Show
Recommended textbook solutions
 Ways of the World: A Global History3rd EditionRobert W. Strayer 232 solutions Tonal Harmony, Workbook8th EditionByron Almen, Dorothy Payne, Stefan Kostka 1,387 solutions America's History for the AP Course8th EditionEric Hinderaker, James A. Henretta, Rebecca Edwards, Robert O. Self 470 solutions
The Language of Composition: Reading, Writing, Rhetoric2nd EditionLawrence Scanlon, Renee H. Shea, Robin Dissin Aufses 661 solutions Colin Ware, in
Information Visualization (Fourth Edition), 2021 The Gestalt principle of continuity states that we are more likely to construct visual entities
out of visual elements that are smooth and continuous, rather than ones that contain abrupt changes in direction. (See Fig. 6.6.). The principle of good continuity can be applied to the problem of drawing diagrams consisting of networks of nodes and the links between them. It should be easier to identify the sources and destinations of connecting lines if they are smooth and continuous. This point is illustrated in Fig. 6.7. A graph drawing method known as a
confluent diagram (Bach, Riche, Hurter, Marriott, & Dwyer, 2017) combines link bundling with the principle of continuity to make paths between nodes clear and unambiguous (Fig. 6.8 illustrates). Figure 6.6. The pattern on the left (a) is perceived as a smoothly curved line overlapping a rectangle (b) rather than as the more angular components shown in (c). Figure 6.7. In (a), smooth continuous contours are used to connect nodes in the diagram; in (b), lines with abrupt changes in direction are used. It is much easier to perceive connections with the smooth contours. Figure 6.8. Confluent diagram from Bach et al., 2017. Smooth continuity makes it clear that there is no direct path from S to c even though they are connected by short lines. In contrast there are clearly direct paths from d to e and d to c. From Bach et al., 2017.Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780128128756000062 Our Vision is Optimized to See StructureJeff Johnson, in Designing with the Mind in Mind (Second Edition), 2014 Gestalt Principle: ProximityThe Gestalt principle of Proximity is that the relative distance between objects in a display affects our perception of whether and how the objects are organized into subgroups. Objects that are near each other (relative to other objects) appear grouped, while those that are farther apart do not. In Figure 2.1A, the stars are closer together horizontally than they are vertically, so we see three rows of stars, while the stars in Figure 2.1B are closer together vertically than they are horizontally, so we see three columns. FIGURE 2.1. Proximity: items that are closer appear grouped as rows (A) and columns (B). The Proximity principle has obvious relevance to the layout of control panels or data forms in software, Web sites, and electronic appliances. Designers often separate groups of on-screen controls and data displays by enclosing them in group boxes or by placing separator lines between groups (see Fig. 2.2). FIGURE 2.2. In Outlook’s Distribution List Membership dialog box, list buttons are in a group box, separate from the control buttons. However, according to the Proximity principle, items on a display can be visually grouped simply by spacing them closer to each other than to other controls, without group boxes or visible borders (see Fig. 2.3). Many graphic design experts recommend this approach to reduce visual clutter and code size in a user interface (Mullet and Sano, 1994). FIGURE 2.3. In Mozilla Thunderbird’s Subscribe Folders dialog box, controls are grouped using the Proximity principle. Conversely, if controls are poorly spaced (e.g., if connected controls are too far apart) people will have trouble perceiving them as related, making the software harder to learn and remember. For example, the Discreet Software Installer displays six horizontal pairs of radio buttons, each representing a two-way choice, but their spacing, due to the Proximity principle, makes them appear to be two vertical sets of radio buttons, each representing a six-way choice, at least until users try them and learn how they operate (see Fig. 2.4). FIGURE 2.4. In Discreet’s Software Installer, poorly spaced radio buttons look grouped in vertical columns. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780124079144000026 Our Vision is Optimized to See StructureJeff Johnson, in Designing with the Mind in Mind (Third Edition), 2021 Gestalt Principle: ContinuitySeveral Gestalt principles describe our visual system’s tendency to resolve ambiguity or fill in missing data so we perceive whole objects. The first such principle, the principle of Continuity, states that when visual elements are aligned with each other, our visual perception is biased to perceive them as continuous forms rather than disconnected segments. For example, in Fig. 2.9A, we automatically see two crossing lines—one blue and one orange. We don’t see two separate orange segments and two separate blue ones, and we don’t see a blue-and-orange V on top of an upside-down orange-and-blue V. In Fig. 2.9B, due to the vertical alignment of the pieces and the fact that they are spaced to match the curvature of the visible pieces, we see a sea monster in water, not three pieces of one. If we misaligned the pieces or spaced the pieces further than the curvature suggests, the illusion of continuity would disappear. Figure 2.9. Continuity: Human vision is biased to see continuous forms, even adding missing data if necessary. A well-known example of the use of the continuity principle in graphic design is the IBM logo. It consists of disconnected blue patches, and yet it is not at all ambiguous. The blue rectangles are stacked vertically with horizontal space between the stacks, so we see three bold letters, perhaps viewed through something like venetian blinds (see Fig. 2.10). Figure 2.10. The IBM company logo uses the Continuity principle to form letters from disconnected patches. Slider controls are a user-interface example of the Continuity principle. We see a slider as depicting a single range controlled by a handle that appears somewhere on the slider, not as two separate ranges separated by the handle (see Fig. 2.11A). Even displaying different colors on each side of a slider’s handle doesn’t completely “break” our perception of a slider as one continuous object, although ComponentOne’s choice of strongly contrasting colors (gray vs. red) certainly strains that perception a bit (see Fig. 2.11B). Figure 2.11. Continuity: we see a slider as a single slot with a handle somewhere on it, not as two slots separated by a handle: (A) Mac OS and (B) ComponentOne. