Modeling the television process

Advances in digital signal processing have led to a renewed interest in basic television research. Much of this attention is devoted to the processing of the video signal in the channel, while past research has focused on analytical models of particular television devices. In this thesis, a general spatiotemporal model is developed that includes all important camera, channel, and display processes. Numerical simulation is used to study the effect of system parameters on image quality at the display. TVSIM, a TV simulation program, is developed as a tool for the design and analysis of a general monochrome baseband TV system. It models (1) the image illuminance falling on a camera target, (2) the physical scanning operation of a tube-type or solid-state camera, (3) the generation, processing, and transmission of the video signal, and (4) the physical scanning operation of a tubetype or solid-state display. The final output is an image representing the detailed luminance pattern on a hypothetical display. This image may be further processed by a model of the human visual system, or it may be compared with other processed images. TVSIM and other simulation software are used to study a number of basic television processes. Several charge scanning models are developed and compared; it is shown that destructive readout in camera tubes enhances high-frequency detail while introducing phase distortion. The effect of camera and display apertures on image quality is investigated. Simulation results show how physically different camera lag processes can yield similar lag responses. Temporal integration patterns are used to explain and model the motion rendition characteristics of common imaging devices. Standard, extended-definition, and high-definition TV systems are simulated and compared. Finally, two adaptive interpolation methods for television are described. One method uses Gaussian beam shaping to reduce scan line visibility; the other uses digital contour interpolation to reduce aliasing. Thesis Supervisor: William F. Schreiber, Professor of Electrical Engineering * AT&T Bell Laboratories Scholar -----'--111 ---