In this paper, we propose a new method to generate a continuousbelief functions from a multimodal probability distribution function definedover a continuous domain. We generalize Smets' approach in the sense thatfocal elements of the resulting continuous belief function can be disjoint setsof the extended real space of dimension n. We then derive the continuousbelief function from multimodal probability density functions using the leastcommitment principle. We illustrate the approach on two examples of probabilitydensity functions (unimodal and multimodal). On a case study of Search AndRescue (SAR), we extend the traditional probabilistic framework of search theoryto continuous belief functions theory. We propose a new optimization criterionto allocate the search effort as well as a new rule to update the informationabout the lost object location in this latter framework. We finally compare theallocation of the search effort using this alternative uncertaintyrepresentation to the traditional probabilistic representation.
