AN ANALOG HOPFIELD NETWORK TO OVERCOME EXCESSIVE SMOOTHING IN SHAPE FROM SHADING

The regularization term used in the cost function of the shape from sh ading (SFS) problem is necessary to make the problem well-posed. Howev er, it has the disadvantage of producing excessive smoothing of the su rface even at places where the gradient of the surface slope is very h igh. To overcome this we propose th use of a line function depending o n the gradient of the surface slope in the regularization term. For an analog line function. the SFS energy functions is then minimized usin g and analog Hopfield network. For a binary line function, the SFS ene rgy minimization problem is formulated as a combinatorial minimization problem; the minimization is implemental using a Boltzmann machine (s imulated annealing). Simulation results are presented to substantiate our approach.