice groups The purpose is to analyze the output of the frame segmenter. The slice group determination and assignment module 5丨8 analyzes the size and geometry of one or more of the identified macroblock groups, such as predictability within the frame or predictability within the frame, and the like. Attributes, assigning the one or more macroblock groups to one or more slice groups, and determining the size of one or more slices within the one or more slice groups 127986.doc -24- 200840368 ( Such as the number of columns occupied by any slice of any slice group). The slice group determination and assignment module 5i 8 determines slice groups, slices, and/or slice types for one or more frames. Coding unit 520 of the non-uniform video access unit (VAU) performs non-coherent encoding on the macroblocks associated with the determined type.

Referring again to Figure 2A, depending on the slice group and/or slice type, the slice will be encoded using inter-coded or intra-frame coding, respectively, by inter-coded encoder 11 or intra-frame code encoder ι2. Accordingly, the encoding device 1 encodes the slices based at least in part on the determined slice group and/or slice and/or slice type according to inter-coded or intra-frame coding techniques. The content-adapted frame division (divided into slice groups and slices) and the non-uniform VAU coding process reduce the coding efficiency caused by the mechanism. Therefore, it is not suitable for a hardware translational motion model (such as an object undergoing rotational motion) such as camera panning or scrolling, reduction or amplification, and B&P-type encoding in v A u encoded by P and B codes. Global motion operations such as complex motion variants in VAUs will necessitate the use of intraframe coding in originally time-predicted access units. This non-uniform VAU coding will use intra-frame coding in the time-predicted access unit with increased efficiency. To achieve this in the most efficient manner, an encoder can employ a process flow similar to that illustrated in Figure 8. As a result of this enhancement process, the slice segmentation structure of the sample case illustrated in Figs. 6A-6D should be modified to the structure illustrated in Figs. 9A-9D. It should be understood that the sample examples in Figures 6A-6D and Figures 9A-9D are provided for illustrative purposes only and may be based on FMO regulations in a fully live manner based on 127986.doc -25- 200840368 Predictive attributes are segmented/divided into groups and slices. Multiple regions (macroblock groups), slice groups Ο

In the various groups below, _ 士 ～ , , θ 不 实施 实施 实施 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 该 流程图 该 流程图 该 流程图 流程图 流程图 流程图 流程图 流程图 流程图 流程图The program sequence implements the block diagram or some of the blocks: a process flow chart for the encoding of the dependent and non-video access units in accordance with certain configurations. Process 600 begins at block 602 with a measure of shot-boundary/scene changes. The horizontal unit %* & is used to identify the basic unit with increased spatial time similarity, and the scene change interrupts the similarity of spatial time and marks the boundary of the unit S. Block 6〇4 is after block 6〇2, where the motion field is calculated. In the case, use bidirectional and single item calculations: identify predictive attributes of one or more areas within the frame or frame (such as inter-frame prediction or not), such as camera panning or scrolling, zooming out or Global motion operations such as zooming in, and identifying areas within the frame with significantly different motion characteristics (macroblock groups), such as static (no change), consistent motion, non-uniform motion region 0 should be noted, in a shot segment The motion field of all frames except the first frame of the video segment is calculated. Typically, a video sequence will include multiple shot segments, a collection of consecutive video frames separated by scene changes. An IBP... arrangement will be more accurately referred to as a "G〇p structure". Although the situation is a matter of aligning the I-frame with the scene change, it is not necessary, and there is a uniform insertion interval and does not necessarily change with the scene. Other reasons for aligning the frame (for example, enabling random access with an upper limit delay). 127986.doc -26 - 200840368 For example, the frame 350 of Figure 5 is an I frame and will not be subject to The field calculation. However, for frame 370 in Figure 5B, it is a p-type frame and the field calculation will be performed. The first frame or I frame will all be intra-frame coded. 块 Block 606 is at block 604. After that, the frame is segmented here. The segmentation of the frame is basically dependent on the temporal predictability and the motion field attribute. Block 6〇8 after block 6〇6 'here, the determination of the slice group and Assignment. Determine the assignment of slice group, slice and slice (code) type for each frame. At block 〇8, the absolute address of the first macro block in each slice (firslmb-injlice) can be identified. And/or anti-scan macro block information within each slice. The specific relationship of Figure 9A shows a frame 7. Here, the identification slice 6 has an I-type code. The slice boundary is determined, for example, the number of vertical macro block rows included in the slice 6. In one aspect The slice 6 is associated with the vertical strip of the macroblock and requires intra-frame coding as the camera pans to the left. Also, slices i, 2, 3, 4, and 5 are determined here. In the middle, the slice i-5 is a P-type slice. Therefore, in summary, the cut type should be type I, and the remaining slices are all type or all type b. The 'cloth,' flat code engine can also contain Such as for error resilience