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780128182024000027 Our Vision is Optimized to See StructureJeff Johnson, in Designing with the Mind in Mind, 2010 Gestalt Principle: Figure/GroundThe next Gestalt principle that describes how our visual system structures the data it receives is Figure/Ground. This principle states that our mind separates the visual field into the figure (the “foreground”) and ground (the “background”). The foreground consists of those elements of a scene that are the object of our primary attention, and the background is everything else. The Figure/Ground principle also specifies that the visual system’s parsing of scenes into figure and ground is influenced by characteristics of the scene. For example, when a small object or color patch overlaps a larger one, we tend to perceive the smaller object as the figure and the larger object as the ground (see Fig. 2.16). Figure 2.16. Figure/Ground: When objects overlap, we see the smaller as figure on ground. However, our perception of figure vs. ground is not completely determined by scene characteristics. It also depends on the viewer’s focus of attention. Dutch artist M. C. Escher exploited this phenomenon to produce ambiguous images in which figure and ground switch roles as our attention shifts (see Fig. 2.17). Figure 2.17. M. C. Escher exploited figure/ground ambiguity in his art. In user interface and Web design, the Figure/Ground principle is often used to place an impression-inducing background “behind” the primary displayed content (see Fig. 2.18). The background can convey information—e.g., the user’s current location—or it can suggest a theme, brand, or mood for interpretation of the content. Figure 2.18. Figure/Ground is used at AndePhotos.com to display a thematic watermark “behind” content. Figure/Ground is also often used to pop up information over other content. Content that was formerly the figure—the focus of the users' attention—temporarily becomes the background for new information, which appears briefly as the new figure (see Fig. 2.19). This approach is usually better than temporarily replacing the old information with the new information, because it provides context that helps keep people oriented regarding their place in the interaction. Figure 2.19. Figure/Ground is used at GRACEUSA.org to pop up a photo “over” the page content. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780123750303000028 Visual Objects and Data ObjectsColin Ware, in Information Visualization (Fourth Edition), 2021 Static LinksWhen text is integrated into a static diagram, the Gestalt principles discussed in Chapter 6 apply, as Fig. 8.24 shows. Simple proximity is commonly used in labeling maps. A line drawn around the object and text creates a common region. A line or common region can also be used to associate groups of objects with a particular label. Arrows and speech balloons linking text and graphics also apply the principle of connectedness. Figure 8.24. Gestalt principles used to guide the linking of text and graphics: (a) Proximity. (b) Continuity/connectedness. (c) Common region. (d) Common region combined with connectedness. [G8.20] Use Gestalt principles of proximity, connectedness, and common region to associate written labels with graphical elements. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780128128756000086 Static and Moving PatternsColin Ware, in Information Visualization (Third Edition), 2013 Gestalt LawsThe first serious attempt to understand pattern perception was undertaken by a group of German psychologists who, in 1912, founded what is known as the Gestalt school of psychology. The group consisted principally of Max Westheimer, Kurt Koffka, and Wolfgang Kohler (see Koffka, 1935, for an original text). The word Gestalt simply means “pattern” in German. The work of the Gestalt psychologists is still valued today because they provided a clear description of many basic perceptual phenomena. They produced a set of Gestalt laws of pattern perception. These are robust rules that describe the way we see patterns in visual displays, and, although the neural mechanisms proposed by these researchers to explain the laws have not withstood the test of time, the laws themselves have proved to be of enduring value. The Gestalt laws easily translate into a set of design principles for information displays. Eight Gestalt laws are discussed here: proximity, similarity, connectedness, continuity, symmetry, closure, relative size, and common fate (the last concerns motion perception and appears later in the chapter). ProximitySpatial proximity is a powerful perceptual organizing principle and one of the most useful in design. Things that are close together are perceptually grouped together. Figure 6.2 shows two arrays of dots that illustrate the proximity principle. Only a small change in spacing causes us to change what is perceived from a set of rows, in Figure 6.2(a), to a set of columns, in Figure 6.2(b). In Figure 6.2(c), the existence of two groups is perceptually inescapable. Proximity is not the only factor in predicting perceived groups. In Figure 6.3, the dot labeled x is perceived to be part of cluster a rather than cluster b, even though it is as close to the other points in cluster b as they are to each other. Slocum (1983) called this the spatial concentration principle; we perceptually group regions of similar element density. The application of the proximity law in display design is straightforward. Figure 6.2. Spatial proximity is a powerful cue for perceptual organization. A matrix of dots is perceived as rows on the left (a) and columns on the right (b). In (c) we perceive two groups of dots because of proximity relationships. Figure 6.3. The principle of spatial concentration. The dot labeled x is perceived as part of cluster a rather than cluster b. [G6.1] Place symbols and glyphs representing related information close together. In addition to the perceptual organization benefit, there is also a perceptual efficiency to using proximity. Because we more readily pick up information close to the fovea, less time and effort will be spent in neural processing and eye movements if related information is spatially grouped. SimilarityThe shapes of individual pattern elements can also determine how they are grouped. Similar elements tend to be grouped together. In Figure 6.4(a, b) the similarity