A new fast efficient non-maximum suppression algorithm based on image segmentation
<span>In this paper, the problem of finding local extrema in grayscale images is considered. The known non-maximum suppression algorithms provide high speed, but only single-pixel extrema are extracted, skipping regions formed by multi-pixel extrema. Morphological algorithms allow to</span><span>extract all extrema but its maxima and minima are processed separately with high computational complexity by iterative processing based on image reconstruction using image morphological dilation and erosion. In this paper a new fast efficient non-maximum suppression algorithm based on image segmentation and border analysis is proposed. The proposed algorithm considers homogeneous areas, which are formed by multi-pixel extrema and are the local maxima or minima in relation to adjacent areas, eliminating iterative processing of non-extreme pixels and assigning label numbers to local extrema during their search. The proposed algorithm allowed to increase the accuracy of local extremum extraction in comparison with known non-maximum suppression algorithms and reduce the computational complexity and the use of RAM in comparison with the morphological algorithms.</span>