A system for generating texture maps for 3D models of real-world objects includes a camera and reflective surfaces in the field of view of the camera. The reflective surfaces are positioned to reflect one or more reflected views of a target object to the camera. The camera captures a direct image of the target object and reflected images from the reflective surfaces. An image processor device separates the reflected views/images from the direct image of the target object in the captured image by detecting distortion in the reflected views. The image processor reduces distortion in the reflected views, and generates a texture map based on 3D space characteristics of the target object and on the reflected views. Reducing distortion in the reflected views may include scaling the reflected views to correspond to a size of the target object in the camera field of view.