UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Implementation of an experimental facility and modeling studies for time varying images. Jensen, Olav Velling

Abstract

A wealth of experiments have been performed studying image encoding techniques as applied to non-time varying or single-frame images. However, to date little work has been done to apply these techniques to time varying images, with most of such works emphasizing various ad hoc redundancy reduction techniques. In this work, a computer based experimental system is implemented which makes more methodological studies of time varying images possible. Particular attention is devoted to obtaining very accurate inter-frame registration and uniform quantization of the images. Using this system, a selection of 35 mm movie film images are digitized and stored on computer magnetic tape in a format compatible with many other computing installations, providing a standard data base for future experiments. An often used model for describing picture data is the stationary Gauss-Markov model. In this work, the appropriateness of this model for describing time varying images is studied by comparing the autocorrelation functions as described by the model and as obtained by computation from the picture data. These results indicate that the autocorrelation function is best described by a function which is separable in the time dimension and nonseparable in the spacial dimensions. A number of DPCM communication systems are then studied as a vehicle for evaluating the effect of using the Gauss-Markov model. These results indicate that, for the sample images studied here, the estimated performance using the Gauss-Markov model is good when the model is a good fit to the first data point of the computed autocorrelation function.

Item Media

Item Citations and Data

License

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

Usage Statistics