Automating the Creation of Parallax Effects from Images
Sofie Kamber
Bachelor's Thesis, August 2023
Supervisors: Henry Raymond, Börge Scheel, Dr. Fabio Zünd, Prof. Dr. Bob Sumner
Abstract
The parallax effect refers to the apparent shift in the position of objects as observed from different viewpoints. In this thesis, we introduce a method for generating parallax effects from images, specifically by creating distinct layers with automation tools. The primary steps of this automation involve inpainting, segmentation, and depth estimation.
To achieve this objective, we developed an editor that integrates automation tools for inpainting and segmentation. This approach enables users to create parallax effects with more ease and efficiency compared to manual creation. Subsequently, we conducted a user study to evaluate the editor’s performance, highlighting the effectiveness of these integrated tools.
Although our study revealed certain limitations in terms of the quality of parallax effects produced by the editor, we successfully created a platform that allows individuals without specialized skills to generate parallax effects in significantly less time compared to manual methods.