Documents
Submitted for publication
[PDF]
- Download complete
set of results and segmentations on Berkeley Segmentation data set.
Abstract:
We introduce robust variational segmentation techniques that are driven by an
Edgeflow vector field. Variational image segmentation has been widely used
during the past ten years. While there is a rich theory of these techniques in
the literature, a detailed performance analysis on real natural images is
needed to compare the various methods proposed. In this context, this paper
makes the following contributions: (a) designing curve evolution and
anisotropic diffusion methods that use Edgeflow vector fields to obtain good
quality segmentation results over a large and diverse class of images, and (b)
a detailed experimental evaluation of these segmentation methods. Our
experiments show that Edgeflow-based anisotropic diffusion outperforms other
competing methods by a significant margin.
Accepted for publication in IEEE Trans. on Pattern Analysis and
Machine Intelligence (PAMI)
[PDF]
Abstract:
In this paper we introduce new variational segmentation cost functions that
are based on pair wise similarities or dissimilarities of the pixels. As a
solution to the minimization problem, we introduce a new curve evolution
framework, the graph partitioning active contours (GPAC). Using global
features, our curve evolution is able to produce results close to the ideal
minimization of such cost functions. New and efficient implementation
techniques are also introduced in this paper. Comparisons to graph
partitioning methods such as normalized cuts show that GPAC solution is more
effective and computationally more efficient. Experiments on gray scale,
color, and texture images show promising segmentation results.
|
