interactive maps

THREE DIMENSIONAL INTERACTIVE PICTORIAL MAPS A Thesis by ASMA NAZ Submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE December 2004 Major Subject: Visualization Sciences THREE DIMENSIONAL INTERACTIVE PICTORIAL MAPS A Thesis by ASMA NAZ Submitted to Texas A&M University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Approved as to style and content by: Ergun Akleman (Chair of Committee) Carol LaFayette (Member) Michael C. Neuman (Member) Phillip Tabb (Head of Department) December 2004 Major Subject: Visualization Sciences iii ABSTRACT Three Dimensional Interactive Pictorial Maps. (December 2004) Asma Naz, B. Arch, Bangladesh University of Engineering & Technology Chair of Advisory Committee: Dr. Ergun Akleman The objective of my research is to revive and practice the art of traditional pictorial maps in 3D cartographic visualization. I have chosen to create both graphical and statistical pictorial maps which can be used for the purpose of tourism and data representation respectively. Some traditional hand-drawn and sculptural pictorial maps of famous artists have been picked out to start as a base for my work. The goal was to recreate or imitate the style, character and features of these traditional hand-drawn and sculptural maps with 3D computer graphics and to analyze how effectively 3D tools can be used to communicate map information. I also wanted to explore ways to make these maps interactive on the Web and have them accessible to a large number of viewers. The results show a number of interactive 3D pictorial maps of different countries and continents. These maps are initially built with Maya, a 3D modeling software, and converted into web pages using the Viewpoint Technology. For statistical maps, Mel scripts have been used in Maya to take input from the user and change the shape of models accordingly to represent data. These maps are interactive and navigable and are designed to be easily accessible on the Web. iv To my family v ACKNOWLEDGMENTS I would like to take a moment to express my sincere gratitude to my thesis committee chair, Dr. Ergun Akleman, for sharing his expertise, enthusiasm and encouragement to explore new creative ideas. Special thanks to my committee members, Carol LaFayette and Michael C. Neuman, for their support, instruction and encouragement throughout this process. I also thank my classmates, faculty, and staff of the Viz Lab, who have provided their continuous assistance. Thanks to Ozan Ozener for his valuable suggestions, Aditya Dafre for his endless support, and Mostaque Bhuiyan for his support and invaluable friendship. Above all I am grateful to my parents and my brothers for their unconditional love, understanding and guidance through the years. vi TABLE OF CONTENTS CHAPTER I INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . I.1. I.2. I.3. II Introduction to Pictorial Maps . . . . . . . . . . . . . . Goal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 1 1 2 3 4 4 5 5 5 6 6 7 7 8 10 10 11 11 11 11 13 13 13 14 15 15 16 16 17 18 19 20 BACKGROUND STUDY . . . . . . . . . . . . . . . . . . . . . . II.1. Pictorial Maps . . . . . . . . . . . . . . . . . . . II.2. Classification of Pictorial Maps . . . . . . . . . . II.2.1. Graphical Maps . . . . . . . . . . . . . . . II.2.2. Statistical Maps . . . . . . . . . . . . . . II.2.3. Metaphorical Maps . . . . . . . . . . . . . II.3. Pictorial Maps in Cartographic Visualization . . . II.3.1. Statistical Pictorial Maps in Visualization II.3.2. Graphical Pictorial Maps in Visualization II.3.3. Interactivity in Pictorial Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III OBJECTIVES AND PROPOSITIONS . . . . . . . . . . . . . . III.1. Objectives . . . . . . . . . . . . . III.2. Types of Maps and Inspirations III.2.1. Graphical Map . . . . . . III.2.2. Statistical Map . . . . . . III.3. Caricature . . . . . . . . . . . . III.4. Aesthetics . . . . . . . . . . . . III.5. Contribution of the Computer . III.6. Scale . . . . . . . . . . . . . . . III.7. Interactivity of 3D Pictorial Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . IV.1. Building the 3D Model . . . . . . . . . . . . . . . . . . IV.2. Modeling the Graphical Map . . . . . . . . . . . . . . . IV.2.1. Applying the Process of Caricature . . . . . . . . IV.2.2. Caricature of Land and Choosing Major Elements IV.2.3. Modeling Mountain Ranges . . . . . . . . . . . . IV.2.4. Creating Rivers, Road Networks and Vegetation . IV.2.5. Caricatures of Buildings . . . . . . . . . . . . . . vii CHAPTER IV.2.6. Text . . . . . . . . . . . . . . . . . . . . . . IV.3. Modeling the Statistical Map . . . . . . . . . . . . IV.3.1. Cognitive Reading of Statistical Maps . . . IV.3.2. Traditional Ways to Read a Statistical Map IV.3.3. Understanding Color Concepts . . . . . . . IV.3.4. Use of Graphs or Bars . . . . . . . . . . . . IV.3.5. Choosing Scale, Type and Method . . . . . IV.3.6. Choosing Color Schemes . . . . . . . . . . IV.3.7. The Process of Caricature . . . . . . . . . . IV.3.8. Using Mel Scripts . . . . . . . . . . . . . . IV.3.9. Create the Surrounding Landform and Text IV.4. Applying Textures and Shaders . . . . . . . . . . . IV.4.1. Process of Baking Shadows in Maya . . . . IV.5. Transferring Model to Viewpoint Media . . . . . . IV.5.1. Advantages of Viewpoint Technology . . . IV.5.2. Applications to Install . . . . . . . . . . . . IV.5.3. Steps to Export a Maya File to Viewpoint . IV.5.4. Discussing Export Options in Maya . . . . . IV.5.5. Assembling in Viewpoint Scene Builder . . . IV.5.6. Billboard Texts and Toon Shader . . . . . . IV.5.7. Publishing on the Web . . . . . . . . . . . . IV.5.8. View Web Page with Media Player . . . . . V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 22 24 24 25 25 26 27 27 28 29 31 32 32 33 34 35 36 36 39 39 40 40 41 41 41 46 46 47 48 50 52 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V.1. Interactive Graphical Maps . . . . . . . . . . . . . . . . V.2. Interactive Statistical Maps . . . . . . . . . . . . . . . . VI CONCLUSIONS AND FUTURE POSSIBILITIES . . . . . . . . VI.1. Evaluation of Size and Appearance . . . . . . . . . . . . VI.2. Limitations of Viewpoint . . . . . . . . . . . . . . . . . VI.3. Future Research Possibilities . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VITA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii LIST OF FIGURES FIGURE 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 An example of average revenue per kilowatthour for utilities. . . . . . Modeling landform of Italy from a base map. . . . . . . . . . . . . . Top view of Italy with important natural and built features. . . . . . Polygonal mesh of three valence objects. . . . . . . . . . . . . . . . . 3D model of mountain ranges in Italy. . . . . . . . . . . . . . . . . . Examples of reduction of number of openings for simplification. . . . Examples of simplification and exaggeration. . . . . . . . . . . . . . . 3D polygonal text in Maya. . . . . . . . . . . . . . . . . . . . . . . . Final 3D model of Italy with text. . . . . . . . . . . . . . . . . . . . Top view of 3D model of France. . . . . . . . . . . . . . . . . . . . . A bar graph shows comparison of data. . . . . . . . . . . . . . . . . . Simplified shapes of states of USA with changing heights. . . . . . . The slider interface in Maya for changing heights of models. . . . . . Melscript for creating the slider interface. . . . . . . . . . . . . . . . Polygonal text in Maya for statistical maps. . . . . . . . . . . . . . . Dome of light and hypershade option in Maya. . . . . . . . . . . . . 