A scene matching algorithm based on the knowledge of object edges

Describes an algorithm which based on the knowledge of image edges to perform a phase-correlation scene matching. Scene matching is a key technique in computer vision. The phase correlation algorithm uses “Fourier phase information”, i.e. the phase function of the Fourier spectrum of two images, to construct the scene matching algorithm. This paper gives a new algorithm in which the matching process is based on the edge knowledge of the image. The computational cost of this algorithm is more than that of the grey phase correlation algorithm, but its matching effect is much better than that of the latter in the presence of noise in the images