RLND

Multivariate density maps

In this research we focused on visualizing large scale, multivariate, moving object data, in particular maritime data. These moving objects generally have a large amount of attributes besides time and position, i.e., size, type, velocity, etc. The challenge is how to convey these additional attributes to a user. We researched a method to interactively explore multiple attributes in trajectory data using density maps, i.e., images that show an aggregate overview of massive amounts of data.

Individual trajectories are convolved by moving a smoothing kernel over the trajectories with the speed of the vessel. The smoothed trajectories are then aggregated into a 'density field', which is then visualized using a combination of color mapping and illumination. The user can compute density fields subsets of the data based on any of its attributes using different parameters such as the size of the smoothing kernel and the weight applied to the resulting density. Using a widget, called a distribution map, the user can interactively define subsets in an effective and intuitive way. These density fields can be combined into a single density map where a multi hue color map is applied or a single hue color map is applied to each individual density field and the resulting images are blended. We use four different blending methods:

Opacity: Each density field is assigned a weight and the color mapped images are blended.
Maximum: For each pixel the color associated with the density field that has the maximum value is chosen.
Block: The screen is divided into a number of blocks, for each block one of the colored images is chosen at random, proportional to the density of the fields.
piechart: A number of piecharts is rendered on the screen to convey the proportion between the underlying density fields for a number of select locations.

One of the things we can do with the above method is visual, spatial anomaly detection. We take a large historic set of trajectories as reference and a live data set and aggregate these. A spatial anomaly is then defined as a 'live' vessel moving in an area where there is little or no traffic in the historic set. We use a density field of the historic data only for the illumination to serve as context and show anomalies with a green to red color map as shown in the image above.

Related publications:

Interactive Visualization of Multivariate Trajectory Data with Density Maps. Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Proceedings of PacificVis 2011, p. 147-154, 2011.

Density Based, Visual Anomaly Detection. Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Proceedings of the international workshop on Maritime Anomaly Detection (MAD) 2011, p. 11-12, 2011.

Interactive Density Maps for Moving Objects. Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Computer Graphics and Applications, IEEE 32, 1 (jan.-feb. 2012), 56-66.

Composite density maps

As an extension to the multivariate density maps shown above, we have created a visualization scheme in which the user can build a network of parameterizable 'blocks' that can each perform a class of functions. This allows an analyst to encode their domain knowledge into a network of these blocks that can then be used by an operator in an operational context. We have six types of blocks:

Convolution: Computes a density field of a subset of the trajectory data where both the convolution function and the filter functions can be parameterized by an arbitrary function on the trajectory attributes.
Counter: Counts the number of unique vessels per area where both the counting expressing and the trajectory filter function can be parameterized by an arbitrary function on the trajectory attributes.
Composition: An operator that can combine multiple density fields using some arbitrary function.
Enhancement: A set of parameterizable, image-based filters such as: Gaussian blur, edge detection and discretization.
Enrichment: A function that adds an attribute to the trajectory data that can be parameterized by an arbitrary function on the attributes.
Iteration: A meta-function that peforms a bounded number of iterations of a block network.

An analyst can build networks for a large variety of use cases such as the extraction of busy shipping lanes from a large set of trajectories (see above: a), making clearer images of areas with slow moving vessels by inversely scaling the kernel size by an underlying density map (see above: b) or by finding complex movement patterns such as areas with vessel movements that are moving against the 'main flow' (see above: c).

Related publications:

Composite Density Maps for Multivariate Trajectories. Roeland Scheepens, Niels Willems, Huub van de Wetering, Gennady Andrienko, Natalia Andrienko, and Jarke J. van Wijk. In Transactions on Visualization and Computer Graphics (Proceedings of InfoVis 2011), p. 147-154, 2011.