3D models with baked shadow. . . . . . . . . . . . . . . . . . . . . . Model of Italy with baked shadow. . . . . . . . . . . . . . . . . . . . The export option window in Maya. . . . . . . . . . . . . . . . . . . The camera setting option window. . . . . . . . . . . . . . . . . . . . Page 6 17 18 19 20 21 21 22 23 23 26 28 29 30 31 32 33 34 37 38 ix FIGURE 21 22 23 24 25 26 27 28 29 30 31 The materials setting option window. . . . . . . . . . . . . . . . . . . The shadow option window. . . . . . . . . . . . . . . . . . . . . . . . The compression setting option window. . . . . . . . . . . . . . . . . Billboard text and toon shader in Scene Builder. . . . . . . . . . . . Graphical pictorial map of Italy. . . . . . . . . . . . . . . . . . . . . Graphical pictorial map of the United Kingdom. . . . . . . . . . . . . Graphical pictorial map of France. . . . . . . . . . . . . . . . . . . . Graphical pictorial map of Europe. . . . . . . . . . . . . . . . . . . . Graphical pictorial map of Western Europe. . . . . . . . . . . . . . . Statistical pictorial map of the United States of America. . . . . . . . Statistical pictorial map of Europe. . . . . . . . . . . . . . . . . . . . Page 38 38 38 39 42 42 43 43 44 45 45 1 CHAPTER I INTRODUCTION Cartographic Visualization is a popular tool in producing traditional as well as interactive maps and cartography. Various methods are being investigated, developed, and evaluated for the effective portrayal of cartographic information. Visualization tools are used effectively to create virtual environments and using geographical information systems to create graphic maps. I.1. Introduction to Pictorial Maps Pictorial map, a genre of cartography, is a new concept and not yet tried and tested in the field of visualization. These are mental or thematic maps, where geography is distorted to reflect the reality in our minds rather than the world around us. Depending on the cartographer’s perception, some regions grow in importance, while others shrink or disappear. Pictorial maps use symbols or pictures to represent a theme. Elements normally used are landmarks, buildings, people, ethnography, plants, animals, modes of transportation, local agricultural and manufactured products, mythical objects, authors etc. including textual explanations of the illustrations. They also include historical events and references, usually characterized by humorous, playful or whimsical touches. The pictorial map does not usually give a very accurate presentation of the landscape. It is basically a caricature of traditional topographic or illustrated maps, and not pure reference forms. Geography is made interesting by combining pictures of things that one can relate to, such as, buildings, trees and The journal model is IEEE Transactions on Visualization and Computer Graphics. 2 people, with texts, symbols and abstract or even metaphorical representations of the land masses [1], [2], [3]. Pictorial maps have a long history and tradition. They have been made for centuries, contributing not only to teach geography, but also history and literature. Illustration artists, painters, sculptors, caricaturists specialized in pictorial maps, which were used for advertising agencies to promote tourism, for classroom walls commemorating historical milestones or some annual events. Popular for their decorative qualities, these maps have been expressions of civic or national pride. But due to lack of effort for collection, many have become scarce. Collecting pictorial maps is in its relatively early stages, even compared to 20th-century road maps. Academic institutions have started to take serious steps in archiving pictorial maps. Harvard University Library put together exhibits of maps such as the 2003 show at Harvard by graphic designer Ernest Dudley Chase [3]. Books are being published in an attempt to collect pictorial maps and study their history and nature. Online commercial websites, such as, the George Glazer Gallery have started putting together galleries of pictorial maps for collectors [3]. In 3D cartographic visualization, representation of pictorial maps, till now, has been limited to abstract geographic or data representation. The practice of hand drawn maps is becoming scarce with the advent of computer-generated maps inclined towards producing accurate traditional maps or virtual maps. I.2. Goal The goal of my research is to identify methods or ways to represent traditional pictorial maps with 3D computer graphics. My aim is to explore ways of representing graphical and statistical aspects of pictorial maps in 3D, imitating the style, char- 3 acter and features, of hand-drawn and sculptural pictorial maps. The focus is also to analyze how effectively 3D tools can be used to communicate map information in the most efficient way in the 3D environment. I also want explore ways of making these maps interactive and navigable on the Web and to have them accessible to a larger number of viewers. I have chosen Maya, a 3D modeling software, to create 3D models of the map, sometimes with the help of Mel scripts. Viewpoint Technology, a multimedia software, is to be used for converting these maps into web pages and make them interactive and navigable. I.3. Purpose The main purpose of my thesis is to revive, sustain and encourage the practice of pictorial maps as a popular form of map-making in the field of cartographic visualization, as well as making them interactive and navigable, and easily accessible to users through the Internet. Pictorial maps are attractively designed, informative pieces which combine the best elements of scientific cartography and decorative art. The aspect of beauty is a personal experience where one’s own imagination plays an important role. Through interesting perspectives, buildings and people, these maps vividly bring back to life the mental picture we have of a place. My aim is to recreate or capture the beauty of the vivid and fascinating portrait of the world, with beautiful landscape and wonderful structures, while transferring that into 3D. These 3D pictorial maps can be used as effective tourists maps, as a welcome tool, or to provide statistical or quantitative information through the use of 3D models, such as, population density, income groups etc. They can also be used to analyze or study the spatial organization of natural and built environment. 4 CHAPTER II BACKGROUND STUDY II.1. Pictorial Maps The heyday of pictorial maps was the 1920s through 1950s, with the resurgence in the 1970s. Pictorial mapmaking became a popular art form in the early 20th century. The style of the maps was mostly Art Deco, and reflected different sensibilities of the times. Pictorial Maps, till now, have been produced on wood, paper, cardboard, leather, ceramic, plastic, fabrics, clay and metal, and even some combinations thereof. The maps range in size from a shower curtain to miniature maps on stamps, and they come in all manner of shapes [4]. Artists frequently remained anonymous. Miguel Covarrubias, Coulton Waugh, Frank Dorn, Jo Mora, Ernest Dudley Chase and John Held were some of the prominent illustrators who developed recognizable cartographic styles [3]. The major practitioners combined artistic ability with historical or cultural research. Various studies have been done on the type of hand-drawn pictorial maps and efforts have been made to collect them. Universities such as Harvard, Dartmouth and Cornell made pictorial maps of their campuses [3]. Inexpensive color printing inspired companies and other institutions to use pictorial maps as promotional purposes, in the form of souvenirs. George and Helen Glazer made an elaborate discussion on the history and illustrators of Pictorial Maps in [3]. The Library of Congress also created online sites for zoom-able 2D Pictorial Maps in [5]. Graphic designer Nigel Holmes discussed different styles and modes of pictorial maps by different artists in [1]. Artist Jankovic created a virtual pictorial map gallery where he has a collection of traditional hand-drawn tourist maps [2]. 