Visual models are often used to analyze the performance of image processing systems. Two of the leading models are the Daly and the Sarnoff model which have been designed to predict the visibility of luminance differences between static input images. They accomplish this by attempting to reproduce the functional responses of every physiological mechanism in the visual pathway of the brain.

These two models are based on the same set of psychophysical facts about human vision. Therefore, they have a similar basic architecture and some similar mechanistic features. However, the Daly and the Sarnoff models take totally different approaches to modeling visual perception: the frequency domain approach and the spatial domain approach respectively.

A comparison of these two models is made based on a detailed description of their structures and on detection test results. Similarities and differences of both models are discussed along with their strengths and weaknesses.