Li is restricted. After the block is called, here, based on the identified slice coding type, the slice is subjected to intra-frame coding and inter- and flat codes. Block 610 ends the process 6〇〇. At block 612, the output of the process 600 is sent to a file of the memory 114 to be written, 1 , 6 , and/or transmitted as a bit stream to the receipt, for use by The network 14 is delivered to the decoder for 15 农. The output of process 600, which is based on the specific criteria being used, will contain information about the macroblocks associated with slice groups and slices. Figure 11 illustrates a flow diagram of a process for non-uniform VAU decoding in the presence of flexible macroblock ordering (FMO). The FM0 specification in the H.264 standard enables full flexibility to partition a macroblock set of a video frame (without implying a continuous raster scan order) or to group one or more slice groups and each slice group. One or more slices. Determining, by the encoder, dividing the macroblock set into one or more slice groups and one or more slices within each slice group, and forming a macro block and a slice group and a slice Contact is provided to the decoder. For example, in the H.264 standard, the association is signaled in the picture parameter set (PPS) by slice group mapping (SGM). The decoding operation will use the information provided by the slice group mapping (SGM) and the header of each slice to send the absolute address of the first macro block in the slice "first_mb_in_sHce" The semantic element reverse scans the information of the macroblocks in each slice from its outer raster scan order to its correct spatial position. Therefore, decoding and pixel are performed according to the slice encoding type also sent in the slice header. Reconstruction process. The decoder device 150 will implement non-uniform VAU decoding using SGM, which is generated by the encoder and written into the bit stream when specified using FMO. In the various configurations below, implemented in the illustrated order The flowchart blocks may implement the flowchart blocks, or portions thereof, simultaneously, in parallel, or in a different order. Process 1000 begins with block 1002, where decoder device 15 receives pps and determines the encoder SGM generated by device 110. According to block 〇〇4, solution 127986.doc -28 - 200840368 coder device 150 also receives semantic elements of each of the non-uniformly encoded VAUs in the slice header. 1006 at block 1004 After that, the absolute address (first-ml3 slice) of the first macroblock in the parent slice is determined here. The block loos is after block 1006, where it is scanned from its outer raster to its correct space. Position to perform inverse scanning of the operation of the macro block location information in each slice. Block 1〇1〇 after block 1〇〇8, where the decoding type is decoded according to the slice coding type sent in the slice header Uniformly encoded VAUs and reconstructed pixels. Figures 9A-9D illustrate example examples of non-uniformly encoded VAUs obtained from content compliant video frame cuts (both for slice geometry and slice encoding types) in accordance with certain configurations. In Fig. 9A, the non-coincidentally encoded VAU 700 includes a vertical slice #6 as a type I vertical block 7 5 located on the left side of the VAU and designated as intraframe coding. 715 begins at the left edge or boundary of the VAU 700 and extends one or more macroblock rows therefrom to define the right edge or boundary of the slice. The vertical strip 715 extends from the top to the bottom of the video frame. The remaining slices at 71〇 are 1-5 Inter-coding a p-type slice. Slices 1-5 are parallel horizontal structures and are grouped into slice group #0. The left boundary of slice group #〇 begins at the rightmost edge or boundary of slice #6 and extends to The right edge of the non-uniformly encoded VAU 700. Therefore, the absolute address of the first macroblock in each of the slices ^ will be set accordingly. In Figure 9B, the non-coincidentally encoded VAU 73A includes one The vertical strip 74 is located as the vertical strip of the macroblock located on the right side of the VAU and is designated as the vertical slice #6 of the type I to be intra-frame coded. The vertical strip 74 is away from the right of the VAu 73〇 127986.doc -29- 200840368 The edges or boundaries extend from - or a plurality of macroblock rows, extending through the edge or boundary of the MO - or a plurality of macroblock rows. The top of the vertical strip 740', VAU 73G extends to the bottom. The remaining slices designated at 735 15 are inter-coded as -p slices. Slices 5 are parallel horizontal structures and are grouped into slice groups. In this example, the left side of the slice group = starting at the left edge or boundary of the VAU 73G and extending to the left edge of the vertical strip. Therefore, the absolute address of the first macroblock in each of the slices κ will be set accordingly.

