We propose a variational framework for the three-dimensional volume reconstruction from 2-D slices. The proposed method is based on modified Cahn-Hilliard equation. To satisfy constraints accurately as well as obtain a smooth result, we propose pre-smoothing procedure. After splitting a grayscale image into binary channels, we perform binary inpainting. Then we introduce smoothing and shock filter when combining binary results. We applied our results for reconstruction of 3-D human body from slices of CT images.

This presentation is part of Contributed Presentation “CP10 - Contributed session 10”

Authors:

Junwoo Kim (KAIST)

Chang-Ock Lee (KAIST)

Keywords:

cahn–hilliard equation, computer vision, image inpainting, image reconstruction, volume reconstruction