of the elements causes us to see rows more clearly. In terms of perception theory, the concept of similarity has been largely superseded. The channel theory and the concepts of integral and separable dimensions provide much more detailed analysis and better support for design decisions. Two different ways of visually separating row and column information are shown in Figure 6.4(c) and (d). In Figure 6.4(c), integral color and grayscale coding is used. In Figure 6.4(d), green is used to delineate rows and texture is used to delineate columns. Color and texture are separate channels, and the result is a pattern that can be more readily visually segmented either by rows or by columns. This technique can be useful if we are designing so that users can easily attend to either one pattern or the other. Figure 6.4. (a, b) According to the Gestalt psychologists, similarity between the elements in alternate rows causes the row percept to dominate. (c) Integral dimensions are used to delineate rows and columns. (d) When separable dimensions (color and texture) are used, it is easier to attend separately to either the rows or the columns. [G6.2] When designing a grid layout of a data set, consider coding rows and/or columns using low-level visual channel properties, such as color and texture. ConnectednessPalmer and Rock (1994) argued that connectedness is a fundamental Gestalt organizing principle that the Gestalt psychologists overlooked. The demonstrations in Figure 6.5 show that connectedness can be a more powerful grouping principle than proximity, color, size, or shape. Connecting different graphical objects by lines is a very powerful way of expressing that there is some relationship between them. Indeed, this is fundamental to the node–link diagram, one of the most common methods of representing relationships between concepts. Figure 6.5. Connectedness is a powerful grouping principle that is stronger than (a) proximity, (b) color, (c) size, or (d) shape. [G6.3] To show relationships between entities, consider linking graphical representations of data objects using lines or ribbons of color. ContinuityThe Gestalt principle of continuity states that we are more likely to construct visual entities out of visual elements that are smooth and continuous, rather than ones that contain abrupt changes in direction. (See Figure 6.6.) The principle of good continuity can be applied to the problem of drawing diagrams consisting of networks of nodes and the links between them. It should be easier to identify the sources and destinations of connecting lines if they are smooth and continuous. This point is illustrated in Figure 6.7. Figure 6.6. The pattern on the left (a) is perceived as a smoothly curved line overlapping a rectangle (b) rather than as the more angular components shown in (c). Figure 6.7. In (a), smooth continuous contours are used to connect nodes in the diagram; in (b), lines with abrupt changes in direction are used. It is much easier to perceive connections with the smooth contours. SymmetrySymmetry can provide a powerful organizing principle. The symmetrically arranged pairs of lines in Figure 6.8 are perceived more strongly as forming a visual whole than the pair of parallel lines. A possible application of symmetry is in tasks in which data analysts are looking for similarities between two different sets of time-series data. It may be easier to perceive similarities if these time series are arranged using vertical symmetry, as shown in Figure 6.9, rather than using the more conventional parallel plots. Figure 6.8. The pattern on the left consists of two identical parallel contours. In each of the other two patterns, one of the contours has been reflected about a vertical axis, producing bilateral symmetry. The result is a much stronger sense of a holistic figure. Figure 6.9. An application designed to allow users to recognize similar patterns in different time-series plots. The data represents a sequence of measurements made on deep ocean drilling cores. Two subsets of the extended sequences are shown on the right. To take advantage of symmetry the important patterns must be small. Research by Dakin and Herbert (1998) suggests that we are most sensitive to symmetrical patterns that are small, less than 1 degree in width and 2 degrees in height, and centered around the fovea. The display on the right in Figure 6.9 is far too large to be optimal from this point of view. We more readily perceive symmetries about vertical and horizontal axes, as shown in Figure 6.10(a, b); however, this bias can be altered with a frame of reference provided by a larger-scale pattern, as shown in Figure 6.10(c) and (d). See Beck et al. (2005). Figure 6.10. Because symmetries about vertical and horizontal axes are more readily perceived, (a) is seen as a square and (b) is seen as diamond. (c, d) A larger pattern can provide a frame of reference that defines the axes of symmetry; (c) is seen as a line of diamonds and (d) as a line of squares. [G6.4] Consider using symmetry to make pattern comparisons easier, but be sure that the patterns to be compared are small in terms of visual angle (<1 degree horizontally and <2 degrees vertically). Symmetrical relations should be arranged on horizontal or vertical axes unless some framing pattern is used. Closure and Common RegionA closed contour tends to be seen as an object. The Gestalt psychologists argued that there is a perceptual tendency to close contours that have gaps in them. This can help explain why we see Figure 6.11(a) as a complete circle and a rectangle rather than as a circle with a gap in it as in Figure 6.11(b). Figure 6.11. The Gestalt principle of closure holds that neural mechanisms operate to find perceptual solutions involving closed contours. In (a), we see a circle behind a rectangle, not a broken ring as in (b). Wherever a closed contour is seen, there is a very strong perceptual tendency to divide regions of space into “inside” or “outside” the contour. A region enclosed by a contour becomes a common region in the terminology of Palmer (1992), who showed common region to be a much stronger organizing principle than simple proximity. Closed contours are widely used to visualize set concepts in Venn–Euler diagrams. In an Euler diagram, we interpret the region inside a closed contour as defining a set of elements. Multiple closed contours are used to delineate the overlapping relationships among different sets. A Venn diagram is a more restricted form of Euler diagram containing all possible regions of