Corner Detection Using Spatial Moment
Chen Xing-lin · Journal of Optoelectronics·laser · 2005
A novel corner detection method based on spatial moments was presented in this paper.Firstly,based on the Kitchen′s corner measure and the Wang′s corner measure,gradient-direction change and image gradient are used as one decision rule of the corner detection.Then,by analyzing the corner-model,spatial-moment polynomials are employed as another decision rule of the corner detection.The corner feature can be detected by the two rules,but vertex of the corner can't be accurately localized.Therefore,non-max suppression method is used for reducing false points and improving localizaion.Experiment results illustrate that spatial moment corner detection approach has more accuracy and robustness than CSS and COP methods.