Suppr超能文献

检测视觉噪声的统计效率。

Statistical efficiency for the detection of visual noise.

作者信息

Kersten D

机构信息

Walter S. Hunter Laboratory of Psychology, Brown University, Providence, RI 02912.

出版信息

Vision Res. 1987;27(6):1029-40. doi: 10.1016/0042-6989(87)90016-2.

Abstract

Visual detection thresholds were measured for one-dimensional static noise patterns centered at 2 c/deg on a logarithmic scale, for several widths and spatial frequency bandwidths in the presence and absence of a one-dimensional dynamic visual noise mask. Human performance was compared with the ideal observer's performance for the same task. The ideal's contrast thresholds increase with the fourth root of bandwidth. For thresholds measured in the presence and absence of noise, the fourth-root-law for bandwidth fitted the human observers' data from 1 to 6 octaves for a space constant of 2 degrees and from 2 to 6 octaves for a space constant of 0.25 degree. Detection efficiencies were quite high (30-60%) for patterns of 1- and 6-octave bandwidths. These results can be interpreted either in terms of a channel with an adjustable spatial frequency bandwidth or as very efficient combination of information between spatial frequency channels.

摘要

在存在和不存在一维动态视觉噪声掩模的情况下,以对数尺度测量了以2 c/deg为中心的一维静态噪声模式的视觉检测阈值,测量了几种宽度和空间频率带宽。将人类的表现与同一任务下理想观察者的表现进行了比较。理想观察者的对比度阈值随带宽的四次方根增加。对于在有噪声和无噪声情况下测量的阈值,带宽的四次方定律对于空间常数为2度的情况,能拟合人类观察者从1到6倍频程的数据;对于空间常数为0.25度的情况,能拟合从2到6倍频程的数据。对于1倍频程和6倍频程带宽的模式,检测效率相当高(30 - 60%)。这些结果既可以根据具有可调节空间频率带宽的通道来解释,也可以解释为空间频率通道之间非常有效的信息组合。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验