Achievement: A multidisciplinary team of researchers from Oak Ridge National Laboratory (ORNL) developed a new online heatmap method, named hilomap, to visualize geospatial datasets as online map layers when low and high trends are equally important to map users. Existing online heatmap tools overlay data points by accumulating density of overlapping points on a map surface, which is ill-suited to highlight low extremes and, thus, unable to render heatmaps for change data, such as population change that researchers need to see both positive and negative changes. The hilomap algorithm conducts three virtual passes of HTML5 canvas rendering and creates a map layer with both low and high trend highlights leveraging computer graphics techniques within web browsers. It performs as fast as existing tools. The 4th ACM SIGSPATIAL International Workshop on Spatial Gems (Spatial GEMS 2022) awarded the best paper of the workshop to the research paper of this work.
Significance and Impact: This work addressed a long-standing technical challenge in existing online heatmap libraries to visualize data with both low and high weights in a single map layer. The new algorithm eliminated technical bottlenecks of this challenge. It has been published as open source software. This work provides a powerful visualization method to enable researchers and general public to visualize geospatial change data or other datasets in which low and high density trends are equally important, such as change in temperature, population, and voltage on power grid. The research and engineering excellence of this work has been recognized by the geographic information science community with the best paper award at Spatial GEMS 2022.
Research Details
- The hilomap algorithm accumulates low density and high density separately on two virtual HTML5 canvases such that cumulation of low values lead to higher absolute values, similar to high value cumulation.
- The final heatmap rendering completes the weight-opacity-alpha channel translation and interpolate cumulative density in both low and high canvases to a correct color in the actual map surface using a global color gradient.
- Performance study shows that the algorithm was slightly slower than one-pass algorithms, but introduced marginal delays in online heatmap rendering of large geospatial datasets.
- Hilomap software has been published as a plugin to a mainstream online geospatial library - OpenLayers.
Sponsor/Funding: LDRD Strategic Hire
PI and affiliation: Yan Liu, Computational Urban Sciences Group, Computational Sciences and Engineering Division, ORNL
Team: Melissa Allen-Dumas, Computational Urban Sciences Group, Computational Sciences and Engineering Division, ORNL
Citation and DOI: Liu, Yan Y. and Melissa Dumas. 2022. "Online Heatmap Generation with Both High and Low Weights." In: the 4th ACM SIGSPATIAL International Workshop on Spatial Gems (SpatialGems 2022). November 01, 2022. Seattle, Washington, USA. RESPub ID: 183039. DOI: https://doi.org/10.48550/arXiv.2212.07820
Summary: Heatmap is a common geovisualization method that interpolates and visualizes a set of point observations on a map surface. Most of online web mapping libraries implement a one-pass heatmap algorithm using HTML5 canvas or WebGL for efficient heatmap generation. However, such implementation applies additive operations that accumulate the rendering of point weights on the map surface grid, making it inappropriate for visualizations that require the highlighting of both low and high weights. We developed hilomap, an online heatmap algorithm that highlights surface areas where points with both low and high trends are located. An HTML5 Canvas-based reference implementation on OpenLayers is presented and evaluated to demonstrate its visualization capability and computational performance.