overlap. The two most important perceptual factors in this kind of diagram are closure and continuity. A fairly complex structure of overlapping sets is illustrated in Figure 6.12, using an Euler diagram. This kind of diagram is almost always used in teaching introductory set theory, and this in itself is evidence for its effectiveness. Students easily understand the diagrams, and they can transfer this understanding to the more difficult formal notation (Stenning & Oberlander, 1994). Figure 6.12. An Euler diagram. This diagram tells us (among other things) that entities can simultaneously be members of sets A and C but not of A, B, and C. Also, anything that is a member of both B and C is also a member of D. These rather difficult concepts are clearly expressed and understood by means of closed contours. When the boundary of a contour-defined region becomes complex, what is inside or outside may become unclear. In such cases, using color, texture, or Cornsweet contours (discussed in Chapter 3) will be more effective (Figure 6.13). Although simple contours are generally used in Euler diagrams to show set membership, we can effectively define more complex sets of overlapping regions by using color and texture in addition to simple contours (Figure 6.14). Figure 6.15 shows an example from Collins et al. (2009) where both transparent color and contour are used to define extremely convoluted boundaries for three overlapping sets. Figure 6.13. When the shape of the region is complex, a simple contour (shown in the upper left) is inadequate. (a) It is not easy to see if the x is inside or outside of the enclosed region. Common region can be defined less ambiguously by means of (b) a Cornsweet (1970) edge, (c) texture, or (d) color. Figure 6.14. An Euler diagram enhanced using texture and color can convey a more complex set of relations than a conventional Euler diagram using only closed contours. Figure 6.15. Both contour- and color-defined regions have been added to make clear the distribution of hotels (orange), subway stations (brown), and medical clinics (purple). From Collins et al. (2009). Reproduced with permission.[G6.5] Consider putting related information inside a closed contour. A line is adequate for regions having a simple shape. Color or texture can be used to define regions that have more complex shapes. [G6.6] To define multiple overlapping regions, consider using a combination of line contour, color, texture, and Cornsweet contours. Both closure and closed contours are critical in segmenting the monitor screen in windows-based interfaces. The rectangular overlapping boxes provide a strong segmentation cue, dividing the display into different regions. In addition, rectangular frames provide frames of reference: The position of every object within the frame tends to be judged relative to the enclosing frame (see Figure 6.16). Figure 6.16. Closed rectangular contours strongly segment the visual field. They also provide reference frames. The positions and sizes of the enclosed shapes are, to some extent, interpreted with respect to the surrounding frame. Figure and GroundGestalt psychologists were also interested in what they called figure–ground effects. A figure is something objectlike that is perceived as being in the foreground. The ground is whatever lies behind the figure. In general, smaller components of a pattern tend to be perceived as objects. In Figure 6.17(a), a black propeller is seen on a white background, as opposed to the white areas being perceived as objects. Figure 6.17. (a) The black areas are smaller and therefore more likely to be perceived as an object. It is also easier to perceive patterns that are oriented horizontally and vertically as objects. (b) The green areas are seen as figures because of several Gestalt factors, including size and closed form. The area between the green shapes in (c) is generally not seen as a figure. The perception of figure as opposed to ground can be thought of as part of the fundamental perceptual act of identifying objects. All of the Gestalt laws contribute to creating a figure, along with other factors that the Gestalt psychologists did not consider, such as texture segmentation. Closed contour, symmetry, and the surrounding white area all contribute to the perception of the two shapes in Figure 6.17(b) as figures, as opposed to cut-out holes. But, by changing the surroundings, as shown in Figure 6.17(c), the irregular shape that was perceived as a gap in Figure 6.17(b) can be made to become the figure. [G6.7] Use a combination of closure, common region, and layout to ensure that data entities are represented by graphical patterns that will be perceived as figures, not ground. Figure 6.18 shows the classic Rubin's Vase figure, in which it is possible to perceive either two faces, nose to nose, or a green vase centered in the display. The fact that the two percepts tend to alternate illustrates how competing active processes are involved in constructing figures from the pattern; however, the two percepts are driven by very different mechanisms. The vase percept is supported mostly by symmetry and being a closed region. Conversely, the faces percept is mostly driven by prior knowledge, not gestalt factors. It is only because of the great importance of faces that they are so readily seen. The result is a competition between high-level and mid-level processes. Figure 6.18. Rubin's Vase. The cues for figure and ground are roughly equally balanced, resulting in a bistable percept of either two faces or a vase. More on ContoursWe now return to the topic of contours to discuss what recent research tells us about how they are processed in the brain. Contours are continuous, elongated boundaries between regions of a visual image, and the brain is exquisitely sensitive to their presence. A contour can be defined by a line, by a boundary between regions of different color, by stereoscopic depth, by motion patterns, or by the edge of a region of a particular texture. Contours can even be perceived where there are none. Figure 6.19 illustrates an illusory contour; a ghostly boundary of a blobby shape is seen even where none is physically present (see Kanizsa, 1976). Because the process that leads to the identification of contours is seen as fundamental to object perception, contour detection has received considerable attention from vision researchers, and contours of various types are critical to many aspects of visualization. Figure 6.19. Most people see a faint illusory contour surrounding a blobby