Automatic target recognition using boundary partitioning and invariant features in forward-looking infrared images

Abstract

We propose an automatic target recognition (ATR) algorithm for recognizing nonoccluded and partially occluded military vehicles in natural forward-looking infrared (FLIR) images. The proposed algorithm consists of global and local feature extraction from partitioned boundaries of a target, and a new classification method using multiple multilayer perceptrons (MLPs). After segmenting a target, the target contour is partitioned into four local boundaries. Radial and distance functions are defined from the target contour and local boundaries, and are used to define global and local shape features, respectively. The global and local shape features are more invariant to similarity transform than traditional feature sets. Four feature vectors are composed of the global and local shape features, and are used as inputs of MLPs. The outputs of MLPs are combined to recognize nonoccluded and partially occluded targets. In the experiments, we show that the proposed features are superior to the traditional feature sets with respect to invariance and recognition performance.