5 II.2. Classification of Pictorial Maps All pictorials maps are thematic. There are mainly three types of pictorial maps. II.2.1. Graphical Maps Graphical maps are sometimes used as tourist maps with visitor attractions, or used to commemorate some historical events and cultural diversity of a country. Familiar metaphors or icons are taken from users’ everyday life to create these maps to make them instantly recognizable and comprehensible. These maps provide a good method to visualize, understand and navigate a large and complex world. II.2.2. Statistical Maps These maps are used for data representation. Statistical maps can display the distribution of a variable over a geographic area, usually defined by political boundaries, by effective use of hue, saturation and brightness (or intensity) of color, symbol type and size; texture (e.g., hatch patterns) and legend to communicate the data. An ordered scale is normally used for quantitative maps where, clearly identifiable shades or patterns represent each group or range of data, besides legends that are categorized in boxes. In a statistical map, each unit (such as, average population, or average rainfall per square kilometer for each county or state) can be illustrated by a color or shading pattern that represents the value (or range of values) for that unit of the variable (see Figure 1). Usually a visual gradation in the shading or color is used, where the densest shading or darkest color (or hue) normally indicates the largest magnitude and the least dense shading or lightest color (or hue) indicates the smallest magnitude. Decreasing (or increasing) darkness or density shows the remaining range of 6 categories. The legend is arranged in an order of values, either from high to low or vice versa, which corresponds to the order of the range of values of the color or shading. Fig. 1. An example of average revenue per kilowatthour for utilities. II.2.3. Metaphorical Maps These maps can be based on visual metaphors of common or serious issues [1]. II.3. Pictorial Maps in Cartographic Visualization Representation of pictorial maps in computer visualization is a relatively new concept. Attempts have been made to use classification, simplification, exaggeration and symbolization in various ways, which are characteristic of pictorial maps. Most of these pictorial cartograms are two-dimensional and have been used to convey quantitative information. The practice of two-dimensional pictorial maps is more popular 7 and wide-spread than that in three-dimension. II.3.1. Statistical Pictorial Maps in Visualization In cartographic visualization, abstract and simplified pictorial maps have been used for data representation, which are not true to scale. Graphical elements that designers normally use for creating these statistical maps are abstract shapes, hue, saturation and brightness (or intensity) of color, symbol type and size, and texture (e.g., hatch patterns). Dorling and Demers (1994) studied codes for the spatial distribution of statistical measurements and created abstract pictorial maps [6], [7]. Beaverstock, Smith and Taylor (1999) created Schematic map of cluster of cities with bullets, boxes, regions and arrows, which represented sizes of cities for Globalization and World Cities (GaWC) Study Group in [6]. Bortins, Demers and Clarke (2002) studied noncontiguous cartograms in visualization in [6]. House and Kocmoud (1998) did a study on continuous area cartogram, representing population distribution with pictorial maps in [8]. II.3.2. Graphical Pictorial Maps in Visualization Virtual Reality Modeling, a popular tool, does not, necessarily, represent reality authentically, but aims for clear and believable representations. Interactive VRML Cartography is used to display abstractness and symbolization to show terrain modeling and data representation. Dykes, Moore and Fairbairn (1999) discussed virtual reality (VRML) and its use in cartography in [9]. The concept of graphical pictorial maps in 3D has been explored by Haeberling (2002), who used the concept of perspective pictorial map and symbolization, together with block diagrams, oblique regional views or schematic maps in [10]. He combined 8 topographic information with legend and tried to revive the tradition of using symbols to design effective and informative Topographic 3D maps. As yet no such cartographic traditions or principles have been evolved. II.3.3. Interactivity in Pictorial Maps Online interactive pictorial maps are mostly used for tourism [11]. Online interactive tourist maps are used to help people find their way in the form of street maps, or find a location of major cities. They can also be used as campus directories, to show transit routes, or find exotic travel locations. Symbols of modes of transportation, city icons and landmarks are used with textual illustrations and legends. These high-tech twodimensional maps are navigable. That means they have a graphical-user interface and respond to user activity such as clicking or scrolling with the mouse of a computer. They also function as an interface to other data. This gives more perspective and context than static maps. Scaling and zooming options work well for printing and can alleviate the problem of static maps that are too small in format. Geographic locations on the map can be further linked to virtual tours, movies, photographs, sound or other imagery. Tourism is an information-intensive and information-sensitive industry. With the tremendous growth of the Web today, WebGIS, electronic commerce, and Web technologies play a significant role in a tourists decision-making and satisfaction, which greatly depends on the way the information is presented and the tourists approach to search for information. Presenting comprehensive description of the tourist destination and presenting tourism products pose new challenges, such as, the geographic searching for tourist objects, dynamically generating scalable and interoperable tourist maps and to offer the users fast, flexible and meaningful access to the data. 9 Static maps are developed in a vector-based graphics application and interactivity is added using Flash, PDF, Java or JavaScript, VRML and Quicktime VR. Interactive maps are hosted on both UNIX and Windows NT servers. 