Enhancing Line Density Plots with Outlier Control and Bin-based Illumination
\authororcid Yumeng Xue0000-0002-8195-517X, \authororcid Bin Chen0009-0005-5047-214X, \authororcid Patrick Paetzold0000-0002-1315-4602, \authororcid Yunhai Wang0000-0003-0059-6580, \authororcid Christophe Hurter0000-0003-4318-6717, and \authororcid Oliver Deussen0000-0001-5803-2185 · 2025
Read Enhancing Line Density Plots with Outlier Control and Bin-based Illumination by \authororcid Yumeng Xue0000-0002-8195-517X, \authororcid Bin Chen0009-0005-5047-214X, \authororcid Patrick Paetzold0000-0002-1315-4602, \authororcid Yunhai Wang0000-0003-0059-6580, \authororcid Christophe Hurter0000-0003-4318-6717, and \authororcid Oliver Deussen0000-0001-5803-2185 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-16017.txt.
Introduction
Yumeng Xue and five co-authors address a chart design problem in this visualization paper, released on arXiv in 2025. Density plots summarize huge sets of lines, such as paths or time series, but they blur the paths themselves. Rare odd lines vanish, and smooth trends become hard to follow.
The team adds a bin-based lighting model that separates structure from density. A new score ranks each line by how unusual it is, and lighting in the brightness channel then highlights either main trends or rare outliers, as the analyst chooses. Tests on real data showed details that simpler plots missed, with less color distortion than standard shading. The method stays interactive for up to 10,000 lines.
Catalog record
| CREATOR | \authororcid Yumeng Xue0000-0002-8195-517X, \authororcid Bin Chen0009-0005-5047-214X, \authororcid Patrick Paetzold0000-0002-1315-4602, \authororcid Yunhai Wang0000-0003-0059-6580, \authororcid Christophe Hurter0000-0003-4318-6717, and \authororcid Oliver Deussen0000-0001-5803-2185 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-16017
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Enhancing Line Density Plots with Outlier Control and Bin-based Illumination \authororcid Yumeng Xue0000-0002-8195-517X, \authororcid Bin Chen0009-0005-5047-214X, \authororcid Patrick Paetzold0000-0002-1315-4602, \authororcid Yunhai Wang0000-0003-0059-6580, \authororcid Christophe Hurter0000-0003-4318-6717, and \authororcid Oliver Deussen0000-0001-5803-2185 Abstract Density plots effectively summarize large numbers of points, which would otherwise lead to severe overplotting in, for example, a scatter plot. However, when applied to line-based datasets, such as trajectories or time series, density plots alone are insufficient, as they disrupt path continuity, obscuring smooth trends and rare anomalies. We propose a bin-based illumination model that decouples structure from density to enhance flow and reveal sparse outliers while preserving the original colormap. We introduce a bin-based outlierness metric to rank trajectories. Guided by this ranking, we construct a structural normal map and apply locally-adaptive lighting in the luminance channel to highlight chosen patterns—from dominant trends to atypical paths—with acceptable color distortion. Our interactive method enables analysts to prioritize main trends, focus on outliers, or strike a balance between the two. We demonstrate our method on several real-world datasets, showing it reveals details missed by simpler alternatives, achieves significantly lower CIEDE2000 color distortion than standard shading, and supports interactive updates for up to 10,000 lines. \onlineid 9953 \vgtccategory Research \vgtcpapertype algorithm/technique \authorfooter Y. Xue, B. Chen, P. Paetzold, and O. Deussen are with University of Konstanz, Germany. E-mail: {firstname.lastname}@uni-konstanz.de. Y. Wang is with Renmin University of China, China. E-mail: [email protected]. C. Hurter is with Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse, France, and with IPAL Singapore International Research Laboratory on Artificial Intelligence, Singapore 138632 E-mail: [email protected]. \teaser Comparison of line-density visualization methods on vessel trajectories. (a) Plain density plot highlights only high-density regions while obscuring continuity and sparse lines; (b) Direct Lambertian shading in RGB introduces color distortion and orientation bias; (c) Our bin-based illumination decouples normals from density, adapts light per bin, and preserves the colormap while enhancing both dense trends and sparse anomalies. (a) Plain density plot (b) Direct Lambertian shading in RGB (c) Our bin-based illumination method Introduction keywords Line plot, density plot, binning technique, illumination model Density plots are widely used to visualize large and complex datasets, particularly when traditional methods such as scatterplots suffer from overplotting and visual clutter. By aggregating data points into discrete bins and encoding their density through color gradients, density plots effectively reveal the global distribution and coarse structures of point-based data. However, this binning-based approach has notable limitations for line-based datasets (e.g., trajectories or time series), where continuity and local shape are essential: fixed bins can disrupt perceived flow, hide smooth trends, and obscure subtle structural differences between nearby trajectories. As illustrated in 1(a) , when visualizing