shape at the center of this figure. A set of experiments by Field et al. (1993) proved to be a landmark in placing the Gestalt notion of continuity on a firmer scientific basis. In these experiments, subjects had to detect the presence of a continuous path in a field of 256 randomly oriented Gabor patches (see Chapter 5 for a discussion of Gabor functions). The setup is illustrated schematically in Figure 6.20. The results showed that subjects were very good at perceiving a smooth path through a sequence of patches. As one might expect, continuity between Gabor patches oriented in straight lines was the easiest to perceive. More interesting, even quite wiggly paths were readily seen if the Gabor elements were aligned as shown in Figure 6.20(a). The theory underlying contour perception is that there is mutual reinforcement between neurons that have receptive fields that are smoothly aligned; there is inhibition between neurons with nonaligned receptive fields. The result is a kind of winner-take-all effect. Stronger contours beat out weaker contours. Figure 6.20. An illustration of the experiments conducted by Field et al. (1993). If the elements are aligned as shown in (a) so that a smooth curve can be drawn through some of them, a curve is seen. If the elements are at right angles, no curve is seen (b). This effect is explained by mutual excitation of neurons (c). Higher order neurophysiological mechanisms of contour perception are not well understood. One result, however, is intriguing. Gray et al. (1989) found that cells with collinear receptive fields tend to fire in synchrony. Thus, we do not need to propose higher order feature detectors, responding to more and more complex curves, to understand the neural encoding of contour information. Instead, it may be that groups of cells firing in synchrony is the way that the brain holds related pattern elements in mind. Theorists have suggested a fast enabling link, a kind of rapid feedback system, to achieve the firing of cells in synchrony (Singer & Gray, 1995). The theory of synchronous firing binding contours is still controversial; however, there is agreement that some neural mechanism enhances the response of neurons that lie along a smoothly connected edge (Li, 1998; Grossberg & Williamson, 2001). Representing Vector Fields: Perceiving Orientation and DirectionThe basic problem of representing a vector can be broken down into three components: the representation of vector magnitude, the representation of orientation, and the representation of direction with respect to a particular orientation. Figure 6.21 illustrates this point. Some techniques display one or two components, but not all three; for example, wind speed (magnitude) can be shown as a scalar field by means of color coding. Figure 6.21. The components of a vector. There are direct applications of the Field et al. (1993) theory of contour perception in displaying vector field data. A common technique is to create a regular grid of oriented elements, such as the one shown in Figure 6.22(a). The theory suggests that head-to-tail alignment should make it easier to see the flow patterns (Ware, 2008). When the line segments are displaced so that smooth contours can be drawn between them, the flow pattern is much easier to see, as shown in Figure 6.22(c). Figure 6.22. The results of Field et al. (1993) suggest that vector fields should be easier to perceive if smooth contours can be drawn through elements representing the flow. (a) A gridded pattern will weakly stimulate neurons with oriented receptive fields but also cause the perception of false contours from the rows and columns. (b) Line segments in a jittered grid will not create false contours. (c) If contour segments are aligned, mutual reinforcement will occur. (d) The strongest response will occur with continuous contours. Instead of the commonly used grid of small arrows, one obvious and effective way of representing vector fields is through the use of continuous contours; a number of effective algorithms exist for this purpose. Figure 6.23 shows an example from Turk and Banks (1996). This effectively illustrates the orientation of the vector field, although it is ambiguous in the sense that for a given contour there can be two directions of flow. In addition, Figure 6.23 does not show magnitude. Figure 6.23. Streamlines can be an effective way to represent vector field or flow data. But here the direction is ambiguous and the magnitude is not shown. From Turk & Banks, 1996; with permission.Comparing 2D Flow Visualization TechniquesLaidlaw et al. (2001) carried out an experimental comparison of the six different flow visualization methods, illustrated in Figure 6.24: (a) arrows on a regular grid; (b) arrows on a jittered grid to reduce perceptual aliasing effects; (c) triangle icons, with icon size proportional to field strength and density inversely related to icon size (Kirby et al., 1999); (d) line integral convolution (Cabral & Leedom, 1993); (e) large-head arrows along a streamline using a regular grid (Turk & Banks, 1996); and (f) large-head arrows along streamlines using a constant spacing algorithm (Turk & Banks, 1996). Figure 6.24. Six different flow visualization techniques evaluated by Laidlaw et al. (2001). From Laidlaw et al. (2001). Reproduced with permission.In order to evaluate any visualization, it is necessary to specify a set of tasks. Laidlaw et al. (2001) had subjects identify critical points as one task. These are points in a vector or flow field where the vectors have zero magnitude. The results showed the arrow-based methods illustrated in Figure 6.24(a) and (b) to be the least effective for identifying the locations of these points. A second task involved perceiving advection trajectories. An advection trajectory is the path taken by a particle dropped in a flow. The streamline methods of Turk and Banks, shown in Figure 6.24(f), proved best for showing advection. The line integral convolution method, shown in Figure 6.24(d), was by far the worst for advection, probably because it does not unambiguously identify direction. It is also worth pointing out that three of the methods do not show vector magnitude at all; see Figure 6.24(d, e, f). Although the study done by Laidlaw et al. (2001) was the first serious comparative evaluation of the effectiveness of vector field visualization methods, it is by no means exhaustive. There are alternative visualizations, and those shown have many possible variations: longer and shorter line segments, color variations, and so