Choosing a Vision Model for Optimizing DPCM Video

The efficient transmission of video signals can be effected by the reduction of statistical redundancy, based on linear prediction. It is clear that economic transmission of video signals must be based on the characteristics of human vision where man is the end user. There are several characteristics of sight which can be used for this purpose. Contrast sensitivity is the important characteristic of vision for the predictive system. We consider the case of the Laplacian curve as an approximation for the contrast sensitivity of vision. It is shown that the differences between a Gaussian and a Laplacian approximation to the contrast sensitivity characteristic are insignificant.