Introduction
In this paper, we present a novel cascaded framework to solve
a self-portrait enhancement problem we call "1+N" problem,
in which a self-portrait is enhanced with the help of N supporting
photos that share the same scene and similar shooting
time. The key idea is to exploit the extra information of these
N photos to expand the field of view of the self-portrait and
improve its lighting style. We achieve this by alternatingly optimizing
two complementary tasks, namely illumination unification
and photo registration. Based on the correspondences
extracted in the input 1+N photos, our method estimates and
updates the illumination and registration coefficients in a cascaded
manner. Then a Markov Random Field formulation
is proposed to globally fuse the aligned photos. Experimental
results demonstrate the proposed method achieves highquality
results in this novel application scenario.
Experimental Results
We evaluated the performance of the proposed method and
compared with Scene Collage [SC] and Photoshop PhotoMerge
tool [PS] on eight photo collections taken by iPhone 6 Plus
and Honor 7. The whole photos and experimental results have
been released on our website1.
Experiment 1:
We show the results of different photo collections in Fig.1~8. In each collection, the first row shows the raw self-portrait and supporting photos.
The lighting style references It are marked with red boxes.
The second row shows the enhanced results obtained by [SC] and [PS].
The third row shows the original self-portrait (we expand it with black region to match the size of the enhanced one for better comparison) and our enhanced result.
References
[SC] Y. Nomura, L. Zhang, and S. K. Nayar, “Scene collages and flexible camera arrays,” in Proc. Eurographics Conf. Rendering Techniques, 2007, pp. 127–138.
http://www1.cs.columbia.edu/CAVE/projects/scene_collage/scene_collage.php.
[PS] Adobe: Photoshop PhotoMerge,
http://helpx.adobe.com/en/photoshop/using/create-panoramic-images-photomerge.html
Back to Projects Page