10 CHAPTER III OBJECTIVES AND PROPOSITIONS III.1. Objectives The goal of my research is to identify methods to represent pictorial maps with the use of 3D Computer Graphics and to reproduce or imitate the style and character of traditional hand-drawn or sculptural pictorial maps when they are translated into 3D. The aim is to find out how graphical and statistical maps can be expressed most effectively with the 3D tools, and the features and attributes of the maps can be communicated to the reader in the most efficient manner. My aim is to recreate or capture the beauty of these informative, and attractively designed portraits of the world, while transferring that into 3D. These pictorial maps will be aesthetically pleasing, and contribute to recreate a mental picture we have of a place. The style of the hand-drawn and sculptural pictorial maps will be imitated as closely as possible, when translated into 3D media. Furthermore, these maps will be interactive, scalable, and navigable with 3D features, texts, and symbols. The steps to achieving my goal are as follows: identifying types of pictorial maps to use choosing features that will visually give the place its character for the graphical maps choosing features that will provide important data or statistical information studying the process of simplification or exaggeration to make it aesthetically pleasing and emphasize the character of the place, and at the same time allow the map to remain smaller in size and easily loadable for viewing finding an efficient way to make scalable and interoperable maps and offer the user fast and meaningful access to the data. 11 III.2. Types of Maps and Inspirations The first step to create a pictorial map is to choose a theme. The graphical maps will be used for tourist attraction, and the statistical maps for data representation. III.2.1. Graphical Map The inspiration for my 3D graphical pictorial maps is the hand-drawn pictorial map of Italy by Artist Saso Aleksander Jankovic [2]. The purpose of these maps is purely to be a welcome tool for tourists and use for planning excursions of a country. The features show the visitor attractions, landmarks, major rivers, road network, mode of transportation, trees and mountain ranges associated with textual illustrations. I will create several examples of tourist maps of different countries or continents. III.2.2. Statistical Map The clay map of New York by Anna Walker [1] is my inspiration to create statistical maps for countries or continents. These maps will be used for data representation. I created the map of the United States of America and Europe. I displayed and compared the population density, and political stature of the states or countries by changing the color or heights of the 3D models of different states of the country or different countries in a continent. III.3. Caricature Caricature is the art of simplification, exaggeration and symbolization. A caricature of a map is an artwork that exaggerates the distinctive features of a country or a place, and is exaggerated for comic effect. The country is portrayed very broadly 12 and in a stereotypical fashion. It is aesthetically pleasing and easily recognizable. Pictorial map is an art of simplification and selection. It is important to find out how much to put in to convey a sense of place, or how much to leave out for the sake of clarity [1]. Highly recognizable individual objects are chosen for simplification, such as, buildings or natural wonders. These are given more emphasis than the landscape and often slightly distorted or exaggerated for that reason. Terrain is not shown true to the size. Important mountain ranges, peaks and landmarks are exaggerated so that it is appealing and easily identifiable to the average man. The boundary of the land is simplified with soft curves and straight lines, showing some major rivers or lakes. Symbols bring life to the map. They are readable, and usually used when there are too many activities or information crammed in a place. Size, shape and simplicity are the three properties of symbols. For the graphical map, there are generic symbols such as, trees, boats or icons for city locations. Clusters of trees will show high vegetation areas. Landmarks can become symbol themselves. There is no generic tower for the Eiffel tower, the Tower of Pisa, or a generic Coliseum in Rome. The characteristic natural poses of the symbols will show the essence of what they represent. The art of caricature is not common among building natural environments or buildings. Current visualization practices do not provide a satisfactory answer to dealing with detailed 3d models. Building details need to be simplified by reducing the complexity of models, and making them easily recognizable at the same time. The speed of computer visualization greatly depend on this. The models need to be simple and easy to upload on the web, so that a wide range of users can view or access them online with ease. 13 III.4. Aesthetics It needs careful consideration as to whether the map is portraying accurate information about the site, retaining the character of the site, not distorting any facts or data, but still is aesthetically interesting and challenging. Distortion, simplification and exaggeration are based on how balanced and aesthetically pleasing it looks. Each case must be decided individually, considering the appearance, effect, size, purpose, etc., of the pictorial elements. Pictorial maps are artworks, as well as maps. III.5. Contribution of the Computer The model, when finished, should be zoom-able and navigable. Similar to the billboard technique, the textual illustrations, when looked at from any direction, will always be oriented towards the camera. III.6. Scale The scale is dependant on the purpose for which the map is used and whether it is suitable to demonstrate the level of detail intended. It can be a city, a country or an entire continent, depending on the purpose of the map. For the graphical map, I started with a single country, Italy, to display places of attraction with landmarks, city icons and mountain ranges. Later on, I created more examples of countries such as, France, the United Kingdom, and the entire continent of Europe. While building the pictorial map of Europe, I had to be selective and choose only some major capitals or cities and some very important and recognizable landmarks that identify those places. Some major roads connect those cities, and the entire continent is treated as one undivided piece of land. To represent statistical information such as population density, or political stature, 14 the purpose is to make comparisons of data between different states or countries. I selected a country, the United States of America, to display such comparisons among the states. I also chose the continent of Europe where I compared population density and political stature of different countries. III.7. Interactivity of 3D Pictorial Maps The practice of three-dimensional interactive maps, both graphical and statistical, is not yet widely explored. Interactivity will allow users to navigate these maps and view them from all directions and distances. The maps also need to have the capability to be easily updated. These maps are interactive. Users can view a city or a region from different perspectives and distance. These maps are printable, and easy to update. With any change of a region, a landmark, or a new road, changes can be made easily using the modeling software and the maps can be updated. They can be interactively viewed on the Internet by exporting the files in a format that is suitable for web viewing. 