SpatialRugs: Enhancing Spatial Awareness of Movement in Dense Pixel Visualizations

Juri F. Buchmüller, Udo Schlegel, Eren Cakmak, Daniel A. Keim, Evanthia Dimara

TL;DR

This paper introduces SpatialRugs, a visualization technique that enhances spatial awareness in dense movement data by mapping movement locations to colors within a pixel-based display. By combining this with a novel time-aware color smoothing method, SpatialRugs allows users to better understand the spatial context of collective movement patterns without sacrificing the compactness of existing methods like MotionRugs.

What Problem Does This Paper Address?

Visualizing large-scale movement data, especially collective behavior, presents a significant challenge. Existing techniques like MotionRugs offer compact, space-efficient summaries by linearizing movement over time and space, allowing for quick identification of trends in features like speed. However, this linearization sacrifices spatial awareness, making it difficult for users to pinpoint the actual locations of the movers. This is a critical limitation, as spatial context (e.g., proximity to food sources or points of interest) is often crucial for understanding the underlying reasons for movement behavior. SpatialRugs aims to bridge this gap by reintroducing spatial awareness into dense pixel visualizations without losing the efficiency of methods like MotionRugs.

Methods

The core of SpatialRugs involves transforming the 2D spatial positions of movers into a 2D color space. This color space is then mapped onto a 1D linearized representation (similar to MotionRugs). Specifically, the process involves:

  1. Color Space Transformation: The 2D color space is adapted to match the extent of the spatial dimensions of the mover dataset.
  2. Color Mapping: Each mover's 2D spatial position is translated to a corresponding color within this transformed color map.
  3. Juxtaposed Display: These colors are then displayed in a dense pixel visualization, typically alongside other feature-based visualizations (like MotionRugs encoding speed).

To address perceptual distortions that can arise from color transitions in dense visualizations, the authors introduce a time-aware color smoothing (TACS) method. This technique uses a pooling matrix to consider the distribution of colors (representing positions) in the current and subsequent time steps. It then orders these colors and applies the median color to correct outliers, effectively smoothing transitions and mitigating the perception of exaggerated spatial distances.

Datasets

The paper uses a "demo dataset of 151 fish moving in a tank for about 90 seconds" for demonstration and evaluation purposes. No new benchmarks are introduced.

Key Insights and Ideas

  • Bridging the Gap: SpatialRugs successfully combines the space-efficiency of MotionRugs with the spatial awareness often lost in such compact visualizations.
  • Color as Location: The novel idea of using color to directly represent spatial location within a dense pixel display is a powerful way to reintroduce context.
  • Perceptual Distortions are Real: The paper highlights how color transitions can create misleading perceptions of spatial relationships, necessitating smoothing techniques.
  • Time-Aware Smoothing is Effective: The proposed TACS method demonstrably reduces perceptual distortions and outperforms standard Gaussian blur in preserving visual structures while smoothing transitions.
  • Colormap Choice Matters: The selection of an appropriate colormap is critical for effective spatial representation, with specific colormaps (e.g., Ziegler et al., Guo et al.) being more suitable than others due to perceptual properties and distribution.

Key Results

  • Qualitative Improvement: Visual comparisons (Figure 1 and 4) show that SpatialRugs, especially with TACS, provides a clearer understanding of mover locations and collective spatial patterns compared to unmodified visualizations or standard smoothing. The difference image (Figure 4C) visually confirms that TACS targets transition areas effectively.
  • Quantitative Assessment: The table in Figure 4 presents quantitative results comparing TACS with Gaussian blur using RMSE, MSE, and SSIM metrics. TACS consistently achieved lower RMSE and MSE values and higher SSIM values across various parameter settings, indicating better preservation of the original image structure and a more accurate representation.
  • Colormap Suitability: The authors identify colormaps by Ziegler et al. and Guo et al. as promising candidates for SpatialRugs, fulfilling requirements for color distinguishability, even distribution, and comparability.

Limitations and Future Work

  • Spatial Distance Ambiguity: While SpatialRugs enhances spatial awareness, users might still face challenges in precisely relating a color back to its exact spatial coordinate, especially with dense displays. Individual differences in color perception can also affect interpretation.
  • Parameter Tuning: Guidelines for optimal parameterization of SpatialRugs and the TACS method need further exploration.
  • Future Directions:
    • Quantify viewer perception of the technique and colormap choices.
  • Investigate the perceptual implications of the color correction process more thoroughly.
  • Develop adaptive colormap approaches tailored to specific movement distributions and user tasks.
  • Explore supporting interactive color anchors for semantically important regions.
  • Extend the technique to other non-spatial 2D point distributions.

Summary updated

Abstract

Compact visual summaries of spatio-temporal movement data often strive to express accurate positions of movers. We present SpatialRugs, a technique to enhance the spatial awareness of movements in dense pixel visualizations. SpatialRugs apply 2D colormaps to visualize location mapped to a juxtaposed display. We explore the effect of various colormaps discussing perceptual limitations and introduce a custom color-smoothing method to mitigate distorted patterns of collective movement behavior.
BibTeX
@inproceedings{buchmuller2020spatialrugs,
  author = {Juri F. Buchmüller and Udo Schlegel and Eren Cakmak and Daniel A. Keim and Evanthia Dimara},
  title = {{SpatialRugs: Enhancing Spatial Awareness of Movement in Dense Pixel Visualizations}},
  booktitle = {EuroVA@Eurographics/EuroVis},
  year = {2020},
  doi = {10.2312/eurova.20201078},
}

Loading related content...

Figure