U in Fig. 9C 'VAU 75G includes - as a horizontal slice 715 on the top side of VAU 75() and designated as a single horizontal slice of (3) to be intra-frame coded. The remaining slice 2_? specified at 760 is inter-coded as a p-type slice. In the fortune, the right and left borders of the horizontal strip 755 coincide with the right and left borders or edges of the VAU 750. The top edge of the horizontal strip 755 coincides with the top edge of the VAU 750. However, the bottom edge of the horizontal strip 755 extends downwardly from the top VAU edge by one or more columns. Slices 2_7 are parallel and horizontally constructed in slice group #1. In this example, the right and left borders of each slice in slice group #1 also coincide with the right and left borders or edges of the VAU 75〇. In this arrangement, the number of columns in slices 1-7 and their size may not be four. The post-slicing slice of slice group #1, in this example, slice #7 has a bottom boundary that coincides with the bottom boundary of VAU 750. In Fig. 9D, the VAU 800 includes a single horizontal slice #8 which is a horizontal strip 804 located at the bottom of the VAU 8〇〇 and designated as type I to be intraframe coded. The bottom edge of the horizontal strip 804 coincides with the bottom edge of the VAU 8〇〇. The horizontal strip 804 extends upwardly from the bottom edge of the top VAU 800 to one or more macro zones 127986.doc -30 - 200840368 block columns. The remaining slices 1-7 designated at 802 are inter-coded as a p-type slice. Slices 1_7 are constructed in parallel and horizontally in slice group #(). In the cloth: the number of columns in the slice W and its size may vary. The first slice of the blade group #〇, in this example, slice #1 has a top boundary that is equal to the top boundary of VAu 75〇. The right and left borders of all slices coincide with the right and left borders of 750. Those skilled in the art will appreciate that information and signals can be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips that may be referred to throughout the above description may be by voltage, current, electromagnetic wave, magnetic field or particle, light field or particle, or any combination thereof. To represent. Those skilled in the art should further appreciate that the various illustrative logic blocks, modules, and algorithm steps set forth in connection with the examples disclosed herein can be constructed as electronic hardware, firmware, computer software, intermediates, microcode, or Its combination. To clearly illustrate the interchangeability of hardware and software, the above illustrative components, blocks, modules, circuits, and steps are summarized in terms of their functionality. Such functionality is implemented as a hardware or software depending on the particular application and design constraints imposed on the overall system. Skilled persons may construct the functionality in different ways for each particular application, but such implementations are not intended to be a departure from the disclosed method. Various illustrative logic blocks, components, modules, and circuits set forth in connection with the examples disclosed herein may use general purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays ( FPGA) or other programmable logic device, discrete gate or transistor logic, 127986.doc -31·200840368 «The hardware component, or any group 5 designed to perform the functions described herein, is constructed or executed. The general purpose processor may be a microprocessor, and the processor may be any conventional processor, controller, microcontroller (4) or state machine. A processor can also be a combination of horse transport devices, examples

:' A combination of a DSP and a microprocessor, a group of multiple microprocessors, one or more microprocessors, one such configuration. Combination of a core, or any other Ο c The steps of the method or algorithm described in connection with the examples disclosed herein can be directly embodied in the hardware, executed by the imaginary, 4 i / ^ self-processing w - or multiple In a software module or a combination of both. A software module can reside in r A memory, flash memory, ROM memory, EPR memory, memory, temporary memory, hard disk, removable disk, cd_r〇 m, or this 2^ knows the way - in other forms of storage media. An exemplary storage medium is coupled to the processor </RTI> to enable the processor to read information from, and write information to, the storage medium. Alternatively, the storage medium may be part of a processing benefit. The processor and the storage medium can reside in an application specific integrated circuit (ASIC). The ASICT resides in a wireless modem. As another" =. The processor and storage medium may reside as discrete components in wireless data. The above description of the disclosed examples is intended to enable any person skilled in the art to make or utilize the disclosed methods and apparatus. Various modifications of the examples will be readily apparent to those skilled in the art, and the principles defined herein may be applied to other examples and additional elements may be added. ^ [Simple description of the schema] 127986.doc -32- 200840368 In combination with the above figures, according to the above detailed description, the aspects and configuration of the disclosure will be more easily solved, and in all the drawings, the same reference characters Indicates the corresponding component. 