Searching for overlapping objects in depth: Depth speeds search, but does not improve response accuracy

Searching for overlapping objects in depth: Depth speeds search, but does not improve response accuracy
复制标题

深度搜索重叠对象:深度加快搜索速度,但不会提高响应精度

DOI:
10.1167/14.10.1190
复制
发表时间:
2014
期刊:
影响因子:
1.8
通讯作者:
Godwin H
Godwin H
中科院分区:
医学4区
文献类型:
--
作者:
Godwin H

文献摘要

相似文献

使用视觉搜索任务,我们研究了重叠和三维深度对搜索性能的综合影响。参与者通过包含不透明多边形目标的显示器进行搜索,同时记录他们的眼球运动。一半的参与者搜索显示器与对象呈现在不同的深度平面彼此(多平面条件)和另一半的参与者搜索显示器与对象呈现在一个单一的深度平面(单平面条件)。随着对象之间重叠程度的增加,RT和错误率也增加。当两个以上的对象可以重叠时,RT和错误率特别受到影响。在图像中增加立体深度使参与者能够更快地做出反应,尽管在重叠程度较高时并不更准确。眼动分析显示了相似的重叠模式,随着重叠水平的增加,固定持续时间增加,固定次数增加,总固定时间增加。立体深度减少了固定包含许多重叠对象的区域所花费的总时间,这表明深度有助于对象分割和识别过程。这些结果对当前的视觉搜索模型都有影响,这些模型尚未探索重叠和深度在调节搜索行为中的作用,以及包含重叠对象的真实搜索任务,如机场X射线筛查和放射照相。
Using a visual search task, we examined the combined effects of overlap and three-dimensional depth on search performance. Participants searched through displays containing opaque polygon targets while their eye-movements were recorded. Half of the participants searched displays with objects presented at different depth planes to one another (the multi-plane condition) and the other half of the participants searched displays with objects presented at a single depth plane (the single-plane condition). As the degree of overlap between objects increased, both RTs and error rates also increased. RTs and error rates were particularly affected when more than two objects could overlap. Adding stereoscopic depth to the images enabled participants to respond more rapidly, though not more accurately, when the level of overlap was high. Eye movement analyses showed a similar pattern for overlap, with increasing fixation durations, increasing number of fixations, and increasing total fixation time as the level of overlap increased. Stereoscopic depth reduced the total time spent fixating regions containing many overlapping objects, suggesting that depth aided in object segmentation and recognition processes. These results have implications both for current models of visual search, which have not yet explored the role that overlap and depth together can play in modulating search behaviour, and for real-world search tasks that contain overlapping objects, such as airport X-ray screening and radiography.