Close [X]

Abstract

Composite Density Maps for Multivariate Trajectories

We consider moving objects as multivariate time-series. By visually analyzing the attributes, patterns may appear that explain why certain movements have occurred. Density maps as proposed by Scheepens et al. are a way to reveal these patterns by means of aggregations of filtered subsets of trajectories. Since filtering is often not sufficient for analysts to express their domain knowledge, we propose to use expressions instead. We present a flexible architecture for density maps to enable custom, versatile exploration using multiple density fields. The flexibility comes from a script, depicted in this paper as a block diagram, which defines an advanced computation of a density field. We define six different types of blocks to create, compose, and enhance trajectories or density fields. Blocks are customized by means of expressions that allow the analyst to model domain knowledge. The versatility of our architecture is demonstrated with several maritime use cases developed with domain experts. Our approach is expected to be useful for the analysis of objects in other domains.

Roeland Scheepens, Niels Willems, Huub van de Wetering, Gennady Andrienko, Natalia Andrienko, Jarke J. van Wijk. In Transactions on Visualization and Computer Graphics (Proceedings of InfoVis 2011), vol. xx, no. xx, p. xxx-xxx, 2011.

Close [X]

Abstract

Interactive Visualization of Multivariate Trajectory Data with Density Maps

We present a method to interactively explore multiple attributes in trajectory data using density maps, i.e., images that show an aggregate overview of massive amounts of data. So far, density maps have mainly been used to visualize single attributes. Density maps are created in a two-way procedure; first smoothed trajectories are aggregated in a density field, and then the density field is visualized. In our approach, the user can explore attributes along trajectories by calculating a density field for multiple subsets of the data. These density fields are then either combined into a new density field or first visualized and then combined. Using a widget, called a distribution map, the user can interactively define subsets in an effective and intuitive way, and, supported by high-end graphics hardware the user gets fast feedback for these computationally expensive density field calculations. We show the versatility of our method with use cases in the maritime domain: to distinguish between periods in the temporal aggregation, to find anomalously behaving vessels, to solve ambiguities in density maps via drill down in the data, and for risk assessments. Given the generic framework and the lack of domain-specific assumptions, we expect our concept to be applicable for trajectories in other domains as well.

Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Proceedings of PacificVis 2011, p. 147-154, 2011.

Close [X]

Abstract

GPU-Based track visualization of multivariate moving object data

Moving object data visualization and analysis is a growing field of research. In this thesis we present a continuation of previous work where paths of moving vessels are convolved with a kernel onto a den-sity map to aid operators of coastal surveillance systems and analysts. The density map is visualized as a shaded height field and reveals hotspots of vessel activity.

Our improvements are twofold: First, we increase the computation speed of the density map by roughly a factor using modern graphics hardware and second, we introduce several visualization methods to reveal spatial and temporal distributions as well as relations between subsets of the data that are otherwise not visible in the traditional density map.

We present a geographic information system that allows an operator to analyze data sets containing moving objects. While the main focus in this thesis is data sets of ocean faring vessels, we show that our method is applicable to other types of moving object data sets as well.

Roeland Scheepens' MSc. thesis. Eindhoven University of Technology. 2010.

Papers

Composite Density Maps for Multivariate Trajectories. Roeland Scheepens, Niels Willems, Huub van de Wetering, Gennady Andrienko, Natalia Andrienko, and Jarke J. van Wijk. In Transactions on Visualization and Computer Graphics (Proceedings of InfoVis 2011), vol. xx, no. xx, p. xxx-xxx, 2011. Abstract [+]
Interactive Visualization of Multivariate Trajectory Data with Density Maps. Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Proceedings of PacificVis 2011, p. 147-154, 2011. Abstract [+]
Interactive Density Maps for Moving Objects. Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Computer Graphics and Applications, IEEE 32, 1 (jan.-feb. 2012), 56-66. Abstract [+]

Abstracts

Density Based, Visual Anomaly Detection. Roeland Scheepens, Niels Willems, Huub van de Wetering, and Jarke J. van Wijk. In Proceedings of the international workshop on Maritime Anomaly Detection (MAD) 2011, p. 11-12, 2011. Abstract [+]

Theses

GPU-Based track visualization of multivariate moving object data. MSc. Thesis Roeland Scheepens. Eindhoven University of Technology. 2010. Abstract [+]