Abstract

We propose to construct a 3D linear image basis which spans an image space of arbitrary illumination conditions, from images of a moving ob ject observed under a static lighting condition. The key ad- vance is to utilize the ob ject motion which causes illumination variance on the ob ject surface, rather than varying the lighting, and thereby sim- plifies the environment for acquiring the input images. Since we then need to re-align the pixels of the images so that the same view of the ob ject can be seen, the correspondence between input images must be solved despite the illumination variance. In order to overcome the problem, we adapt the recently introduced geotensity constraint that accurately gov- erns the relationship between four or more images of a moving ob ject. Through experiments we demonstrate that equivalent 3D image basis is indeed computable and available for recognition or image rendering.