vessel trajectories, only a few high-density patterns remain clearly visible, while sparse trajectories become barely visible, and the continuity of dense regions can only be inferred through Gestalt principles of continuity [ 8 ] rather than through explicit line details. Previous work has explored alternative enhancements to line density plots. For example, Xue et al. [ 41 ] proposed a color-based differentiation method to reduce ambiguity between different trends. Complementary to this, several approaches [ 32 , 33 , 40 ] have attempted to address continuity loss and low-density invisibility by applying shading models such as Phong shading [ 27 ] , typically reduced to Lambertian shading by omitting specular and ambient terms. However, these shading-based techniques suffer from three core limitations that impede their applicability to general line data. First, previous line shading methods derive normal maps from gradients of the density field. Because these normals are tightly coupled to the density encoding, they rarely reveal structural information beyond what the density map already conveys, as they highlight main trends. Such trends are visually dominant in 1(b) . Thus, fine structural details remain underrepresented. Second, fixed global light directions create an orientation bias: lines nearly parallel to the light vector are much less emphasized than those orthogonal to the light direction (detailed in Section 2.1 ). In 1(b) , horizontal line segments are emphasized, while vertical ones appear flat and indistinct. In addition, applying shading directly in RGB space produces chromatic shifts (color distortions) which interfere with the density values of the colormap, as seen in 1(b) . Motivated by these observations, we propose the bin-based illumination approach shown in 1(c) . It advances the state of the art in three directions. First, we define a bin-based outlierness measure that quantifies how much a trajectory’s tangent orientation deviates from those of its local neighbors; informally, parallel neighbors indicate high similarity and low outlierness, while crossing or shape-distinct neighbors indicate low similarity and high outlierness. This measure provides a basis for ranking trajectories and distinguishing between dominant trends and rare outliers. Second, guided by these scores, we construct a structural normal map that combines trajectory-level cues with density-based gradients, decoupling normals from density and revealing structural details that density alone cannot capture. Because trajectories are ranked by outlierness, users can interactively adjust which patterns are emphasized by shifting the focus toward trend-conforming trajectories, strongly deviating ones, or any balance in between. Third, we adapt the illumination itself: instead of a fixed light source, we determine a local light direction per bin based on the dominant trajectory orientation and, following Chen et al. [ 6 ] , apply shading primarily to the luminance channel of CIELAB to minimize color distortion. Together, these components yield shading that enhances continuity in dense regions, makes sparse deviations more perceptible, and preserves the interpretability of the density colormap. As illustrated in 1(c) , our approach provides both enhanced structural visibility and flexibility of emphasis. The key contributions of this work are: • A bin-based similarity metric designed for line data that ranks trajectories by their degree of outlierness, providing a foundation for smoothly adjustable emphasis on either high-density trends or low-density outliers • A combined structural-normal and dynamic-lighting model, together with a luminance-only shading strategy, that enhances continuity and detail while preserving colors used in the density map • An image-synthesis pipeline that enables users to balance trend visibility and anomaly perception By bridging the gap between the discrete representation of density plots and the continuous character of trajectory data, our method provides a density-based overview while making local shape deviations and sparse anomalies more perceptible. We validate our approach through a comprehensive evaluation comparing it against simpler alternatives, quantitative analyses of its color fidelity and runtime, and diverse case studies on real-world datasets. 1 Related Work In this section, we survey density-based visualizations and discuss their application to continuous line data. We describe density estimation methods, ranging from foundational approaches to advanced techniques, while addressing challenges such as visual clutter and interpretability issues of complex datasets. Additionally, we present our proposed illumination method. 1.1 Density-based Visualization Density-based visualization techniques play a critical role in simplifying large datasets by reducing clutter, uncovering trends, and revealing distributions. Line-density plots have applications in time series and trajectory visualization across various domains, including finance [ 23 , 44 ] and transportation [ 40 , 13 ] . In early examples, sequential data of time series was aggregated by density visualizations to highlight value distributions over time. Work by Carr et al. [ 5 ] introduced density overlays in scatterplot matrices to summarize