Footprint Boundary Finding Based on DFD Parametrically Deformable Models
Cunbao Li · Journal of Information Engineering University · 2007
For image objects constrained by a certain shape mode and smooth locally,this paper proposes a boundary finding method based on DFD(Discrete Fourier Descriptors)parametrically deformable model.For extracting boundary with fine information,one must optimize the model on a larger set of parameters,which causes trouble in the balance of computing cost and accuracy.By applying genetic algorithms,a global optimal searching algorithm with better numerical stability and implied parallel computing mode,and introducing prior shape information in selecting initial group,this paper improves the performance of GAs as well as the speed of optimal searching.Better results are achieved in the application of this method on the boundary finding of weight bearing areas in footprint images.