Visual Pattern Recognition Models of Infrastructure Elements vs. Depth-Encoded Hough Voting