Object-based image labeling through learning by example and multi-level segmentation

Xu Y., Duygulu P., Saber E., Tekalp A., Yarman-Vural F.

PATTERN RECOGNITION, vol.36, no.6, pp.1407-1423, 2003 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 36 Issue: 6
  • Publication Date: 2003
  • Doi Number: 10.1016/s0031-3203(02)00250-9
  • Journal Name: PATTERN RECOGNITION
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1407-1423
  • Middle East Technical University Affiliated: No

Abstract

We propose a method for automatic extraction and labeling of semantically meaningful image objects using "learning by example" and threshold-free multi-level image segmentation. The proposed method scans through images, each of which is pre-segmented into a hierarchical uniformity tree, to seek and label objects that are similar to an example object presented by the user. By representing images with stacks of multi-level segmentation maps, objects can be extracted in the segmentation map level with adequate detail. Experiments have shown that the proposed multi-level image segmentation results in significant reduction in computation complexity for object extraction and labeling (compared to a single fine-level segmentation) by avoiding unnecessary tests of combinations in finer levels. The multi-level segmentation-based approach also achieves better accuracy in detection and labeling of small objects. (C) 2003 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.