Adaptive fusion for satellite images is not as easy as simple conjunctive or disjunctive fusions . The problem arises from:

• the nature of the data to be fused (scalar values and not possibility distributions) and
• a great number of sources to be fused.

Adaptive fusion was defined to fuse distributions of possibilities while values to fuse are scalar which represent the degrees of membership of a pixel to a class, given by each source. It is thus necessary to modify the scheme of calculation of the consensus degree h between the sources to deal with this important constraint.

• Global Degree of Consensus
• Degree of Consensus Local to each Pixel