- 您現(xiàn)在的位置：買賣IC網(wǎng) > PDF目錄363910 > W9961CF (WINBOND ELECTRONICS CORP) H.263/H.261 Video Codec PDF資料下載
參數(shù)資料
| 型號： | W9961CF |
| 廠商： | WINBOND ELECTRONICS CORP |
| 元件分類： | 消費家電 |
| 英文描述： | H.263/H.261 Video Codec |
| 中文描述： | SPECIALTY CONSUMER CIRCUIT, PQFP208 |
| 封裝： | 28 X 28 MM, 2.60 MM HEIGHT, PLASTIC, QFP-208 |
| 文件頁數(shù)： | 29/65頁 |
| 文件大小： | 344K |
| 代理商： | W9961CF |
第1頁第2頁第3頁第4頁第5頁第6頁第7頁第8頁第9頁第10頁第11頁第12頁第13頁第14頁第15頁第16頁第17頁第18頁第19頁第20頁第21頁第22頁第23頁第24頁第25頁第26頁第27頁第28頁當前第29頁第30頁第31頁第32頁第33頁第34頁第35頁第36頁第37頁第38頁第39頁第40頁第41頁第42頁第43頁第44頁第45頁第46頁第47頁第48頁第49頁第50頁第51頁第52頁第53頁第54頁第55頁第56頁第57頁第58頁第59頁第60頁第61頁第62頁第63頁第64頁第65頁
W9961CF
- 29 -
The quantized blocks are also inverse quantized (IQ) and transformed into the spatial domain by an
inverse DCT (IDCT). This operation yields a copy of the encoded picture as it will be seen by the
decoder. That copy is then stored into the video memory and will be used for future predictive coding.
Since the VLE operation is lossless, there is no need to include the VLE unit in the feedback path.
7.2.1.2 P-pictures INTER Coding
P-pictures macroblocks may be coded by INTRA or INTER coding mode. For each macroblock in the
current P-picture, the motion estimation is performed on the luminance (Y) macroblock. A full search
or fast search is made with integer pixel displacement in the Y component at first. The comparisons
are made between the incoming macroblock and the displaced macroblock in the previous
reconstructed picture. The encoder makes a decision on whether to use INTRA or INTER prediction in
the coding after the integer pixel motion estimation. If INTRA mode is chosen, no further operation is
necessary for the motion search. We will describe P-picture INTRA coding in the following section
7.2.1.3.
If INTER mode is chosen the motion search continues with half-pixel search around the integer pixel
motion vector, MV0, position. After the half-pixel search, the best match motion vector is coded using
a variable-length encoder (VLE), and stored into the video memory. Motion vector is included for all
INTER macroblocks and consists of horizontal and vertical components, both measured in half pixel
units.
P-pictures INTER coding does not code the picture macroblocks directly. Instead it codes the
prediction errors. For each INTER coding macroblock in the current picture, the best match
macroblock in the previous reconstructed picture is loaded into the MC and half-pixel motion
compensation is performed according to the motion vector. After half-pixel motion compensation, the
two macroblocks are subtracted to produce prediction errors (their difference) which will be DCT
transformed. DCT coefficients are quantized, zig-zag scanned, coded using a variable-length
encoder, and stored into the video memory. The quantized blocks are also inverse quantized and
transformed into the spatial domain by an inverse DCT. The IDCT results and the previous
reconstructed blocks are added and stored into the video memory as the current reconstructed picture
for future predictive coding.
7.2.1.3 P-pictures INTRA Coding
If INTRA mode is chosen for current P-pictures macroblock, no further half-pixel motion search is
performed and no motion vector is coded. Each INTRA macroblock is coded as that for I-pictures
INTRA coding.
7.2.2 Video Decoding
Video decoding operation is very similar to the feedback loop of the video coding. After optional error
correction, the compressed bit stream is processed by the variable length decoder (VLD). The
decoded data are parsed, inverse zig-zag scanned (IZZ), and then processed by an inverse quantizer
and an inverse DCT. Depending on the transmission mode (INTRA or INTER), macroblocks from the
previous reconstructed picture may also be added to the current data to form the reconstructed
picture.
For each INTRA-coded macroblock of I-pictures or P-pictures, no motion compensation is performed.
The IDCT results are stored into the video memory as reconstructed picture.
相關(guān)PDF資料 |
PDF描述 |
|---|---|
| W9967BCF | JPEG USB Camera Chip(運用JPEG壓縮技術(shù)和通用串行總線結(jié)構(gòu)技術(shù)的照相電路數(shù)字處理芯片的攝像照相機接口控制器) |
| W9967CF | JPEG USB Camera Chip(運用JPEG壓縮技術(shù)和通用串行總線結(jié)構(gòu)技術(shù)的照相電路數(shù)字處理芯片的攝像照相機接口控制器) |
| W99681CF | JPEG USB Dual Mode Camera Chip(運用JPEG壓縮技術(shù)、通用串行總線結(jié)構(gòu)技術(shù)、雙制式的照相電路數(shù)字處理芯片的攝像照相機接口控制器) |
| W9971CF | Integrated Video Graphics Accelerator(集成的視頻圖形加速器) |
| W9980 | Digital Voice Recorder(數(shù)字式錄音器) |
相關(guān)代理商/技術(shù)參數(shù) |
參數(shù)描述 |
|---|---|
| W9966CF | 制造商:未知廠家 制造商全稱:未知廠家 功能描述: |
| W9966TF | 制造商:未知廠家 制造商全稱:未知廠家 功能描述: |
| W99681CF | 制造商:WINBOND 制造商全稱:Winbond 功能描述:JPEG USB Dual Mode Camera Chip |
| W99685 | 制造商:未知廠家 制造商全稱:未知廠家 功能描述: |
| W99685BS | 制造商:WINBOND 制造商全稱:Winbond 功能描述:SYSTEM CAMERA DEVICE |
發(fā)布緊急采購，3分鐘左右您將得到回復(fù)。