Abstract
In this paper we propose classi er ensembles that use multiple Pareto image features for invariant image identi cation. Di erent from traditional ensembles that focus on enhancing diversity by generating diverse base classi ers, the proposed method takes advantage of the diversity inherent in the Pareto features extracted using a multi-objective evolutionary Trace Transform algorithm. Two variants of the proposed approach have been implemented, one using multilayer perceptron neural networks as base classi ers and the other k-Nearest Neighbor. Empirical results on a large number of images from the Fish-94 and COIL-20 datasets show that on average, ensembles using Pareto features perform much better than traditional classi er ensembles using the same features and data randomization. The better classi cation performance of the proposed ensemble is further supported by diversity analysis using a number of measures, indicating that the proposed ensemble consistently produces a higher degree of diversity than traditional ones. Our experimental results demonstrate that the proposed classi er ensembles are robust to various geometric transformations in images such as rotation, scale and translation, and to additive noise.