Wavelet-based cascade model for intermittent structure in terrestrial environments

TL;DR

This paper introduces a wavelet-based cascade model that captures the self-similar, intermittent distribution of objects in terrestrial environments, useful for realistic scene synthesis and system performance prediction.

Contribution

It presents a novel self-similar cascade model using wavelet and fractal concepts to describe complex spatial distributions of objects in natural and man-made environments.

Findings

Model applies to diverse phenomena like turbulence and urban landscapes

Enables synthesis of realistic terrestrial scenes

Predicts sensing and communication system performance in complex environments

Abstract

A wavelet-like model for distributions of objects in natural and man-made terrestrial environments is developed. The model is constructed in a self-similar fashion, with the sizes, amplitudes, and numbers of objects occurring at a constant ratios between parent and offspring objects. The objects are randomly distributed in space according to a Poisson process. Fractal supports and a cascade model are used to organize objects intermittently in space. In its basic form, the model is for continuously varying random fields, although a level-cut is introduced to model two-phase random media. The report begins with a description of relevant concepts from fractal theory, and then progresses through static (time-invariant), steady-state, and non-steady models. The results can be applied to such diverse phenomena as turbulence, geologic distributions, urban buildings, vegetation, and arctic ice floes. The model can be used as a basis for synthesizing realistic terrestrial scenes, and for predicting the performance of sensing and communication systems in operating environments with complex, intermittent distributions of scattering objects.