Image-based simulation of degradation events

Some events can be described or supervised using a visual approach, so that a simulation process could assist operations of both image synthesis and analysis. Appearance features can be extracted from source images and then inserted into computational models for analyzing scenes subject to some perturbation functions. Among the different parameters which can be considered, texture is particularly appropriate to characterize surfaces and to be processed for bothimage generation and examination. In this paper a model is proposed able to define a class of simulators useful to study evolving events represented by sequences of images, eventually being degraded by some noise processes. In particular, a simulator has been implemented to evaluate the behaviour of materialsubject to conditions of downgrading or depreciation due to natural and/or artificial agents. This simulator performs texture-based image interpretation and generation.