Texture mapping adds a (photo-)realistic impression to a give n 3-D model by linking each surface path of the model with an image, called texture. Often, pre-defined, synthetic textures are used for the 3-D models. Here, the real texture of the model is gathered by moving a camera around the real object in question.
The sequence of texture images is then mapped to the real 3-D model acquired by the hand-guided scanning device. Technically speaking, the texture mapping process requires
- A calibrated camera,
- A device tracking the camera position,
- A 3-D object model,
- A common position reference with the 3-D model.
By knowing the exact position and direction of view of the camera shots, the images can be projected onto the 3-D model located at a known position. Fig 1-5 show the single steps in texture mapping.
First, a 3-D model of the object is build. Then, images are shot with the hand-guided device from different positions around the object. The images are undistorted and finally mapped to the surface by knowing the exact position of the cameras and the object.
The process leads to a photorealistic virtual model of the object. Generally, the texture images and 3-D model of the object can be acquired either contemporaneously or sequentially thanks to the highly integrated multi-sensor hand-guided scanning device.