Robust algorithm for detection of image features

Abstract

Authors proved existence of uniformly most powerful invariant algorithm based on the t-test. Conducted study allowed to synthesize decision rule for detection of image features on 3×3 pixel patch was found. Simulation proved stability of the proposed feature point detection algorithm to change of mean value and standard deviation of background pixels' intensity. Versatility of detection algorithm determined only by set of pixels in the signal sample. Uniqueness of detected features determined by formation of support and analyzed samples. Authors obtained equations that allow to assess the effectiveness of the robust feature detector.