Abstract: A method, system and computer-readable medium for converting objects in two-dimensional images into three-dimensional objects. Vanishing points for an object selected from a digitized two-dimensional perspective image are determined. Then, a three-dimensional coordinate space for the selected object is derived based on the determined vanishing points. The system then generates a three-dimensional object in the generated three-dimensional coordinate space based on user-designated vertices at significant features of the selected object in the two-dimensional perspective image. The three-dimensional coordinate space and three-dimensional object are generated using non-linear optimization algorithms. The system extracts surface texture information from the two-dimensional images and maps the extracted texture information onto the generated three-dimensional object.