Learning a Classification Model for Segmentation

  • Authors:
  • Xiaofeng Ren;Jitendra Malik

  • Affiliations:
  • -;-

  • Venue:
  • ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

We propose a two-class classification model for grouping. Humansegmented natural images are used as positive examples. Negativeexamples of grouping are constructed by randomly matching humansegmentations and images. In a preprocessing stage an image is oversegmented into superpixels. We define a variety of features derivedfrom the classical Gestalt cues, including contour, texture,brightness and good continuation. Information-theoretic analysis isapplied to evaluate the power of these grouping cues. We train alinear classifier to combine these features. To demonstrate thepower of the classification model, a simple algorithm is used torandomly search for good segmentations. Results are shown on a widerange of images.