15 CHAPTER IV METHODOLOGY There are three major steps to making an interactive pictorial map. The first step is to build the 3D model using a modeling software, Maya. The second step is to apply texture maps to objects, using the process of baking shadows (i.e. retaining the shadows they cast). The last step is to make these 3D models of maps interactive and accessible to the user on the Web by transferring the built model from Maya to Viewpoint media, using a viewpoint plug-in. All these steps are being discussed in detail in the following sections. IV.1. Building the 3D Model Maya, a 3D modeling software by AliasWavefront, is used for modeling maps, their landform regions, features, symbols, and text. I built a wide variety of models of natural and built structures such as mountains, buildings, roads and vegetation. These are simplified versions of actual structures. The interactive speed of computer visualization greatly depends on reducing the complexity of modeled environments. Every detail of the model present need not be shown. They need to be oversimplified, yet preserving their personalities and characteristics. The models need to be aesthetically pleasing, with smooth surfaces, sometimes with soft curves and should use less disk space to be stored. In Maya, there are three kinds of modeling surfaces, polygons, subdivision surfaces and nurbs to represent the shape of an object. The choice of the method for representation is based on how well it represents the objects of interest, achieving satisfactory level of smoothness and effect with low mesh structure, how easy is it to 16 create, render or apply texture maps, transfer to viewpoint media, and to modify or animate it. And finally, how cheap it is to store and transfer. Considering all these factors, I used polygon surfaces to create the model, with less poly counts. All the above mentioned operations are easier to do with polygons. The memory and disk space is also cheap, which is imperative while transferring the map to viewpoint media, or downloading the map on the Web. Even though nurb and subdivision surfaces are smoother than polygon mesh surfaces, their mesh structure have increased density and thus become heavier. Thus no nurbs or subdivision surfaces are used, since they take too much disk space. Moreover, Viewpoint Technology does not support nurb surfaces. A satisfactory level of smoothness can be achieved with polygonal meshes, to make the models aesthetically pleasing, as well as lightweight. IV.2. Modeling the Graphical Map My inspiration for creating graphical pictorial map was the hand-drawn map of Italy by artist Saso Jankovic [2]. I aimed to recreate the hand-drawn style and character of that map. After I finished the map of Italy, I created some more examples of countries, such as, France, and the United Kingdom and the continent Europe. IV.2.1. Applying the Process of Caricature Simplification is a process of how much to put in and how much to leave out, as well as identifying major landmarks, city locations, road network, mountains and rivers. Exaggeration is portraying the distinctive features of a country or a place very broadly, and making it aesthetically pleasing and easily recognizable. Landmarks are used as symbols, as well as generic symbols such as, cities and modes of transportation. 17 IV.2.2. Caricature of Land and Choosing Major Elements I started with a base map of Italy. While making the 3D polygonal model of the land, I slightly distorted the shape than the original shape, so that the features can be displayed more in detail, the shape of the land looks simplified, aesthetically pleasing, and still easily recognizable (see Figure 2). Soft curvatures and straight lines simplify and beautify the shorelines and boundaries. Base map of Italy Caricature of base map in Maya Fig. 2. Modeling landform of Italy from a base map. Important features are identified, such as some major city locations, mountain ranges, rivers, road networks (see Figure 3) and also major landmarks that symbolize 18 that place. Fig. 3. Top view of Italy with important natural and built features. IV.2.3. Modeling Mountain Ranges Highly recognizable mountain ranges have been chosen to model. These are caricatures, or simplified shapes with sharp edges and rock-cut faces, based on the way the hand-drawn mountains look in [2]. The chosen shape and scale of the mountains is an artistic decision to make them look aesthetically pleasing. The scale of the mountains is relative to each other. And viewers can easily perceive the higher and lower mountain ranges from a distance. I have used three valence polygonal mesh structures for mountains to achieve the 19 sharp rock-cut look. A polygonal object is formed by a closed set of mesh. A mesh has several components, or geometric entities, such as, faces, edges (the boundary between faces) and vertices (the intersection points of edges). Valence is the number of edges adjacent to each vertex in a mesh. The corners of a cube have valence of three (see Figure 4). In the polygonal mesh of mountains, all vertices have a valence of three. It displays effective and realistic effect of rugged mountains, with rock-cut faces, which is closely similar to the look of the hand-drawn mountains. These units of mountains, when juxtaposed with one another can demonstrate realistic mountain ranges (see Figure 5). Polygonal mesh of a cube Polygonal mesh of mountains Fig. 4. Polygonal mesh of three valence objects. IV.2.4. Creating Rivers, Road Networks and Vegetation The major rivers are identified and they are formed as part of the main landform (see Figure 2). The simple and sharp corners of the rivers bring dynamism to the shape of the landmass and break it into some major areas. The roads have simple treatment with relaxed lines, connecting some major cities. The roads are polygonal objects that 20 Fig. 5. 3D model of mountain ranges in Italy. give a flat, 2D effect which separates them from the buildings. Vegetation is showed by groups of spherical polygonal objects, with low poly count. These are grouped around mountain ranges, or major landmarks, to create a backdrop or setting for the landmarks (see Figure 5). IV.2.5. Caricatures of Buildings Viewers, when navigating maps, identify places visually by identifying highly recognizable landmarks. Such highly recognizable and important landmarks are identified and exaggerated to symbolize particular cities. For example, the colisseum of Rome, Big Ben of London, the Eiffel Tower of Paris or the Parthenon of Athens. Caricatures of buildings is an artwork where the features are simplified, or exaggerated and at the same time are aesthetically pleasing and easily recognizable. The number of openings, windows, or number of columns is reduced, retaining the proportion of the building (see Figures 6 and 7). The biggest advantage of caricaturing is that these simplified 21 models become less heavy and easy to upload on the Web. Tower of Pisa Colisseum in Rome Fig. 6. Examples of reduction of number of openings for simplification. Parthenon in Athens Hagia Sophia in Turkey Fig. 7. Examples of simplification and exaggeration. 22 IV.2.6. Text Text is an integral part of pictorial maps. I created 3D polygonal text Maya. The fonts are chosen carefully, so that these stylized texts are suitable with the character of hand-drawn maps. For clarity, I have used two different styles of fonts, one for naming the sea or water bodies and another style of text for naming the cities. After creating 2D polygonal text, I gave it a slight thickness to make it 3D, and placed them right above the location (see Figures 8, 9 and 10). It is important that the text is 3D, because it is easier to render 3D text in Viewpoint. The text should not clash with the models of buildings, but should blend in with the background. Fig. 8. 