Perception and painting: a search for effective, engaging visualizations

C.G. Healey, J.T. Enns
2002 IEEE Computer Graphics and Applications  
S cientific visualization represents information as images that let us explore, discover, analyze, and validate large collections of data. Much research in this area is dedicated to designing effective visualizations that support specific analysis needs. Recently, though, we've considered visualizations from another angle. We've started asking, Are visualizations beautiful? Can we consider visualizations works of art? You might expect answers to these questions to vary widely depending on an
more » ... ividual's interpretation of what it means to be artistic. We believe that the issues of effectiveness and aesthetics may not be as independent as they seem initially. We can learn much from studying two related disciplines-human psychophysics and art theory and history. Human psychophysics teaches us how we see the world around us. Art history shows us how artistic masters capture our attention by designing works that evoke an emotional response. The common interest in visual attention provides an important bridge between these domains. We're using this bridge to produce effective and engaging visualizations, and we'd like to share some of the lessons we've learned along the way. Multidimensional visualization Through our lab work, we've studied various issues in scientific visualization for much of the last 10 years. A large part of our effort focused on multidimensional visualization, the need to visualize multiple layers of overlapping information simultaneously in a common display or image. We often divide this problem into two steps: I the design of a data-feature mapping M, a function that defines visual features (such as color, texture, or motion) to represent the data and I an analysis of a viewer's interpretation of the images M produces. An effective M generates visualizations that let viewers rapidly, accurately, and effortlessly explore their data. One promising technique we discovered is using results from human perception to predict the performance of a particular M. The low-level visual system identifies certain properties of what we see very quickly, often in only a few tenths of a second or less. Perhaps more importantly, this ability is display-size insensitive, so visual tasks are completed in a fixed length of time that's independent of the amount of information being displayed. Obviously, these findings are attractive in a multidimensional visualization context. We can combine different visual features to represent multiple data attributes and pack large numbers of multidimensional data elements into an image. A viewer then rapidly analyzes sequences of images in a movie-like fashion. Figure 1 shows two example visualizations of multidimensional weather data. We constructed the first image by taking traditional visualizations of each attribute, then compositing them. Hue represents temperature (yellow for hot, green for cold), luminance represents pressure (bright for high, dark for low), directed contours represent wind direction, and Doppler radar traces represent precipitation. We built the second image using simulated brush strokes that vary their perceptual color and texture properties to visualize the data. Here, color represents temperature (bright pink for hot, dark green for cold), density represents pressure (denser for lower pressure), stroke orientation represents wind direction, and size represents precipitation (larger strokes for more rainfall). Although viewers often gravitate toward the first image because of its familiarity, any attempt to perform real analysis tasks leads to a rapid appreciation of the careful selection of colors and textures in the second image. Our experiments showed that viewers prefer the second image for the vast majority of the tasks we tested. Using perceptual guidelines can dramatically increase the amount of information we can visualize. We can't take advantage of these strengths with an ad-hoc choice of M, however. Certain combinations of visual features actively mask information by interfering with our ability to see an image's important properties. A key goal, therefore, is to build guidelines on designing effective visualizations and to present these findings in a way that makes them accessible to other visualization researchers and practitioners. During the last year, we asked ourselves, How can we make our visualizations engaging or aesthetically pleasing? Although this issue has only recently received attention in the visualization community, we feel it's an important factor worthy of study. An image regarded as interesting or beautiful can encourage viewers to study it in detail. We might use stylistic techniques that capture and focus attention on certain areas of a painting to
doi:10.1109/38.988741 fatcat:wbywg2gohzbizl6v2tozjtjolq