I see in a paper mentioning ycbcr as well as ycbycr. I am wondering what is the difference between the two? In the figure, the first two lines show ycbycr but the two bottom lines show ycbycr?
Brightness (Y) component is sampled at a twice rate compared to the color components in YCbYCr coding. Such coding makes use of a fact that human eye is more sensitive to brightness than to color and thus helps to limit bitrate while maintaining acceptable visual quality. Decimated color component samples are reconstructed by interpolation. Alternatively, YCbCr honestly transmits color components for each pixel.
That can be compared to a well-known Bayer mask that samples G component twice as fast compared to R and B because human eye is more sensitive to it.