Optimal path finding with space variant metric weights via multilayer CNN-UM

Analogic CNN-based optimal path finding algorithm is proposed to solve the problem with space variant metric weights. The algorithm is based on the analog version of modified dynamic programming which is implemented with nonlinear templates on the multi-layer CNN-UM. The structure is composed of distance computing layer, intermediate layer, and path finding layer. Employment of intermediate layers allows the space variant distance weights to be provided externally and the CNN structure to be compact.