Romero Mario, Summet Jay, Stasko John, Abowd Gregory
Georgia Tech.
IEEE Trans Vis Comput Graph. 2008 Nov-Dec;14(6):1261-8. doi: 10.1109/TVCG.2008.185.
In the established procedural model of information visualization, the first operation is to transform raw data into data tables [1]. The transforms typically include abstractions that aggregate and segment relevant data and are usually defined by a human, user or programmer. The theme of this paper is that for video, data transforms should be supported by low level computer vision. High level reasoning still resides in the human analyst, while part of the low level perception is handled by the computer. To illustrate this approach, we present Viz-A-Vis, an overhead video capture and access system for activity analysis in natural settings over variable periods of time. Overhead video provides rich opportunities for long-term behavioral and occupancy analysis, but it poses considerable challenges. We present initial steps addressing two challenges. First, overhead video generates overwhelmingly large volumes of video impractical to analyze manually. Second, automatic video analysis remains an open problem for computer vision.
在既定的信息可视化程序模型中，首要操作是将原始数据转换为数据表[1]。这些转换通常包括对相关数据进行聚合和分割的抽象操作，且通常由人、用户或程序员定义。本文的主题是，对于视频而言，数据转换应由低级计算机视觉提供支持。高级推理仍由人类分析师负责，而部分低级感知则由计算机处理。为说明这种方法，我们展示了Viz-A-Vis，这是一个用于在自然环境中对可变时间段内的活动进行分析的头顶视频捕获和访问系统。头顶视频为长期行为和占用情况分析提供了丰富的机会，但也带来了相当大的挑战。我们介绍了应对两个挑战的初步措施。第一，头顶视频生成的视频量极其庞大，手动分析不切实际。第二，自动视频分析对于计算机视觉来说仍然是一个未解决的问题。
