Local contour characteristics as determinants of pattern discrimination

Area of Science:

• Visual perception
• Cognitive psychology
• Computational neuroscience

Background:

• Understanding how humans perceive and discriminate complex shapes is crucial in visual cognition.
• Local contour features significantly influence global shape perception.
• Previous research has explored various visual features, but the role of contour complexity remains nuanced.

Purpose of the Study:

• To investigate the impact of local contour characteristics on the discrimination of complex random polygons.
• To determine which types of local contour features (e.g., complexity, element size) are most critical for successful shape discrimination.
• To analyze the effects of presentation time and interstimulus interval on discrimination performance.

Main Methods:

• Participants discriminated between complex random polygons sharing global shapes but differing in local contour details.
• Systematic variation of local contour features, including element size and complexity.
• Controlled presentation times and interstimulus intervals were employed.

Main Results:

• Discrimination accuracy was highest when local contour features consisted of a few long line segments or were highly complex.
• Features with intermediate element sizes and complexity were less effective for discrimination.
• Significant effects of presentation time and interstimulus interval, along with their interaction, were observed.

Conclusions:

• The effectiveness of local contour features in shape discrimination is non-linear, favoring extremes of simplicity or complexity.
• A two-factor model can explain the influence of contour variables on visual discrimination.
• These findings contribute to models of visual feature processing and shape recognition.