on. In addition, the tasks studied by Laidlaw et al. did not include all of the important visualization tasks that are likely to be carried out with flow visualizations. Here is a more complete list: •Judging the speed, orientation, and direction at an arbitrary point •Identifying the location and nature of critical points •Judging an advection trajectory •Perceiving patterns of high and low speed (or magnitude) •Perceiving patterns of high and low vorticity (sometimes called curl) •Perceiving patterns of high and low turbulence Both the kinds and the scale of patterns that are important will vary from one application to another; small-scale detailed patterns, such as eddies, will be important to one researcher, whereas large-scale patterns will interest another. The problem of optimizing flow display may not be quite so complex and multifaceted as it would first seem. If we ignore the diverse algorithms and think of the problem in purely visual terms, then the various display methods illustrated in Figure 6.24 have many characteristics in common. They all consist principally of contours oriented in the flow direction, although these contours have different characteristics in terms of length, width, and shape. The shaft of an arrow is a short contour. The line integral convolution method illustrated in Figure 6.24(d) produces a very different-looking, blurry result; however, something similar could be computed using blurred contours. Contours that vary in shape and gray value along their lengths could be expressed with two or three parameters. The different degrees of randomness in the placement of contours could be parameterized; thus, we might consider the various 2D flow visualization methods as part of a family of related methods—different kinds of flow-oriented contours. Considered in this way, the display problem becomes one of optimizing the various parameters that control how vector magnitude, orientation, and direction are mapped to contour in the display. Showing DirectionIn order to show direction, something must be added to a contour to give it asymmetry along its path. A neural mechanism that can account for the perception of asymmetric endings of contours is called the end-stopped cell. Many V1 neurons respond strongly to a contour that ends in the receptive field of the cell, but only coming from one direction (Heider et al., 2000). The more asymmetry there is in the way contour segments terminate, the greater the asymmetry in neural response, so this can provide a mechanism for detection of flow direction (Ware, 2008). Figure 6.25 illustrates this concept. Figure 6.25. (a) An end-stopped cell (shown as a green blob) will not respond when a line passes through it. (b) It responds only when the line terminates in the cell from a particular direction. (c) This asymmetry of response will weakly differentiate the heads of arrows from their tails. (d) It will more strongly differentiate the ends of a broad line with a gradient along its length. The little bars represent neuron firing rates. Conventional arrowheads are one way of providing directional asymmetry, as in Figure 6.25(c), but the asymmetric signal is relatively weak. Arrowheads also produce visual clutter because the contours from which they are constructed are not tangential to the vector direction. An interesting way to resolve the flow direction ambiguity is provided in a 17th-century vector field map of North Atlantic wind patterns by Edmund Halley (discussed in Tufte, 1983). Halley's elegant pen strokes, illustrated in Figure 6.26, are shaped like long, narrow airfoils oriented to the flow, with the wind direction given by the blunt end. Halley also arranges his strokes along streamlines. These can produce a stronger asymmetric signal than an arrowhead. We verified experimentally that strokes like Halley's are unambiguously interpreted with regard to direction (Fowler & Ware, 1989). Figure 6.26. Drawing in a style based on the pen strokes used by Edmund Halley (1696), discussed in Tufte (1983), to represent the trade winds of the North Atlantic. Halley described the wind direction as being given by “the sharp end of each little stroak pointing out that part of the horizon, from whence the wind continually comes.” Fowler and Ware (1989) developed a new method for creating an unambiguous sense of vector field direction that involves varying the gray level along the length of a stroke. This is illustrated in Figure 6.27. If one end of the stroke is given the background gray level, the stroke direction is perceived to be in the direction of change away from the background gray level. In our experiments, the impression of direction produced by lightness change completely dominated that given by shape. This is what the end-stopped cell theory predicts—the greater the asymmetry between the two ends of each contour, the more clearly the direction will be seen. Unfortunately, the perception of orientation may be somewhat weakened. The problem is to get both a strong directional response and a strong orientation response. Figure 6.27. Vector direction can be unambiguously given by means of lightness change along the particle trace, relative to the background. This gives the greatest asymmetry between the different ends of each trace. We can distill the above discussion into two guidelines. [G6.8] For vector field visualizations, use contours tangential to streamlines to reveal the orientation component. [G6.9] To represent flow direction in a vector field visualization, use streamlets with heads that are more distinct than tails, based on luminance contrast. A streamlet is a glyph that is elongated along a streamline and which induces a strong response in neurons sensitive to orientations tangential to the flow. To reveal the magnitude component of a vector field, we can fall back on the basic principle of using something that produces a stronger neural signal to represent fast flow or a stronger field. Figure 6.28 gives an example that follows both guidelines G6.8 and G6.9, and in addition uses longer and wider graphical elements to show regions of stronger flow (Mitchell et al., 2009). Figure 6.28. The surface currents in the Gulf of Mexico from the AMSEAS model. Head-to-tail elements are used, with each element having a more distinct head than tail. Speed is given by width, length, and background color. [G6.10] For vector field visualizations, use more distinct graphical elements to show greater field strength or speed. They can be wider, longer, more contrasting, or faster moving.
Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780123814647000065 Köhler, Wolfgang (1887–1967)V. Sarris, in International Encyclopedia of the Social & Behavioral Sciences, 2001 1.3 Köhler's Professorship at the University of BerlinIn 1922 Köhler became full professor of philosophy and director of the Psychological Institute at the University of Berlin, where he worked together with other eminent scientists like Max Wertheimer and Kurt Lewin (1890–1947). Here the international fame of the Berlin School of Gestalt psychology grew until its peak during the early 1930s. It was during this time that the Gestalt principles of the organization of perception were published and extended to new research issues (Wertheimer 1923) like memory, cognition, motivation, and aesthetics. During these years Köhler also lectured abroad several times, in the USA at Clark University (1925–6) and at Harvard University (1934–5), in South America in Uruguay (1930), as well as Brazil and Argentina (1932). When Köhler encountered the harassment of university scholars by the Nazi government in 1933, he fought racial discrimination and struggled against the political authorities of those days; he may have been the only psychology professor in Germany who protested in public against the totalitarian system (Henle 1978). When in 1935 his assistants were fired, he left Germany and followed his colleagues into exile in the USA. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B0080430767002795 Vision, High-level Theory ofJ. Wagemans, in International Encyclopedia of the Social & Behavioral Sciences, 2001 2.1 Biederman (1987)Object recognition concerns the identification of an object as a specific entity (i.e., semantic recognition) or the ability to tell that one has seen the object before (i.e., episodic recognition). Interest in object recognition is at least partly caused by the development of a new theory of human object recognition by Biederman (1987). Building on Marr's recognition ideas, where perceptually derived 3-D volumetric descriptions are matched to those stored in memory, Biederman in addition incorporated Gestalt principles of perceptual grouping and some newer ideas from computer vision into his theory. Biederman proposed that objects are identified at their entry level on the basis of a structural description of their components and their spatial relations. All object parts are modeled by a limited number of simple primitives called ‘geons,’ a specific subset of generalized cones. A generalized cone is the volume swept out by a cross section moving along an axis. Geons are defined by the categorical attributes on the following dimensions: the axis (straight or curved), the shape of the cross-section (straight or curved, symmetric or asymmetric), and the size of the cross-section (constant, expanded, or expanded and contracted). One major claim in Biederman's Recognition-by-Components (RBC) theory is that geons themselves are viewpoint-invariant because the underlying dimensions can be distinguished on the basis of so-called ‘nonaccidental properties’ or NAPs (a notion developed in computer vision). These are special features in the image (such as collinearity, curvilinearity, cotermination, parallelism, and symmetry) that are reliable in the sense that they are most likely caused by similar characteristics in 3-D space (under the assumption of a general viewpoint). For example, rarely do nonparallel structures in 3-D space project to parallel structures in the image. Biederman argues that Gestalt principles of grouping support the extraction of these NAPs and that they are thus detected fast and reliably enough to support bottom-up identification of geons. The spatial relations between object components also have to be specified by a number of parameters and in later versions of the theory, a few coarse metric attributes of the geons themselves must be specified too. The combinatorial possibilities of even only two or three geons are then enormous. This allows Biederman to argue that the recovery of two or three geons is enough to recognize complex objects quickly even when they are occluded, rotated in depth, degraded, or lacking color or texture. This claim is supported empirically by experiments in which parts of objects (in fact, line drawings of objects) were deleted until only two or three parts were left. For most moderately complex objects consisting of six or nine parts (e.g., an airplane, an elephant) three parts were still enough to recognize the object. Other supporting evidence comes from recognition of contour-deleted line drawings (which is possible as long as the geons are recoverable), priming with contour-deleted pictures (which is situated at the level of the parts, not the image fragments and not the whole picture), and viewpoint-invariance of 3-D object recognition (under certain limited conditions). While Biederman and his associates were accumulating evidence for his RBC theory, others were accumulating evidence for strong effects of viewpoint changes on 3-D object recognition (see, e.g., Bülthoff et al. 1995). This has given rise to a lot of controversy between viewpoint-dependent and independent theories of object recognition, often centered around the relevance of specific types of objects used as experimental stimuli (e.g., line drawings vs. rendered 3-D objects, all sorts of novel objects such as paperclips, amoebas, and ‘greebles’) or specific types of experimental tasks (e.g., matching, naming, priming). Because the focus in this review is on theory instead of empirical results, this debate will not be reviewed here (but see also Binocular Space Perception Models). Instead, two promising new theoretical developments will be highlighted. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B0080430767006707 Engaging Learners Through Rational Design of Multisensory EffectsDebbie Denise Reese, ... Curtis R. Taylor, in Emotions, Technology, and Design, 2016 Facilitation Through Innate Perceptual MechanismsInnate principles of perceptual organization in the different sensory domains can scaffold learner transactions with multisensory metaphors to aptly discover, define, and apply targeted knowledge. Each sensory system has a separate, finite amount of “perceptual memory.” Multisensory characteristics such as attention and grouping can be engineered to produce an automatic perceptual relationship between appropriately sensory-represented cognitive concepts (multisensory metaphors). For example, people perceptually organize things that are closer together differently than things that are further apart. These perceptual organization elements are preattentive. Preattentive elements are separate from game analogs, which are representations of the concept. Preattentive elements would complement relational structure represented by the modality metaphor and facilitate a natural or automatic making of connections. For example, players naturally process relationships perceptually, based upon proximity. So spatial clusters of connected game elements representations would capitalize on natural perception. Perceptual organization is a method for controlling the flow of thought processes that places relatively low demands on cognitive load. This hypothesis is supported by previous research. For example, Wallace, West, Ware, and Dansereau (2010) used gestalt principles (color, shape, and proximity) to organize information within concept maps. Participants who processed information from these enhanced maps recalled statistically significant greater amounts of information than did those who learned from text narrative or from unenhanced maps. Within learning games, this gestalt organization is both spatial and temporal. Some features of an object are processed more quickly and are more likely to be used than others. They guide subsequent attention even if multiple features are used (Wolfe & Horowitz, 2004). These preattentive object features and grouping principles have been described for vision and haptics in the sensory perception literature (Lederman & Klatzky, 1997; Overliet, Krampe, & Wagemans, 2012; Wolfe & Horowitz, 2004; Wolfe et al., 2011) and are summarized in Table 6.1. Although preattentive features have not been identified in audition, basic parameters for sonification (the use of nonspeech sounds) and similar ones for tactile vibrations have been developed (see Table 6.2), along with the determination of grouping principles for those modalities (Brewster & Brown, 2004; Denham & Winkler, 2013; Schertenleib & Candey, 2013). The proposed research program should also consider and investigate cross-modality interactions. Cross-modality interactions have been found to direct overall spatial attention (Driver & Spence, 1998; Wesslein, Spence, & Frings, 2014). Table 6.1. Preattentive Vision and Haptics Dimensions (Features for Guiding Attention) and Grouping (Association) Principles
Table 6.2. Audition and Tactile Dimensions (Parameters) and Association (Grouping) Principles
Our approach to multisensory presentation is novel in that it considers highly salient preattentive object features and grouping principles in three sensory systems (vision, audition, and haptics/touch) for implicitly guiding attention and presenting metaphors effectively to facilitate learning. Although multisensory HCI interfaces are in existence, most do not consider haptics/touch as well as vision and audition, nor do these interfaces consider the use of “preattentive” object features, grouping principles have only been used for static visual design, and no previous system to our knowledge has considered perceptual aspects of attention and grouping processes to facilitate the flow of learning. Read full chapter URL: https://www.sciencedirect.com/science/article/pii/B9780128018729000065 Edge and line oriented contour detection: State of the artGiuseppe Papari, Nicolai Petkov, in Image and Vision Computing, 2011 8.2 Interdipendence of computational stepsAnother important limitation of most of the existing algorithms for contour defection, which leaves space for further research, is that the steps which they consist of are optimized separately, rather than considering the entire process as a whole (though there are some exceptions, such as some methods based on Markov random fields). An important example is given by the algorithms to extract local edge features, which are not designed in terms of subsequent steps such as computing a contour saliency from local responses, or grouping edges according to Gestalt principles. In particular, much research has been made to distinguish from long chains of collinear edges and randomly scattered oriented stimuli [136,137,291]. Biological evidence suggests that the human visual system deploys similar mechanisms [139,140]. However, the local edge direction extracted with most of the existing local techniques, shows high correlation even when the input image is just random noise.7 This implies that edges originated by both object contours and noise or texture tend to be in good continuation with each other, thus weakening the effectiveness of the aforementioned post-processing technique. As a conclusion, we think that a more elaborate optimization of contour detection schemes, in which the interdependence of the different sub-algorithms is taken into account, will result in a considerably higher effectiveness and efficiency. Read full article URL: https://www.sciencedirect.com/science/article/pii/S0262885610001253 Which Gestalt principle means grouping nearby figures together?Principle #3: proximity
The principle of proximity states that things that are close together appear to be more related than things that are spaced farther apart.
Which Gestalt groups things that are closest to each other?The Law of Proximity is the gestalt grouping law that states elements that are close together tend to be perceived as a unified group. This straightforward law states that items close to each other tend to be grouped together, whereas items further apart are less likely to be grouped together.
What does it mean to close a Gestalt?Closure (a term used in Gestalt psychology) is the illusion of seeing an incomplete stimulus as though it were whole. Thus, one unconsciously tends to complete (close) a triangle or a square that has a gap in one of its sides.
What are the four types of grouping according to Gestalt theory?The Gestalt principles of grouping include four types: similarity, proximity, continuity, and closure. Similarity refers to our tendency to group things together based upon how similar to each other they are.
|