1 illustrates a block diagram of an exemplary multimedia communication system in accordance with certain configurations. 2A illustrates a block diagram of an exemplary encoder acquisition that may be used in a graphics system. Figure 2B illustrates a block diagram of one exemplary decoder device that may be used in the system of Figure i. Figure 3A illustrates a first example frame with a sample slice partitioning according to some configurations using a slice structure according to the MPEG] standard. Figure 3B illustrates a second example frame with a sample slice partitioning according to certain configurations using a slice structure according to the MPeg-2 standard. Figure 4A illustrates a sample frame having a partition based on Type 2 Flexible Macro Block Sorting (FMO) according to the H.264/AVC standard. 4B illustrates a sample frame having a partition based on type i FMO according to the H.264/AVC standard. Figure 5A illustrates an I-coded vau (frame #0). Figure 5B illustrates a P-type encoded VAU (frame #3) of a Type B VAU between frame #〇 and frame #3 when a camera pans to the left. 6A illustrates a single frame having a single vertical stripe structure of intra-coded macroblock groups and a plurality of parallel horizontally extending frames 127986.doc -33 - 200840368 coded slices, wherein the message The vertical strip of the in-frame encoded macroblock begins on the left border of the frame. 6B illustrates a single frame of a frame-coded macroblock group having a vertical stripe structure and a plurality of parallel horizontally extending inter-frame coded slices, wherein the frame encodes a macroblock. The vertical strip begins on the right border of the frame. 6C illustrates a frame of a frame-incorporated macroblock block having a single horizontal stripe structure and a plurality of parallel horizontally extending inter-frame coded slices, wherein the frame of the horizontal stripe structure The group of intracoded macroblocks begins at the bottom boundary of the frame. 6D illustrates a single frame of a group of intra-frame coded macroblocks having a single horizontal stripe structure and a plurality of parallel horizontally extending inter-frame coded slices, wherein the frame of the horizontal stripe structure The group of intracoded macroblocks begins on the top boundary of the frame. Figure 7 illustrates an encoding engine for implementing content tunable frame division (divided into slice groups and slices) and non-uniform VAU coding. Figure 8 illustrates a flow diagram of a process for implementing content dependent frame partitioning and encoding of non-uniform video access units in accordance with certain configurations. Figures 9A-9D illustrate an illustrative example of a non-uniform VAU obtained from a content-adapted video frame cut (both for slice geometry and slice coding type) in accordance with certain configurations. Figure 10 illustrates a multi-region composite scene VAU having a plurality of segments that are semantically different. Figure 11 illustrates a flow chart of a process for non-uniform VAU decoding. 127986.doc -34- 200840368 Simplifies the images in these figures for illustrative purposes, and such images are not proportional. To aid understanding, the same reference numbers are used wherever possible to specify the same components that are common to the illustrations, except that suffixes can be added as needed to distinguish the components. The exemplary configurations of the present invention are illustrated by the accompanying drawings, and thus should not be considered as limiting The invention contemplates that a configuration 2 or step can be advantageously incorporated into other configurations without further recitation. ,test

[Major component symbol description] Multimedia communication system external source encoder device processor memory transceiver frame intra-coded encoder intra-frame code encoder component reference data generator component transmitter component network decoder device processor memory Transceiver 100 102 110 112 114 116 118 120 122 124 140 150 152 154 156 127986.doc • 35- 200840368 158 160 162 164 166 200 202 204 210 250 254 256 258 300 350 370 372 374 376 377 400 410 415 Receiver element Selective Decoder Component Reference Data Determiner Component Channel Switching Detector Component Error Detector Component Frame Macroblock Block Macro Frame Frame Sample Frame Background Front Hall Partition Front Partition Sample Frame VAU VAU Macro Area Block boundary macroblock small arrow frame vertical strip structure frame coding macroblock block inter-frame coding slice 127986.doc -36- 200840368 420 425 430 450 455 460 470 475 ί- 480 500 510 512 514 516 518 520 700 715 730 740

Frame single vertical stripe structure intra-frame coding macroblock block group inter-frame coding slice frame inter-frame coding slice single horizontal stripe structure intra-frame coding macroblock block group frame single level Strip structure intra-frame coding macroblock block inter-frame coding slice coding engine content adjustable frame division unit lens boundary detector motion field calculator frame segmentation slice group determination and assignment module Non-uniform video access unit (VAU) coding unit frame vertical strip VAU macro block vertical strip VAU 127986.doc -37- 750 200840368

