Abstract: The least-squares model is capable of reconstructing smooth surfaces,but the depth discontinuity features can be blurred.In order to solve this problem,a spatially adaptive parameter is constructed by introducing the productivity term and Geary-Hinkley transform into the least squares model,an anisotropic weighted least-squares model is established;then the model is discretized by the 4-direction discretization formula,and the optimality condition is given;finally,the Krylov subspace method is used to solve the model.The numerical experimental results show that the anisotropic weighted least-squares model can keep the depth discontinuity of the surface,and the reconstruction effect is better than that of the least squares model.

Key words: integrability term, Geary-Hinkley transform, depth discontinuity, adaptive parameters