November 2002
Volume 2, Issue 7
Vision Sciences Society Annual Meeting Abstract  |   November 2002
Automatic line-drawings extraction from complex scenes
Author Affiliations
  • Chunhong Zhou
    University of Southern California, USA
Journal of Vision November 2002, Vol.2, 501. doi:10.1167/2.7.501
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Chunhong Zhou, Gary Holt, Bartlett Mel; Automatic line-drawings extraction from complex scenes. Journal of Vision 2002;2(7):501. doi: 10.1167/2.7.501.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Line drawings contain most of the shape-related information in complex scenes. From a technical standpoint, however, automatic extraction of line drawings from raw images has proven extremely difficult, primarily because shape contours are influenced by widely dispersed visual features acting in subtle geometric combinations. To address this problem, we developed a recurrent network architecture inspired by the interconnection circuitry of primary visual cortex, which incorporates several representational biases tailored to the task of long-range visual contour integration. A learning scheme was used to train the modifiable parameters of the network to capture the statistical regularities of contour shape. Two separate inhibitory subsystems, one feedforward, one feedback, modulate contour predictions. We find that when the trained network is applied to complex images, well-organized contours are selectively boosted and texture edges are selectively suppressed, leading to a rough line-drawing-like sketches of the visual scene.

Zhou, C., Holt, G., Mel, B.(2002). Automatic line-drawings extraction from complex scenes [Abstract]. Journal of Vision, 2( 7): 501, 501a,, doi:10.1167/2.7.501. [CrossRef]

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.