3D polygonal text in Maya. The goal is to convert the text into billboard text, which means they will always be oriented towards the camera. In modeling stage in Maya, the text is kept static, and is made billboard later after being transferred into a Viewpoint media file. The procedure of making the text billboard is discussed in a later section. Scale and 23 Fig. 9. Final 3D model of Italy with text. Fig. 10. Top view of 3D model of France. 24 readability of text is of utmost importance. It needs to be readable, and noticeable from a certain distance. IV.3. Modeling the Statistical Map My inspiration for creating statistical pictorial map was the clay map of New York city by Anna Walker [1]. I aim to recreate the style and methods of the simplified sculptural shapes and pastel shades of color of that map in my 3D statistical maps. I created statistical maps of the United States and the continent of Europe based on the style of this particular sculpture. IV.3.1. Cognitive Reading of Statistical Maps There are several cognitive stages to the attainment of the goals of map reading of statistical maps. These are dominated by sensory, especially visual, and perceptual processes and finally cognitive process to comprehension, encoding information into memory and reasoning [12]. First step is to identify which geographic area the map represents or is designed to communicate. The second step is to understand the use of legend and the technique to represent of values of the mapped data. It is also important to understand the symbolization of the legend, identification of the number and ordering of the categories used. The final step is to extract information from the map, such as, the data values associated with specific locations, or identifying areas with the highest rates etc. This involves, comparing visual properties of the map, understanding the legend symbology corresponding to the particular geographic area, and furthermore understanding the rules for grouping neighboring geographic units for recognizing a geographic pattern. 25 IV.3.2. Traditional Ways to Read a Statistical Map Graphical elements available to the designer of a statistical map are: hue, saturation, brightness (or intensity) of color, symbol type, size, and texture (e.g., hatch patterns). Color can be represented in several ways, such as, use of the hue,saturation and brightness scale to illustrate the concepts. Incase of 3D statistical maps, an added attribute is the change of height of models. The shades or patterns that represent each data must be clearly identifiable, not only on the legend, where the category boxes are ordered, but also on the map, where there is no predetermined order. If the shades are too similar, it will lead to difficulty in integrating the legend and map, even when separately the legend and the map appear to be clear. IV.3.3. Understanding Color Concepts In order to use color most effectively, we need to understand the ways in which readers mentally process the map. William Playfair wrote in 1802 that regarding numbers and proportions, the best way to catch the imagination is to speak to the eyes [13]. In a monochromatic (gray) scale map, dark to light shades usually correspond to high to low rates. It can be convenient for cheap color printing of maps, but if the progression from dark to light shades is not clear, or the discriminability of the gray shades is poor, its use as a category shade can confuse some readers, who are unsure whether black or white represents the highest rate. Changes in the hue of color can alleviate the discriminability problem. A combination of hue and saturation is where the saturation is greater for higher rates. For example, less to highly saturated green can be followed by less to highly saturated brown as rates increase from lowest to highest. 26 It is advisable that hues not be used for a quantitative ordering. Use of different hues, without changing the saturation, to convey an ordering on a map usually confuses the reader to perceive the intended order of the hues. However, highly discriminable hues can be selected and can be used to convey a qualitative (or equivalently a dichotomized quantitative) characteristic of the mapped data [13]. IV.3.4. Use of Graphs or Bars Graphs are indispensable for examining variables across geographic areas. A bar or column chart or graph uses rectangles (bars) to represent different amounts or data (see Figure 11). This gives the relationship between quantities that vary with one Fig. 11. A bar graph shows comparison of data. another. Charts or bars can be used to illustrate a single variable data, or compare the differences between objects or to show change over time. 27 IV.3.5. Choosing Scale, Type and Method I have created statistical maps for both a country (the United States) and a continent (Europe). For the United States of America, I am providing two sets of information. One is the population density of states, and the other is, grouping states according to the election results of 2000 (Democrats and Republicans). In case of Europe, there are also two sets of variables, population density of countries, and their political status (members and non-members of the European union). In my statistical map, I am using highly discriminable hues for conveying one set of variable (e.g., political status of states or countries). To show the population density of states or countries, I am changing of heights of 3D models. This will allow the reader to perceive general patterns by mentally grouping hues for reading one set of data, and also permit rate readout by the change of heights of 3D models. The change of height of 3D models will act similar to a 2D bar or column chart, where a single variable data can be compared between states or countries, or differences can be shown. This comparison is mostly visual, where the reader can understand which state or country has the highest rate and which has the lowest. IV.3.6. Choosing Color Schemes Color schemes should be chosen carefully, giving due consideration to the size or scale of the map, along with the amount of reproduction and reduction expected. I have used complementary colors, such as red, blue and yellow so that the reader would not have trouble differentiating. The same color is not used to portray both the categories of data and a geographic (i.e., lake or ocean) or other feature (i.e., borders) in the map, as this can be confusing to readers and blur the clear presentation of the data. 28 IV.3.7. The Process of Caricature I used polygonal mesh structure for modeling the statistical maps in Maya. I started with a base map of the United States of America and created models of all the states of the country. The shapes of the states are simplified, the details have been reduced, Fig. 12. Simplified shapes of states of USA with changing heights. keeping the size accurate, so that they are still easily recognizable and yet aesthetically pleasing (see Figure 12). Soft curves have been used to make it aesthetically pleasing. 