公开(公告)号：US5754235A

公开(公告)日：1998-05-19

申请号：US410490

申请日：1995-03-24

申请人： Takashi Urano ,  Koichi Tsuchikane ,  Satoko Kobayashi

发明人： Takashi Urano ,  Koichi Tsuchikane ,  Satoko Kobayashi

IPC分类号： G06T9/00 ,  H04N7/26 ,  H04N11/02

CPC分类号： H04N19/40

摘要： An input bitstream (a first bitstream) is divided into a main code having a changeable length and the remaining data. The main code is subjected to a bit-rate conversion and combined with the remainig data to output a second bitstream. By this arrangement, the configuration ofthe bit-rate conversion circuit can be simplified and delay in the conversion process can be reduced. Incidentally, the main code corresponds to the DCT-coefficient data(e) and the quantizer scale(d), while the remaining data corresponds to the header data(a), the handling type data(b) and the motion vector data(c) respectively. Furthermore, by using the fact that if the quantizer scale is enlarged and the compression rate is raised, the degradation of a reproduced picture with a large DCT-coefficient's value of higher frequency terms is not conspicuous but the degradation of a reproduced picture with a small DCT-coefficient's value of higher frequency terms is conspicuous, and by changing the compression rate based on the DCT-coefficient's value of higher frequency terms, a bitstream with a low bit-rate but a small degradation in the reproduced picture quality can be obtained.

摘要翻译： 输入比特流（第一比特流）被划分为具有可变长度的主码和剩余数据。 对主码进行比特率转换并与剩余数据组合以输出第二比特流。 通过这种布置，可以简化比特率转换电路的配置，并且可以减少转换处理中的延迟。 顺便提及，主码对应于DCT系数数据（e）和量化器刻度（d），而剩余数据对应于标题数据（a），处理类型数据（b）和运动矢量数据（c ） 分别。 此外，通过使用如果量化器尺度扩大并且压缩率升高的事实，则具有较高频率项的较大DCT系数值的再现图像的劣化不显着，但是具有小的再现图像的劣化 较高频率项的DCT系数的值是显着的，并且通过基于较高频率项的DCT系数的值改变压缩率，可以获得具有低比特率但是再现的图像质量的小的劣化的比特流。

公开(公告)号：US06011598A

公开(公告)日：2000-01-04

申请号：US825528

申请日：1997-03-28

申请人： Yasunori Tsuji ,  Takashi Urano ,  Koichi Tsuchikane ,  Yasuhachi Hamamoto

发明人： Yasunori Tsuji ,  Takashi Urano ,  Koichi Tsuchikane ,  Yasuhachi Hamamoto

IPC分类号： H04N7/26 ,  H04N7/50 ,  H04N21/234 ,  H04N21/44

CPC分类号： H04N21/23406 ,  H04N19/102 ,  H04N19/146 ,  H04N19/42 ,  H04N19/44 ,  H04N19/61 ,  H04N19/70 ,  H04N21/44004

摘要： A decoding start controller is capable of appropriately decoding pictures up to the final picture, even if the amount of the data stored in a bit-stream buffer is reduced, as a result of partial storage of the data in an internal local buffer. Further, the decoding start controller is capable of decoding extremely short bit-stream data. If completion information, which is sent from an input side of the bit-stream data and represents completion of input of the bit-stream data, is detected, a decoding start instruction signal is output to a variable-length decoder regardless of a value of VBV occupancy. In short, when the amount of the data stored in the bit-stream buffer connected to the decoder having the decoding start controller, is equal or greater than the value of VBV occupancy, the decoding start instruction signal will usually be sent to the variable-length decoder: namely, so-called decoding start control operation based on value of VBV occupancy. If the completion information is detected, the decoding start control operation based on the value of VBV occupancy is interrupted, and the decoding start instruction signal is output.

摘要翻译： 即使存储在位流缓冲器中的数据量减少，解码开始控制器能够适当地解码直到最终图像的图像，这是由于数据在内部本地缓冲器中的部分存储的结果。 此外，解码开始控制器能够解码极短的比特流数据。 如果检测到从比特流数据的输入侧发送并表示比特流数据的输入的完成的完成信息，则将解码开始指令信号输出到可变长度解码器，而不管 VBV占用。 简而言之，当存储在连接到具有解码开始控制器的解码器的比特流缓冲器中的数据量等于或大于VBV占有率时，解码开始指令信号通常被发送到可变 - 长度解码器：即所谓的基于VBV占用率的解码启动控制操作。 如果检测到完成信息，则基于VBV占用的值的解码开始控制操作被中断，并且输出解码开始指令信号。