摘要:
An image decoding device which increases decoding efficiency and can be easily implemented includes: a division unit dividing coded image data into first and second coded image data; a frame storage unit; first and second decoding units decoding, in parallel, the first and second coded image data; and an information storage unit. The first decoding unit decodes the first coded image data using second decoding result information stored in the information storage unit and stores, as first decoding result information, a part of information generated by the decoding into the information storage unit. The second decoding unit decodes the second coded image data using the first decoding result information stored in the information storage unit and stores, as the second decoding result information, a part of information generated by the decoding into the information storage unit.
摘要:
To provide a simply-structured image decoding apparatus which appropriately executes parallel decoding processing. The image decoding apparatus includes: a stream segmentation unit (130) which generates plural segment streams by segmenting each of the pictures included in a bit stream into plural MB lines, and assigning each of the plural MB lines to a corresponding one of the plural segment streams to be generated; and plural decoding engines (120) which decode the respective plural segment streams in parallel. In the case where at least two consecutive MB lines in the bit stream have a dependency indicated by a predetermined code word, the stream segmentation unit (130) sets the code word to the segment streams so as to update the dependencies to new dependencies depending on the contexts of MB lines in the respective segment streams.
摘要:
An image decoding device (C100) for increasing decoding efficiency and reducing the number of memory accesses includes: a division unit (C101) dividing a picture into first and second coded image data; and first and second decoding units (C103 and C104) decoding, in parallel, the first and second coded image data and storing decoding results into a frame storage unit (C102). The first and second decoding units (C103 and C104) decode the first and second coded image data using second and first decoding result information, respectively, and store the resulting first and second decoding result information into an information storage unit (C105). When decoding a target macroblock, each of the first and second decoding units performs image processing on: the corresponding second or first decoding result information indicating part of the decoded macroblock included in the macroblock line adjacent to the macroblock line including the target macroblock; and the target macroblock.
摘要:
An image decoding device for increasing decoding efficiency and reducing the number of memory accesses includes a division unit dividing a picture into first and second coded image data; and first and second decoding units decoding, in parallel, the first and second coded image data and storing decoding results into a frame storage unit. The first and second decoding units decode the first and second coded image data using second and first decoding result information, respectively, and store the resulting first and second decoding result information into an information storage unit. When decoding a target macroblock, each of the first and second decoding units performs image processing on the corresponding second or first decoding result information indicating part of the decoded macroblock included in the macroblock line adjacent to the macroblock line including the target macroblock, and the target macroblock.
摘要:
An image decoding device which increases decoding efficiency and can be easily implemented includes: a division unit dividing coded image data into first and second coded image data; a frame storage unit; first and second decoding units decoding, in parallel, the first and second coded image data; and an information storage unit. The first decoding unit decodes the first coded image data using second decoding result information stored in the information storage unit and stores, as first decoding result information, a part of information generated by the decoding into the information storage unit. The second decoding unit decodes the second coded image data using the first decoding result information stored in the information storage unit and stores, as the second decoding result information, a part of information generated by the decoding into the information storage unit.
摘要:
To provide a simply-structured image decoding apparatus which appropriately executes parallel decoding processing.The image decoding apparatus includes: a stream segmentation unit (130) which generates four segment streams by segmenting each of the pictures included in a bit stream into plural MB lines, and assigning each of the plural MB lines to a corresponding one of the four segment streams to be generated; and four decoding engines (120) which decode the respective four segment streams in parallel. When the stream segmentation unit (130) segments a slice in the picture into slice portions and assigns the respective slice portions to plural segment streams, the stream segmentation unit (130) reconstructs, as a new slice, the slice portion group including at least one slice portion to be assigned to a corresponding one of the segment streams so that the slice portion group is recognized as a slice by the associated one of the decoding engines (120).
摘要:
A buffer management apparatus that sequentially receives L (L>1) types of data and transmits the L types of data to an external device, including: a reception unit that receives data; M (M
摘要:
A resource request arbitration device is connected with each of a plurality of masters, and arbitrates transfer requests issued by the masters. The resource request arbitration device includes a plurality of counters each indicating a slack time of a transfer request issued by a master corresponding to the counter, and compares counter values stored in the counters by a tournament method, and specifies a master that has issued a transfer request having the highest priority. The resource request arbitration device grants access permission to the specified master to permit the specified master to use a slave.
摘要:
In layout design step of the semiconductor integrated circuit manufacturing method, when it is found that the wiring length between an external terminal and an IO block (external terminal I/F circuit) corresponding to the external terminal increases after a floorplan of a circuit including a functional block and the IO block is determined, placement of the IO block is determined such that the IO block is placed close to the external terminal to alleviate constraints on the wiring between the IO block and the external terminal, and timing adjustment circuits whose number is determined according to the wiring length of a bus (or a shared bus) connecting a data transfer circuit and the IO block is inserted into the bus.
摘要:
A resource request arbitration device is connected with each of a plurality of masters, and arbitrates transfer requests issued by the masters. The resource request arbitration device includes a plurality of counters each indicating a slack time of a transfer request issued by a master corresponding to the counter, and compares counter values stored in the counters by a tournament method, and specifies a master that has issued a transfer request having the highest priority. The resource request arbitration device grants access permission to the specified master to permit the specified master to use a slave.