Multifrequency absolute phase estimation via graph cuts

Many imaging systems, e.g., interferometric synthetic aper- ture radar (InSAR), yield phase images. These systems re- trieve the phase up to a modulo-2π rad ambiguity, i.e., the phase is wrapped into the principal interval ( π π). Phase unwrapping (PU) is, then, a crucial inverse problem to ob- tain absolute phase, which is what embodies physical infor- mation. If the phase difference between neighboring pixels is less than π rad, then, phase unwrapping can be obtained un- ambiguously. This, however, is not always the case. For ex- ample, in InSAR, where absolute phase is proportional to the terrain elevation, we often face neighbor phase differences much larger than π rad. The PU problem is even more chal- lenging for noisy images. This paper proposes a diversity approach, which consists of using two (or more) images of the same scene acquired with different frequencies. Diver- sity grants an enlargement of the ambiguity interval ( π π), thus, allowing to unwrap images with high phase rates. Fur- thermore, this paper presents a multi-resolution techniqu e to make denoising. We formulate both tasks as integer op- timization problems, which we tackle by using graph cuts techniques. We illustrate the effectiveness of our method- ology by showing experimental results, which are, to our knowledge, state-of-the art competitive.