Elsevier

Cognition

Volume 99, Issue 3, April 2006, Pages 275-325
Cognition

Segmentation of object outlines into parts: A large-scale integrative study

https://doi.org/10.1016/j.cognition.2005.03.004Get rights and content

Abstract

In this study, a large number of observers (N=201) were asked to segment a collection of outlines derived from line drawings of everyday objects (N=88). This data set was then used as a benchmark to evaluate current models of object segmentation. All of the previously proposed rules of segmentation were found supported in our results. For example, minima of curvature (i.e. locations along the contour where negative curvature takes an extreme value) were often used as segmentation points. The second point of a pair connected by a segmentation line often depended on more global shape characteristics such as proximity, collinearity, symmetry, and elongation. Based on these results, a framework is presented in which all of the previously proposed (and now empirically validated) segmentation rules or rules for part formation are integrated.

Section snippets

The minima rule (Hoffman & Richards, 1984)

Hoffman and Richards (1984) proposed that objects are segmented according to the minima rule. Because in their view object segmentation into parts precedes object recognition, it must be based on very generic regularities that hold for almost any object. These regularities are found in the transversality and singularity principle. The transversality principle entails that a concave crease on the surface of an object is a very likely candidate for the segmentation of two forms. The singularity

Aims

In the light of the above state-of-the-art, we thus believe that the field would benefit from a study in which a large set of outlines is segmented by a large number of subjects. In such a benchmark data set, the relative strengths and weaknesses of the different models would be easier to evaluate, and even relatively rare types of parts could reveal themselves. To provide such a benchmark data set is the most important goal of this paper.

We investigate how people segment outlines of existing

Subjects

Two hundred and one subjects, all first-year psychology students at the University of Leuven, participated in this study as a mandatory component of their curriculum. It took place in a large movie theatre used as a lecture hall, following two hours of lectures on the history of philosophy and metaphysics. Our study was only part of a longer session of several (unrelated) paper-and-pencil tests and experiments. It is important to note that these conditions are not ideal to motivate subjects for

Results and discussion

A fairly large number of data was excluded from analysis. There were three categories of faults that led to exclusion. (1) The segmentation line did not cut two points on the outline. Sometimes this was true for the majority of the 22 outlines, sometimes for only a number of outlines per subject. (2) Some stimuli were not segmented at all. (3) There was also a rest category of other ‘errors’, like filling-in the inner details of an outline (e.g. drawing an eye on the horse). In addition to

General discussion

This study started out by formulating three aims. The first goal of this study was to supplement the large literature on object segmentation into parts with good (i.e. reliable and valid) benchmark data. The current literature consists of a wealth of different segmentation principles and rules for part formation but relatively little strong empirical evidence. Most of the experiments so far had used only a small number of subjects and stimuli (mostly meaningless outline shapes). Using outlines

Acknowledgements

This research was supported by a research grant from the University Research Council (OT/00/007) and from the Fund for Scientific Research (FWO-Vlaanderen G.0189.02) to JW. This study is part of a larger research program with financial support from the University Research Council (GOA/2005/03-TBA) to the Laboratory of Experimental Psychology. We would also like to thank Felix Wichmann, Sven Panis, Donald Hoffman and an anonymous reviewer for their helpful comments on a previous draft.

We would

References (95)

  • J. Feldman

    Curvilinearity, covariance, and regularity in perceptual groups

    Vision Research

    (1997)
  • J. Feldman

    The role of objects in perceptual grouping

    Acta Psychologica

    (1999)
  • D.J. Field et al.

    Contour integration by the human visual system: Evidence for a local ‘association field’

    Vision Research

    (1993)
  • M. Hendrickx et al.

    A critique of Leyton's theory of perception and cognition

    Journal of Mathematical Psychology

    (1999)
  • S. Hochstein et al.

    View form the top: Hierarchies and reverse hierarchies in the visual system

    Neuron

    (2002)
  • D.D. Hoffman et al.

    Parts of recognition

    Cognition

    (1984)
  • D.D. Hoffman et al.

    Salience of visual parts

    Cognition

    (1997)
  • P.J. Kellman et al.

    A theory of visual interpolation in object perception

    Cognitive Psychology

    (1991)
  • I. Kovács

    Gestalten of today: Early processing of visual contours and surfaces

    Behavioural Brain Research

    (1996)
  • M. Kubovy et al.

