Automatic photo pop-up

TLDR

The model consists of texture‑mapped planar billboards and mirrors the complexity of a typical children's pop‑up book illustration, yet the inherent ambiguity and statistical nature of the approach mean it may not work on every image. The study introduces a fully automatic method for generating a 3D pop‑up model from a single photograph by statistically modeling geometric classes defined by scene orientations. The algorithm first labels image regions as ground, sky, or vertical, then uses these coarse categories to cut and fold the image into a pop‑up model under simple assumptions. The method performs surprisingly well across a wide range of scenes from typical personal photo albums. However.

Abstract

This paper presents a fully automatic method for creating a 3D model from a single photograph. The model is made up of several texture-mapped planar billboards and has the complexity of a typical children's pop-up book illustration. Our main insight is that instead of attempting to recover precise geometry, we statistically model geometric classes defined by their orientations in the scene. Our algorithm labels regions of the input image into coarse categories: "ground", "sky", and "vertical". These labels are then used to "cut and fold" the image into a pop-up model using a set of simple assumptions. Because of the inherent ambiguity of the problem and the statistical nature of the approach, the algorithm is not expected to work on every image. However. it performs surprisingly well for a wide range of scenes taken from a typical person's photo album.