755 800 804 900 902 904 906 908 910

Horizontal strip VAU horizontal strip

VAU fragment fragment fragment fragment fragment 127986.doc -38-

Claims (1)

200840368 X. Patent application scope · · · · A device that includes: processing theft's operation to perform division into a tangible adjustable frame division, and using one or: a single VAU within a slice to perform non-action, p - The slice, the flat horse type is encoded in a consistent view access unit (VAU); and a memory coupled to the processor. 2. In the case of a request for the device, the processor operates in the execution of the frame division, the store, the valley, the adjustment boundary, the previous, or the plurality of frames - or multiple shots n or The motion field of the plurality of frames, the segmentation of the one or more degrees, and the exact slice or the slice type for the one or more frames. a slice group, the arrangement of the I items 2, wherein the processor is performing the non-VAU code = less: the code is encoded based on the determined slice group, the slice or the slice type One or more frames. 4 St device 3, wherein the processor encodes a non-fox code = a portion based on the determined slice groups belonging to a slice group and an early slice group to encode a respective one Frame. 5·=device of claim 4, wherein the processor is performing the non-VAU encoding, encoding each frame into a first group having a single slice, and having a plurality of parallel horizontal extensions Two groups. 6·= Item 3, wherein the -the-slice group comprises-arranged in the early frame of the early frame to encode the macroblock strip, and the single frame contains the macroblock block The strip has three edges that coincide with the parallel horizontally extended boundary edges or portions of the three 127986.doc 200840368 one slice group blocks of the respective frame, and includes a plurality of inter-frame coding macros Sliced area. 7. The device of claim 6, wherein the content is adapted to include the detection of the sports field containing the global motion to which the associated macroblocks are assigned. In-frame coding 8. As in the device of claim 7, the camera pans to the right by 0. The global motion includes a device that shakes the camera to the left, the camera scrolls up, or the camera scrolls down to item 1, where the processor operates In response to a composite/or multiple changes in the scribing system, the block sorting uses all of the different w coding classes in the encoding. ~ 曰 The VAU - or multiple parts rather than the 10. multimedia system, including · · 内谷 可调 适 框 彳 - - - - - - - - - - - - - 兀 兀 兀 兀 兀 兀 兀Dividing a frame into a slice group and a slice; and a j-to-video access unit (trans) coding unit, which encodes the first part of the frame into a single intra-frame coding macroblock strip, The second portion of one of the frames is encoded into a plurality of parallel horizontally extending inter-frame coded slices. 11. The system of claim 10, wherein the content adjustable frame division unit comprises: 127986.doc 200840368 one or more lens groups and a predator for detecting one or more boundaries Or a plurality of frames of motion fields or a plurality of frames; and or a plurality of frames determining a slice-calculator for calculating the one-segment segmenter for segmenting the one-to-one determiner for use For this one and slice or slice type. The non-uniform vau coding unit includes a f code: V operative to encode the one or more frames based at least in part on the determined slice group or slice types. .:: The system of item 12 wherein the encoder operates to encode a respective frame based at least in part on the determined group of slices belonging to a p-type and-single-ship slice group. % is the system of claim 13, wherein the encoder operates to: - each one busy, flat code into a first slice having the single slice and one having a plurality of parallel horizontally extending interframe coding The second group of slices. The system of claim 14 wherein one of the first slice groups includes a plurality of intra-frame coding macroblocks disposed in the 2 solid strips, the single strip having three of the respective frames A boundary edge or a portion of the coincident edge, and a second slice group includes a plurality of parallel horizontally extending inter-frame coded slices. A system of months 1 wherein the calculator detects a sports field comprising global motion for assigning the macroblocks associated with the global motion to the intra-coded macroblocks. For example, the system of claim 15, wherein the global motion includes the camera shaking left 127986.doc 200840368, the camera panning to the right, the camera scrolling up, or the camera scrolling down 18, - for processing multimedia materials The method includes: dividing a frame into a slice and a slice by performing content-tunable frame division; and performing a non-uniform video access unit (VAU) and a flat code on the first part of the frame The coded macro in a single frame and one of the frames _ 邡 抽 — μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ μ Intercoded slices. 