Segmentation of Natural Images by Texture and Boundary Compression

  • Authors:

  • Hossein Mobahi;Shankar R. Rao;Allen Y. Yang;Shankar S. Sastry;Yi Ma

  • Affiliations:

  • Coordinated Science Lab, University of Illinois, Urbana, USA 61801;HRL Laboratories, LLC, Malibu, USA 90265;Cory Hall, Department of EECS, University of California, Berkeley, USA 94720;Cory Hall, Department of EECS, University of California, Berkeley, USA 94720;Coordinated Science Lab, University of Illinois, Urbana, USA 61801 and Visual Computing Group, Microsoft Research Asia, Beijing, China

  • Venue:
  • International Journal of Computer Vision
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present a novel algorithm for segmentation of natural images that harnesses the principle of minimum description length (MDL). Our method is based on observations that a homogeneously textured region of a natural image can be well modeled by a Gaussian distribution and the region boundary can be effectively coded by an adaptive chain code. The optimal segmentation of an image is the one that gives the shortest coding length for encoding all textures and boundaries in the image, and is obtained via an agglomerative clustering process applied to a hierarchy of decreasing window sizes as multi-scale texture features. The optimal segmentation also provides an accurate estimate of the overall coding length and hence the true entropy of the image. We test our algorithm on the publicly available Berkeley Segmentation Dataset. It achieves state-of-the-art segmentation results compared to other existing methods.