1简介YUV是种颜色编码方法，是彩色电视为了兼容黑白电视而发展起来的。在现代彩色电视系统中，通常采用三管彩色摄影机或彩色CCD摄影机进行取像，然后把取得的彩色图像信号经分色、分别放大校正后得到RGB，再经过矩阵变换电路得到亮度信号Y和两个色差信号R－Y（即U）、B－Y（即V），然后对这三个信号分别进行编码。Y就是所谓的流明(luminance)，表示光的浓度且为非线性，使用伽马修正(gammacorrection)编码处理，而CB和CR则为蓝色和红色的浓度偏移量成份。 
2YUV采样格式 （1）YUV4:4:4 YUV三个信道的抽样率相同，因此在生成的图像里，每个象素的三个分量信息完整（每个分量通常8比特），经过8比特量化之后，未经压缩的每个像素占用3个字节。 下面的四个像素为:[Y0U0V0][Y1U1V1][Y2U2V2][Y3U3V3] 存放的码流为:Y0U0V0Y1U1V1Y2U2V2Y3U3V3 （2）YUV4:2:2 每个色差信道的抽样率是亮度信道的一半，所以水平方向的色度抽样率只是4:4:4的一半。对非压缩的8比特量化的图像来说，每个由两个水平方向相邻的像素组成的宏像素需要占用4字节内存。 下面的四个像素为：[Y0U0V0][Y1U1V1][Y2U2V2][Y3U3V3] 存放的码流为：Y0U0Y1V1Y2U2Y3V3 映射出像素点为：[Y0U0V1][Y1U0V1][Y2U2V3][Y3U2V3] （3）YUV4:1:1 4:1:1的色度抽样，是在水平方向上对色度进行4:1抽样。对于低端用户和消费类产品这仍然是可以接受的。对非压缩的8比特量化的视频来说，每个由4个水平方向相邻的像素组成的宏像素需要占用6字节内存。 下面的四个像素为:[Y0U0V0][Y1U1V1][Y2U2V2][Y3U3V3] 存放的码流为:Y0U0Y1Y2V2Y3 映射出像素点为：[Y0U0V2][Y1U0V2][Y2U0V2][Y3U0V2] （4）YUV4:2:0 4:2:0并不意味着只有Y，Cb而没有Cr分量。它指得是对每行扫描线来说，只有一种色度分量以2:1的抽样率存储。相邻的扫描行存储不同的色度分量，也就是说，如果一行是4:2:0的话，下一行就是4:0:2，再下一行是4:2:0...以此类推。对每个色度分量来说，水平方向和竖直方向的抽样率都是2:1，所以可以说色度的抽样率是4:1。 下面八个像素为：[Y0U0V0][Y1U1V1][Y2U2V2][Y3U3V3] [Y5U5V5][Y6U6V6][Y7U7V7][Y8U8V8] 存放的码流为：Y0U0Y1Y2U2Y3Y5V5Y6Y7V7Y8 映射出的像素点为：[Y0U0V5][Y1U0V5][Y2U2V7][Y3U2V7] [Y5U0V5][Y6U0V5][Y7U2V7][Y8U2V7] 3YUV存储格式YUV格式分为packed格式和planar格式，packed格式会将YUV数据按顺序放进数组里，planar则分将YUV的数据分为三部分放进数组里，即先放Y部分的数据，再放U，接着是V的数据。 3.1PackedYUVFormatsLabel | Bitsperpixel | Description | AYUV | 32 | CombinedYUVandalpha | CLJR | 8 | CirrusLogicformatwith4pixelspackedintoau_int32.AformofYUV4:1:1wihtlessthan8bitsperY,UandVsample. | cyuv | 16 | EssentiallyacopyofUYVYexceptthatthesenseoftheheightisreversed-theimageisupsidedownwithrespecttotheUYVYversion. | GREY | 8 | ApparentlyaduplicateofY800(andalso,presumably,"Y8") | IRAW | ? | InteluncompressedYUV.Ihavenoinformationonthisformat-canyouhelp? | IUYV | 16 | InterlacedversionofUYVY(lineorder0,2,4,....,1,3,5....)registeredbySilviuBrinzeiofLEADTechnologies. | IY41 | 12 | InterlacedversionofY41P(lineorder0,2,4,....,1,3,5....)registeredbySilviuBrinzeiofLEADTechnologies. | IYU1 | 12 | 12bitformatusedinmode2oftheIEEE1394DigitalCamera1.04spec.ThisisequivalenttoY411 | IYU2 | 24 | 24bitformatusedinmode0oftheIEEE1394DigitalCamera1.04spec | HDYC | 16 | YUV4:2:2(Ysampleateverypixel,UandVsampledateverysecondpixelhorizontallyoneachline).Amacropixelcontains2pixelsin1u_int32.ThisisasuplicateofUYVYexceptthatthecolorcomponentsusetheBT709colorspace(asusedinHDvideo). | UYNV | 16 | AdirectcopyofUYVYregisteredbyNVidiatoworkaroundproblemsinsomeoldcodecswhichdidnotlikehardwarewhichofferedmorethan2UYVYsurfaces. | UYVP | 24? | YCbCr4:2:2extendedprecision10-bitspercomponentinU0Y0V0Y1order.RegisteredbyRichEhlersofEvans&Sutherland.(Awaitingconfirmationofcomponentpackingstructure) | UYVY | 16 | YUV4:2:2(Ysampleateverypixel,UandVsampledateverysecondpixelhorizontallyoneachline).Amacropixelcontains2pixelsin1u_int32. | V210 | 32 | 10-bit4:2:2YCrCbequivalenttotheQuicktimeformatofthesamename. | V422 | 16 | IamtoldthatthisisanupsidedownversionofUYVY. | V655 | 16? | 16bitYUV4:2:2formatregisteredbyVitecMultimedia.Ihavenoinformationonthecomponentorderingorpacking. | VYUY | ? | ATIPackedYUVData(formatunknownbutyoucangetholdofacodecsupportingithere) | Y422 | 16 | DirectcopyofUYVYasusedbyADSTechnologiesPyroWebCamfirewirecamera. | YUY2 | 16 | YUV4:2:2asforUYVYbutwithdifferentcomponentorderingwithintheu_int32macropixel. | YUYV | 16 | DuplicateofYUY2 | YUNV | 16 | AdirectcopyofYUY2registeredbyNVidiatoworkaroundproblemsinsomeoldcodecswhichdidnotlikehardwarewhichofferedmorethan2YUY2surfaces. | YVYU | 16 | YUV4:2:2asforUYVYbutwithdifferentcomponentorderingwithintheu_int32macropixel. | Y41P | 12 | YUV4:1:1(Ysampleateverypixel,UandVsampledateveryfourthpixelhorizontallyoneachline).Amacropixelcontains8pixelsin3u_int32s. | Y411 | 12 | YUV4:1:1withapacked,6byte/4pixelmacroblockstructure. | Y211 | 8 | PackedYUVformatwithYsampledateverysecondpixelacrosseachlineandUandVsampledateveryfourthpixel. | Y41T | 12 | FormatasforY41PbutthelsbofeachYcomponentisusedtosignalpixeltransparency. | Y42T | 16 | FormatasforUYVYbutthelsbofeachYcomponentisusedtosignalpixeltransparency. | YUVP | 24? | YCbCr4:2:2extendedprecision10-bitspercomponentinY0U0Y1V0order.RegisteredbyRichEhlersofEvans&Sutherland. | Y800 | 8 | Simple,singleYplaneformonochromeimages. | Y8 | 8 | DuplicateofY800asfarasIcansee. | Y16 | 16 | 16-bituncompressedgreyscaleimage. |
UYVY存储结构UYVY是一种最常用的YUV4:2:2格式，其数据排列如下： 
3.2PlanarYUVFormatsLabel | Bitsperpixel | Description | YVU9 | 9 | 8bitYplanefollowedby8bit4x4subsampledVandUplanes.RegisteredbyIntel. | YUV9 | 9? | RegisteredbyIntel.,thisistheformatusedinternallybyIndeovideocode | IF09 | 9.5 | AsYVU9butanadditional4x4subsampledplaneisappendedcontainingdeltainformationrelativetothelastframe.(Bppisreportedas9) | YV16 | 16 | 8bitYplanefollowedby8bit2x1subsampledVandUplanes. | YV12 | 12 | 8bitYplanefollowedby8bit2x2subsampledVandUplanes. | I420 | 12 | 8bitYplanefollowedby8bit2x2subsampledUandVplanes. | IYUV | 12 | DuplicateFOURCC,identicaltoI420. | NV12 | 12 | 8-bitYplanefollowedbyaninterleavedU/Vplanewith2x2subsampling | NV21 | 12 | AsNV12withUandVreversedintheinterleavedplane | IMC1 | 12 | AsYV12excepttheUandVplaneseachhavethesamestrideastheYplane | IMC2 | 12 | SimilartoIMC1exceptthattheUandVlinesareinterleavedathalfstrideboundaries | IMC3 | 12 | AsIMC1exceptthatUandVareswapped | IMC4 | 12 | AsIMC2exceptthatUandVareswapped | CLPL | 12 | FormatsimilartoYV12butincludingalevelofindirection. | Y41B | 12? | Weitekformatlistedas"YUV4:1:1planar".Ihavenootherinformationonthisformat. | Y42B | 16? | Weitekformatlistedas"YUV4:2:2planar".Ihavenootherinformationonthisformat. | Y800 | 8 | Simple,singleYplaneformonochromeimages. | Y8 | 8 | DuplicateofY800asfarasIcansee. | CXY1 | 12 | Awaitingclarificationofformat. | CXY2 | 16 | Awaitingclarificationofformat. |
Y42B存储结构 YCbCr4:2:2的方式存储 
4YUV与RGB的转换关系yuv422planartorgb565
转换公式：
R=Y+1.4075*(V-128)
G=Y-0.3455*(U-128)-0.7169*(V-128)
B=Y+1.779*(U-128)
为了加快运算速度，采用下面的整形计算法：
u=YUVdata[UPOS]-128;
v=YUVdata[VPOS]-128; rdif=v+((v*103)>>8);
invgdif=((u*88)>>8)+((v*183)>>8);
bdif=u+((u*198)>>8); r=YUVdata[YPOS]+rdif;
g=YUVdata[YPOS]-invgdif;
b=YUVdata[YPOS]+bdif;
r=r>255?:255:(r<0:?0:r);
g=g>255?:255:(g<0:?0:g);
b=b>255?:255:(b<0:?0:b);
以上得到的是rgb888的数据，再将rgb888转为rgb555
RGBdata[1]=((r&0xF8)|(g>>5));
RGBdata[0]=(((g&0x1C)<<3)|(b>>3));
YUV422PlanarFORMATsize:
size=width*height*2;
YSIZE=size/2;
USIZE=size/4;
VSIZE=size/4;
YPOS=0;
UPOS=YPOS+size/2;
VPOS=UPOS+size/4; | 
转播