19. If request 1 &amp; s, +, , go, it further includes detecting a global motion detection. The method of claim 18, wherein the content adjustable frame division comprises: detecting one or more lens boundaries of the plurality of frames, calculating the one or more busy sports fields, segmentation The one or more frames, and determining the slice groups, the slices, or the slice types for the one or more frames. The method of claim 18, wherein the non-uniform VAU encoding comprises encoding the one or more frames based at least in part on the determined group of slices, the slices, or the two types. The method of claim 21, wherein the non-stationary coding comprises at least ortho (4) encoding a respective frame based on the determined 4 slice groups belonging to the -p type and -single (3) slice group. 23. The method of claim 22, wherein the non-VAU encoding comprises: - encoding each frame into a first slice group having the single strip 127986.doc 200840368 = ^ a single strip having Each of the three boundary edges or portions of the three frames coincides with three edges; and the code_ has a plurality of parallel horizontally extending inter-frame coded slices of the second group. The method of the month of the mussel 23, wherein the encoding package of the first slice group • #使四间-frames-horizontal or vertical boundary edges and the same, is the code macroblock A single strip is associated with a code-packet' ^ a number of parallel second horizontally stretched inter-frame coded slices. 25. The method of claim 24 wherein the content adjustable frame partition comprises a sports field of the squad, J 〇 ^ king motion for assigning the E set blocks associated with the global motion to the # News (4) coding macro block. 26. The method of claim 25, wherein the predicate of the global motion comprises a predicate - one of panning to the left, one panning to the right, one camera scrolling up, or one camera scrolling down . 27· ^ encoding device 'which includes an encoding engine that can operate U in response to a camera panning or - scrolling global motion detection to the social-single video access unit (VAU) Flexible slice ordering (FM〇) is applied for different slice coding types. 28. The encoding device of claim 27, further comprising: a component for a plurality of frames or a plurality of shot boundaries; for calculating the one or more frames including the global motion a component of the sports field; a means for segmenting the one or more frames; and for determining a slice group and a slice or slice for the one or more frames 127986.doc 200840368 m dust ι component. Based on: the further step comprising: a computer program product for at least partially encoding the - or more two: box:: and the slice or the slice type 3°. readable medium, the computer readable Electricity: an instruction to process multimedia data, wherein the instructions are: macro: macro block sorting (fm〇) to divide the content of a frame into a slice group and slice, and: and: The slice encoding type performs the non-31. The computer program product of claim 30, wherein the instructions for performing the content=frame division include causing the computer to perform the following job detection. Or one or more shot boundaries of the plurality of frames; calculating a motion field of the one or more frames; segmenting the one or more frames; and determining the slice groups for the one or more frames and These slices are the same type of slice. The computer program product of claim 31, wherein the instructions for performing the non-uniform VAU encoding comprise causing the computer to execute the following job instruction: μ based at least in part on the determined slice group and the The slice or the slice types are used to encode the one or more pivots. 127986.doc 200840368 33. The computer program product of claim 31, wherein the instructions for calculating the motion field include instructions for causing the computer to detect a global motion detection. 34. The computer program product of claim 30, wherein the VAu is a multi-regional scene having a semantically different segment, and the instruction to perform the content tunable frame partitioning comprises causing the computer to cause the vau An instruction that is divided into segments that are semantically different. 35. The computer program product of claim 34, wherein the instructions for dividing further comprise causing the computer to determine a change in the cut scene, to fade, fade in or fade out, zoom in or out, and diversify in global motion. Any one or more instructions. 36. The computer program product of claim 31, wherein the instructions for performing the non-uniform VAU encoding comprise causing the computer to belong to a P-type or B-type slice group and a single 1 based at least in part on the determination. The slice groups of the type slice group encode an instruction for a respective frame. 37. The computer program product of claim 36, wherein the instructions for performing the non-uniform VAU encoding comprise instructions for causing the computer to perform the operation of: encoding a respective frame into a single intraframe coded macro a first group of block strips, the coded macro block strips of the single frame having three edges coincident with three boundary edges or portions of the respective frame; and the code one has a complex number A second group of coded slices between parallel horizontally extending frames. 