Optimal decision trees for local image processing algorithms

In this paper we present a novel algorithm to synthesize an optimal decision tree from OR-decision tables, an extension of standard decision tables, complete with the formal proof of optimality and computational cost analysis. As many problems which require to recognize particular patterns can be modeled with this formalism, we select two common binary image processing algorithms, namely connected components labeling and thinning, to show how these can be represented with decision tables, and the benefits of their implementation as optimal decision trees in terms of reduced memory accesses. Experiments are reported, to show the computational time improvements over state of the art implementations.

► We show how to convert OR-decision tables to optimal decision trees.
► We provide the pseudocode of a dynamic programming solution.
► We give a formal proof of optimality for the dynamic programming algorithm.
► We show how to use the optimized tree to speedup local image processing algorithms.