Chapter 21
Video Sampling Rate and Standards Conversion
What Will We Learn?
- What is sampling rate conversion?
- What are the main practical aspects involved in converting a video sequence from one format to another?
- Which steps are involved in PAL to NTSC (and vice versa) standard conversion?
21.1 Video Sampling
As discussed in Chapter 20, although video signals vary continuously in space and time, most TV cameras capture a video sequence by sampling it in the temporal and vertical dimensions. The resulting signal is stored in a 1D raster scan, which is a concatenation of the intensity (or color) variations along successive horizontal scan lines. A third type of sampling (along each line) takes place when an analog video is converted to a digital representation. In all three cases, the fact that selected intensity values are being sampled (while others are being left off) implies that there might be some loss in the process. To minimize such losses, the original video sequence should be sampled at the highest possible sampling rate.
Based on Nyquist's sampling theorem, the sampling rate in each dimension (
, 
, or 
) should be at least twice the highest frequency along that direction. ...