    On the lawfulness of grouping by proximity

    Cognitive Psychology

    (1998)
  • B. Landau et al.

    Object perception and object naming in early development

    Trends in Cognitive Sciences

    (1998)
  • M. Leyton

    Symmetry-curvature duality

    Computer Vision, Graphics, and Image Processing

    (1987)
  • M. Leyton

    Inferring causal history from shape

    Cognitive Science

    (1989)
  • S. Pinker

    Visual cognition: An introduction

    Cognition

    (1984)
  • W. Richards et al.

    Codon constraints on closed 2D shapes

    Computer Vision, Graphics, and Image Processing

    (1985)
  • B.J. Scholl

    Objects and attention: The state of the art

    Cognition

    (2001)
  • P.G. Schyns et al.

    The ontogeny of part representation in object concepts

  • R. Van Lier et al.

    Effects of physical connectivity on the representational unity of multi-part configurations

    Cognition

    (1998)
  • J. Wagemans

    Characteristics and models of human symmetry detection

    Trends in Cognitive Sciences

    (1997)
  • M. Bar

    A cortical mechanism for triggering top–down facilitation in visual object recognition

    Journal of Cognitive Neuroscience

    (2003)
  • G.C. Baylis et al.

    Obligatory edge assignment in vision: The role of figure and part segmentation in symmetry detection

    Journal of Experimental Psychology: Human Perception and Performance

    (1995)
  • Bertamini, M., & Farrant, T. (2005). Detection of change in shape and its relation to part structure: An advantage for...
  • M. Bertamini et al.

    Early computation of contour curvature and part structure: Evidence from holes

    Perception

    (2004)
  • I. Biederman

    Recognition-by-components: A theory of human image understanding

    Psychological Review

    (1987)
  • M.L. Braunstein et al.

    Parts of visual objects: An experimental test of the minima rule

    Perception

    (1989)
  • S.L. Brincat et al.

    Underlying principles of visual shape selectivity in posterior inferotemporal cortex

    Nature Neuroscience

    (2004)
  • C.B. Cave et al.

    The role of parts and spatial relations in object identification

    Perception

    (1993)
  • Claessens, P. M., & Wagemans, J. (2005a). Perceptual grouping in Gabor lattices: Proximity and alignment. Perception...
  • Claessens, P. M., & Wagemans, J. (2005b). Perceptual grouping in discrete periodic patterns and perturbed zigzag...
  • M. Davi et al.

    Frames of reference and distinctive figural characteristics affect shape perception

    Journal of Experimental Psychology: Human Perception and Performance

    (1993)
  • J. De Winter et al.

    Contour-based object identification and segmentation: Stimuli, norms and data, and software tools

    Behavior Research Methods, Instruments, and Computers

    (2004)
  • De Winter, J., & Wagemans, J. (2005a). Part priming of object naming, submitted for...
  • De Winter, J. & Wagemans, J. (2005b). Perceptual salience of points along the contour of everyday objects: A...
  • R. Dhandapani et al.

    Role of scale in partitioning shape

    Proceedings of the International Conference on Image Processing II

    (2002)
  • J. Driver et al.

    One-sided edge-assignment in vision: 2. Part decomposition, shape description, and attention to objects

    Current Directions in Psychological Science

    (1995)
  • J. Driver et al.

    Preserved figure-ground segregation and symmetry perception in visual neglect

    Nature

    (1992)
  • J. Feldman et al.

    Information along contours and object boundaries

    Psychological Review

    (2005)
  • Cited by (101)

    • Near-convex decomposition of 2D shape using visibility range

      2021, Computer Vision and Image Understanding
    • Revisiting the medial axis for planar shape decomposition

      2019, Computer Vision and Image Understanding
      Citation Excerpt :

      Comparing to the conventional definition, which requires part-cuts to cross an axis of local symmetry (Singh and Hoffman, 2001), this is a stronger definition in agreement with the definition of necks (Siddiqi and Kimia, 1995). We actually show that this can be in accordance to psychophysical evidence (De Winter and Wagemans, 2006) in some cases. In general, some ground-truth cuts may be lost but we introduce a way to recover them.

    View all citing articles on Scopus
    View full text