Modelling and identification of characteristic intensity variations

摘要

An approach is introduced for the modelling and identification of a certain class of characteristic intensity variations resulting essentially from polygonal structures of a depicted 3D-scene. We develop a general analytical model for the structural grey-value transitions in an image which is the superposition of elementary model functions. Special cases of this general model are the grey-value variations of step edges, grey-value corners (L-junctions), T-, Y-, ARROW-junctions, and all other junction types represented in the labelling system of Waltz1. It will be shown that this parametric model agrees fairly well with real image intensities. Estimates of the unknown model parameters are found by minimizing the sum of squared differences between the model and the observed grey values. The approach has been tested on real image data.