A new shape descriptor for object recognition
Misheng Xue, Licheng Xu, Gang Zhou · 2011
In this paper ,a new shape descriptor for object recognition which is invariant to translation ,rotation,scale and mirror mapping is proposed. According to the results obtained by a edge detector , the distance between the edge points and the centroid of the object is calculated, and a series of standardization are cited by which the new shape descriptor was established. The experiment demostrated that the new method achieves a significant performance comparing to the Fourier descriptor and moment invariants.