J. Edward Swan II

Relative Effects of Color-, Texture-, and Density-Coding on Visual Search Performance and Subjective Preference

Gary Perlman and J. Edward Swan II. Relative Effects of Color-, Texture-, and Density-Coding on Visual Search Performance and Subjective Preference. In Proceedings of the 38th Annual Meeting of the Human Factors and Ergonomics Society, pp. 235–239, HFES, October 1994.

Download

[PDF] 

Abstract

Previously, it had been found that texture-coding was ineffective at reducing search time (Perlman & Swan, 1993). In the experiment reported here, 16 subjects searched for blank-, color-, texture-, and density-coded targets of varying complexity in a naturalistic task. The data showed that all non-blank methods were significantly and about equally more effective at reducing search time than blank-coding (no coding). The difference of outcome with previous results is explained by task simplification and by the control of possibly confounding factors. The difference suggests that coding techniques using texture, and possibly other methods, should be evaluated in context. The similar performance of color-, texture-, and density-coding is explained by the use of equal-saturation and equal-brightness colors. Recommendations for the design of effective coding methods and for future research are discussed.

BibTeX

@InProceedings{HFES94-bar, 
  author =      {Gary Perlman and J. Edward {Swan~II}}, 
  title =       {Relative Effects of Color-, Texture-, and Density-Coding on 
                 Visual Search Performance and Subjective Preference}, 
  booktitle =   {Proceedings of the 38th Annual Meeting of the 
                 Human Factors and Ergonomics Society}, 
  location =    {Nashville, Tennessee, USA}, 
  date =        {October 24--28}, 
  month =       {October}, 
  year =        1994, 
  publisher =   {HFES}, 
  pages =       {235--239}, 
  abstract =    { 
Previously, it had been found that texture-coding was ineffective at reducing 
search time (Perlman & Swan, 1993). In the experiment reported here, 16 subjects 
searched for blank-, color-, texture-, and density-coded targets of varying 
complexity in a naturalistic task. The data showed that all non-blank methods 
were significantly and about equally more effective at reducing search time than 
blank-coding (no coding). The difference of outcome with previous results is 
explained by task simplification and by the control of possibly confounding 
factors. The difference suggests that coding techniques using texture, and 
possibly other methods, should be evaluated in context. The similar performance 
of color-, texture-, and density-coding is explained by the use of 
equal-saturation and equal-brightness colors. Recommendations for the design of 
effective coding methods and for future research are discussed. 
}, 
}