29 IV.3.8. Using Mel Scripts Mel scripts are specialized programming language used in Maya. I have created a Mel script for changing the height of the 3D models (see Figure 13). This script allows me Fig. 13. The slider interface in Maya for changing heights of models. to create an interface, where if a value is entered, the height of the models will increase or decrease according to the input value. This interface can be used for inputting data 30 for each of the states, and change their heights accordingly. Figure 14 is an example of Fig. 14. Melscript for creating the slider interface. the Mel script to build the window or interface with sliders and a button, for changing the height of models. I named the interface as “height builder”. The interface has sliders for each of the 3D models of the states, such as Alabama, California, Texas, etc. The sliders have a minimum value of 0, and a maximum value of 7. There is also a button that resets heights and sets all values to 0. I created an object 31 “controller” which has attributes named after all the states of the United States. For example, in the Mel Script, the attributes are referred to as “controller.Alabama” or “controller.California” etc. These attributes are connected to the height (translate Y) of the corresponding models of Alabama or California. Changing the values of these attributes directly change the height of the corresponding models of Alabama or California. IV.3.9. Create the Surrounding Landform and Text The surrounding landforms are created with 3D polygonal meshes. These are kept in monochrome, so that it does not clash or interfere with reading the map information clearly. Static 3D polygonal texts have been used for the surrounding sea or water bodies (see Figure 15). Fig. 15. Polygonal text in Maya for statistical maps. 32 IV.4. Applying Textures and Shaders Once the model is created, the next step is to apply texture maps to objects. These texture maps are “baked” using a baking procedure in Maya. Baking the shadows give the models a rendered look. The textures of the objects are integrated with the shadows that the surrounding lights create. So, the models look similar to a rendered image. Once the shadows are baked, the model can be directly transferred into the Viewpoint media, where no added light sources are necessary to create the shadows. Once in Viewpoint, a “toon shader” can be applied to give the graphical map a hand-drawn look. This is discussed later in the chapter. IV.4.1. Process of Baking Shadows in Maya A dome of light in Maya The hypershade option Fig. 16. Dome of light and hypershade option in Maya. The first step is to create a dome of lights surrounding the model, which lights up all the surfaces of the model (see Figure 16). These are directional lights, where there is one main key light which acts as the sunlight to create the main shadow. The next step is to open the “file texture” options in hypershade. I chose each model, and its corresponding material from the hypershade and converted it into a 33 file texture (see Figure 16). Each objects need to be “automatic mapped” before the conversion process is done. After the conversion is done, it will create a file texture for that particular model or object, with the shades and shadows baked into it (see Figures 17 and 18). The United States Europe Fig. 17. 3D models with baked shadow. IV.5. Transferring Model to Viewpoint Media The main purpose of transferring the model to Viewpoint Media is to make it accessible and navigable to a larger number of viewers or users on the Web. The Maya models are converted as .html files or web pages on the web, and they are interactive. The user can view the model or the map from various angles, zoom in and out, pan and rotate by dragging the mouse. A “toon-shader” is applied in Viewpoint, where the models will look very similar to the hand-drawn maps. Also the text is turned into “billboard texts” which is always oriented towards the camera. The files are compressed into a smaller size, to make them user-friendly and to download in less time. 34 Fig. 18. Model of Italy with baked shadow. IV.5.1. Advantages of Viewpoint Technology Viewpoint Technology is a unique technology of Viewpoint Corporation. It allows enhanced perspective, animation, integrated media messaging, such as 2D graphics, text, flash movies, 3D models, audio, video, and interactivity for displaying 2D and 3D content on the web. It displays content from all angles, allowing the user to move, rotate, pan and zoom 3D models on the web and expand video display to the size of the desktop. Users can view it from any angle or to stand in the center of a 360degree, by moving the mouse. For installing Viewpoint applications and utilities, and to see examples, visit http://www.viewpoint.com/ Viewpoint Technology is supported by Windows, Mac and Unix, and browsers such as, Netscape, Internet Explorer and AOL. The latest versions of Netscape and Internet Explorer (7.0 and up) have integrated viewpoint media player. Viewpoint 35 supports a lot of features in Maya. It exports geometry (polygon mesh), vertex normals, UV texture map coordinates, 2D textures, color and transparency, translate, rotate, and scale animations of a model, camera animation, instanced geometry, and multiple cameras. It compresses the file to a much smaller size, making it friendlier for the web user. Downloading Viewpoint Technology is free if used for educational purpose. A key advantage of Viewpoint Media Player is the platforms superb rendering of 3D models. Viewpoint Media Player delivers rich media content over any standard network connection, enabling non-broadband Internet users to view the interactive content. IV.5.2. Applications to Install There are several applications necessary to install for exporting a 3D model from Maya to Viewpoint as an .html file or a web page on the web. The “Viewpoint Media Exporter” is a Maya plug-in which exports 3D polygonal models, and animation directly from Alias Wavefront Maya 3.0 or 4.0, by converting them to Viewpoint media files (.mts and .mtx formats). The .mts file contains a compressed collection of resources, such as, 3D objects, material properties, sound, object movies, animators, interactors, and environmental lightmaps, orchestrated by an .mtx (XML-based) file to create a scene. The .mtx file is an XML-based file that contains the hierarchical relationships between elements in the scene and is the script for staging them. The compressed version of this file is binary and has the extension .mtz. The “Viewpoint Scene Builder” is an application which is used to assemble and edit the .mtx files created and publish it into html files for the web. To view the interactive html files online, one also needs to install the “Viewpoint Media Player”. It is a web browser plug-in that reads both the .mtx and .mts files to display the rich 36 media scene. To publish Viewpoint Technology scenes on the web, the user needs to obtain a “Broadcast Key”. A “Broadcast Key” is a license or a serial number, which can be obtained free online. It is necessary to display the content without watermarks (the name VIEWPOINT). It can be embedded in the html file by publishing the html file using “Viewpoint Scene Builder” or by manually changing the script of the html file. Another important step to publish Viewpoint Technology content to the web is to register MIME types that you intend to use. Viewpoint