Approximate orientation steerability based on angular Gaussians

Junctions are significant features in images with intensity variation that exhibits multiple orientations. This makes the detection and characterizati on of junctions a challenging problem. The characterization of junctions wo uld ideally be given by the response of a filter at every orientation, This can be achieved by the principle of steerability that enables the decompos ition of a filter into a linear combination of basis functions. However, cu rrent steerability approaches suffer from the consequences of the uncertain ty principle: In order to achieve high resolution in orientation they need a large number of basis filters increasing, thus, the computational complex ity, Furthermore, these functions have usually a wide support which only ac centuates the computational burden, In this paper we propose a novel alternative to current steerability approa ches, It is based on utilizing a set of polar separable filters with small support to sample orientation information, The orientation signature is the n obtained by. interpolating orientation samples using Gaussian functions w ith small support. Compared with current steerability techniques our approa ch achieves a higher orientation resolution with a lower complexity In addi tion, we build a polar pyramid to characterize junctions of arbitrary inher ent orientation scales.