198.

[Cover] [Contents] [Index]

Page 276

Figure 7.2 General steps in multisource classification based on evidential reasoning and extension of Bayesian theory.

alternative way of normalisation of the data scale, as each different source is mapped on to the probability domain.

7.3.1.3 Multisource consensus

In this final stage, all the probability (or evidence) measures are combined together in order to generate a classified image (Figure 7.2). Where the evidential reasoning method is used for evidence consensus, the data sources are considered in a pair-wise fashion while if the consensus mechanism is based on the extension of Bayesian classification theory then all the probability measures are combined simultaneously.

[Cover] [Contents] [Index]