Many applications of computer graphics involve the display of a three-dimensional solid reconstructed from a sequence of two-dimensional planar contours. Surface construction algorithms accomplish this by mapping individual pairs of contours, forming triangular surface patches, that approximate the original three-dimensional solid. In this paper, we present an expanded algorithm that not only handles the mappings of multiple contours per plane and partial contour mappings, but also allows human interaction to resolve mapping problems. We include a discussion of our algorithm's limitations and the proposed solutions to those limitations.