A Neural Network Approach for Three-Dimensional Object Recognition
Volker Tresp · Neural Information Processing Systems · 1990
The model-based neural vision system presented here determines the position and identity of three-dimensional objects. Two stereo images of a scene are described in terms of shape primitives (line segments derived from edges in the scenes) and their relational structure. A recurrent neural matching network solves the correspondence problem by assigning corresponding line segments in right and left stereo images. A 3-D relational scene description it then generated and matched by a second neural network against models in a model base. The quality of the solutions and the convergence speed were both improved by using mean field approximations.