Scene Shape Priors for Superpixel Segmentation

Abstract

Unsupervised over-segmentation of an image into superpixels is a common preprocessing step for image parsing algorithms. Superpixels are used as both regions of support for feature vectors and as a starting point for the final segmentation. In this paper we investigate incorporating a priori information into superpixel segmentations. We learn a probabilistic model that describes the spatial density of the object boundaries in the image. We then describe an over-segmentation algorithm that partitions this density roughly equally between superpixels whilst still attempting to capture local object boundaries.

We demonstrate this approach using road scenes where objects in the center of the image tend to be more distant and smaller than those at the edge. We show that our algorithm successfully learns this foveated spatial distribution and can exploit this knowledge to improve the segmentation. Lastly, we introduce a new metric for evaluating vision labeling problems. We measure performance on a challenging real-world dataset and illustrate the limitations of conventional evaluation metrics

Title: Scene Shape Priors for Superpixel Segmentation

Author: Alistair P. Moore
Author: Simon J.D. Prince
Author: Jonathan Warrell
Author: Umar Mohammed
Author: Graham Jones

Publication: Computer Vision, IEEE 12th International Conference 2009 | full text (PDF)

Year: 2009

D.O.I: 10.1109/ICCV.2009.5459246

ISBN: Insert ISBN Here

Tags: Alistair P. Moore Simon J.D. Prince Graham Jones Jonathon Warrell object boundary Sharp Laboratories Europe superpixel Umar Mohammed