Robust Corner Detection Algorithm Based on Angle Accumulation

For detecting image corner better,a robust corner detection algorithm based on the angle accumulation is proposed.Firstly,the angle accumulation concept is defined to reveal the feature of edge points; then one can obtain the candidate corners under single-chord setting for avoiding loss of adjacent corners.Secondly,a local adaptive threshold as well as the extraction method of supporting region and coefficients is presented to remove the round corners.Finally,the product of angle accumulations is computed for obtaining corners’ feature value under multi-chord,and a global threshold is constructed for removing false corners.The experimental results show that the proposed algorithm achieves better accuracy of corner detection,and behaves better in repeatability and localization error than state-of-the-art algorithms.