Image segmentation assisted by a triple-camera setup

Image segmentation assisted by a triple-camera setup

Liu, Na (TU Delft Electrical Engineering, Mathematics and Computer Science)

Eisemann, Elmar (mentor)

Delft University of Technology

2018-02-27

Photographers suffer from many challenging problems when separating and compositing images, such as the color camouflage and unpleasant color bleeding artifacts. To tackle these issues, we present a solution to fore- and background segmentation for images taken from multiple cameras, which introduce the parallax that leads to the foreground model projecting to dif- ferent background locations, which has been designed to exhibit a special pattern to ease the object extraction. We optimize the camera distance and model position to ensure the silhouette of the model falls on different colors of the background pattern on the wall. Furthermore, we develop a visual- ization tool based on our algorithm to help photographers effectively setup their studios.

Image segmentation
Multiple Cameras
Background pattern

http://resolver.tudelft.nl/uuid:6783789c-712c-43c9-9e87-bbaac68e2c5a

Student theses

master thesis

© 2018 Na Liu