38. The computer program product of claim 31, wherein the instructions for calculating include causing the computer to detect global motion for use in linking the macros associated with the global 127986.doc 200840368 The block is assigned as an instruction to encode a macroblock within the frame. 39. The computer program product of claim 38, wherein the instructions for detecting the global motion comprise causing the computer to detect that a camera pans to the left, a camera pans to the right, and a camera scrolls up And a camera scrolling down one of the instructions. 40. A device for processing multimedia material, comprising: a component for dividing a frame into a slice group and a slice by performing content-tunable frame division; and for performing a non-uniform video access unit on the divided frame using - or a plurality of slice coding types (VAU) The component of the code. For example, the device of claim 40, wherein the means for performing content tunable frame segmentation comprises: means for making - or a plurality of frames - or a plurality of lens boundaries; for calculating the one or more a component of a motion field of a frame; a component for segmenting the one or more frames; and means for determining the slice groups, the slices, or the components of the frame for the frame or frames. 1颂42.=:1°: a device, wherein the means for performing the non-coding comprises encoding the _ or more based at least in part on the determined slice or the like of the slice group D or the slice types The apparatus of claim 40, wherein the member 3 instructed to perform the non-VAU encoding is based, at least in part, on a determined type or a type of switching group and A single type I slice group slice - a frame. (4) Slice group to encode - each 127986.doc 200840368 44. The device of claim 4, wherein the means for performing non-uniform 编码 编码 encoding comprises: encoding a respective frame into one a first group of coded macroblock strips in a single frame and a second group of inter-frame coded slices having a plurality of parallel horizontal extensions, the blocklet macroblock block strips in the single frame having Three edges that coincide with the three boundary edges or portions of the respective frame. 45. The device of claim 4, wherein the device has a multi-regional scenario of two segments that are semantically different, and the means for performing the content adjustable frame division comprises dividing the VAU into the semantics Different segments. 46. A decoding device that includes a decoding engine operable to combine a different slice encoding class within a single video access unit (then) to apply a live macroblock ordering (FMQ) The single non-coded VAU is decoded. 47. The decoding device of claim 46, wherein the decoding engine receives the set of picture parameters through the touch and receives an absolute address of each of the different coding classes. 48. The decoding device of claim 46, wherein the decoding engine includes a selective solution: the second is used for decoding - having - a single intraframe coding macroblock U f # slice group ' $ a single message The in-frame coded macroblock stripe has three edges that coincide with the three boundary edges or portions of the single one; and decodes - a second slice group having a plurality of parallel horizontally extending inter-frame coded slices. 49. A computer-readable media medium comprising a computer program product for processing multimedia, the computer readable data instructions, wherein the instructions are 127986.doc 200840368 enabling a computer: combining a single video access unit (VAU) Flexible macroblock ordering (FMO) is applied to decode the single non-coherently encoded VAU. 5. The computer program product of claim 49, wherein the instructions for decoding comprise causing the computer to receive a set of picture parameters by the FM frame to receive a first macro region of each of the different coding types. Block - Absolute: Instruction. 5: The computer program product of claim 49, wherein the instruction packet for decoding is used to cause the computer to perform the following operations: ♦ the stone horse has one - a single frame intra-coded macro block strip a _th slice group, the coded macroblock strip in the single frame has three edges U·u 琼 edges that coincide with three boundary edges or portions of the single VAU, and the decoding one has a plurality of parallels A second slice group of sliced slices between horizontally extending frames. 52. An encoding device comprising an encoding engine operable to respond to a single viewing 5-hole access unit (vau) in response to a $_day, dry or multiple change in a composite scene The different slices within the IJ slice encoding type apply the flexible giant block sorting (FMO), which affects one or more parts of the VAU rather than all parts of the VAU. 53. The or changes include shearing, zooming in or out, and globally, as in the encoding device of claim 52, wherein the scene changes, fades, fades in, or fades out any one or more of the sports diversity. 127986.doc