recommends to register .mtx, .mtz, .mts, and .mzv as MIME types. IV.5.3. Steps to Export a Maya File to Viewpoint The first step is to set up the “Viewpoint Media Exporter” plug-in in Maya. The “export” option in Maya (see Figure 19) gives the user options to set camera, shadows, materials, animation controls and compression options while export. IV.5.4. Discussing Export Options in Maya I chose the “perspective” view as the main camera (see Figure 20). The “navigation mode” lets you choose how the camera should respond when the user drags the mouse to view the scene in Viewpoint Media Player. The “orbit” mode is where the camera is attached to an invisible stick, anchored at a pivot point at some distance in front of the camera. The camera always looks toward the pivot point. The user can swing the camera around to look at the pivot point from every direction. “Normalize the scene” adds a scaling transformation in the .mtx file to shrink or expand the whole scene to fit in a cube one unit in size. Materials exported are anistropic, blinn, lambert, phong, phongE and surfaceShader. All materials are treated alike. But their lighting and surface characteristics 37 Fig. 19. The export option window in Maya. are ignored. The only attributes used are color, transparency, and diffuse. I used the “default” texture size which affects the chosen material only (see Figure 21). The “height” option lets you set the location of the shadow’s ground plane in relation to the world plane (see Figure 22). I have aligned the shadow with bottom of lowest object. I applied compression to textures and geometry. Generally, TrixelsNT is the best compression for visual fidelity and small overall file sizes (see Figure 23). 38 Fig. 20. The camera setting option window. Fig. 21. The materials setting option window. Fig. 22. The shadow option window. Fig. 23. The compression setting option window. 39 IV.5.5. Assembling in Viewpoint Scene Builder Viewpoint Scene Builder is an essential application for assembling a scene, editing and publishing it into .html/.mtx/.mts format. Once the mtx file is opened, the 3D model can be seen. It can be directly edited, applied textures or shaders, change material colors for all objects or shadows. I chose texture maps for all 3D objects, and the default material color for the texts. I did not use any additional light sources as the textures are already baked with shadows. From the global parameters, I chose the shadow color to blue, which indicates the sea. IV.5.6. Billboard Texts and Toon Shader Applying Billboard text Applying toon shader Fig. 24. Billboard text and toon shader in Scene Builder. I selected all text with the right mouse button, and turned on the “billboard” option from the menu to have the text move with camera. 40 I applied a “toon-shader” that will make the model look as close as a hand-drawn image. By turning on the “edges” option from the materials menu, the objects will have visible edges (see Figure 24). I applied the toon-shader only for the graphical maps. IV.5.7. Publishing on the Web When publishing the file, scene builder creates a set of .mtx or .mtz, .mts, and .html file. A broadcastkey.mtx file will be referred to by the .html file. The .html file is the web page created to display the Viewpoint scene. A resource folder is automatically created when the .html file is published. It contains a JavaScript library file, MTS3Interface.js, required for the proper display and functioning of the Viewpoint scene across supported browsers and platforms, and also provides a standard API for scripting with a Viewpoint scene. IV.5.8. View Web Page with Media Player Once a Viewpoint Media Player is installed on the computer, the user can view the interactive pictorial map (.html file) with Netscape Navigator, Internet Explorer, and AOL. With versions 7.0 and higher, the media player is automatically installed. 41 CHAPTER V RESULTS The results are interactive 3D Pictorial Maps which can be viewed as web pages online. These maps are scalable, zoom-able and navigable from all directions. The user can navigate these maps with a click of a mouse. V.1. Interactive Graphical Maps Here are some still images or “screen captures” from the interactive and navigable 3D graphical pictorial maps on the Web. Toon shader has been applied to give these maps a look that is close to a hand-drawn one. The style of the maps is also similar to the graphical pictorial map of Jankovic in [2]. The following images are graphical pictorial map of Italy (see Figure 25), the United Kingdom (see Figure 26), France (see Figure 27) and Europe (see Figures 28 and 29). V.2. Interactive Statistical Maps I created statistical maps of the United States of America and Europe (see Figures 30 and 31). The complementary colors demonstrate one set of data (political stature), whereas the change of heights compares another (population). The shape of the states and countries have been simplified. The user can change the height of models using the Mel script interface in Maya and then transfer the model to Viewpoint to view as a web page. The colors in the map of the United States of America show election results of 2000. Whereas, the colors in the map of Europe show members, non-members and applicants of the European Union. 42 Fig. 25. Graphical pictorial map of Italy. Fig. 26. Graphical pictorial map of the United Kingdom. 43 Fig. 27. Graphical pictorial map of France. Fig. 28. Graphical pictorial map of Europe. 44 Fig. 29. Graphical pictorial map of Western Europe. 45 Fig. 30. Statistical pictorial map of the United States of America. Fig. 31. Statistical pictorial map of Europe. 46 CHAPTER VI CONCLUSIONS AND FUTURE POSSIBILITIES The 3D models, compared to the hand-drawn sketch by Saso Jankovic in [2] and the clay model by Anna Walker in [1], can be evaluated whether these maps are more appealing and interesting than 2D maps and if the 3D tools are used in an effective way to convey the information. 2D pictorial maps are a mixture of perspective and plan. The 3D pictorial maps retain the same character of story-telling, yet is zoomable and navigable. VI.1. Evaluation of Size and Appearance The graphical and sculptural maps are aesthetically pleasing to look at. The 3D graphical map, with the toon-shader, looks very close to the hand-drawn graphical map. It has the similar simplified, cartoon-like, hand-painted look, but, remaining true to its medium, the 3D graphical map does not completely imitate the 2D graphical map. It retains its identity as a 3D map, with the shadows and textures. In case of the 3D sculptural map, the look of the 2D map has been imitated, in terms of simplified shapes and soft curves. The colors chosen for the models are also light pastel shades, close to the colors used in the 2D map. Integrated shadows in the 3D models emphasize the sculptural look. The final Viewpoint output files are smaller in size than the actual Maya files. Compared to other software, such as Macromedia Director, that are also used as Maya plug-ins to create on line interactive 3Dmodels, Viewpoint create files much smaller in size. The actual Maya files of my 3D models range from approximately 2.0 - 8.0 megabyte in size, along with textures and shaders. Once transferred into Viewpoint, 47 the final output .mtx files range from 15.0 - 100.0 kb. It can be compressed further into .mtz files to reduce the size. Avoiding complex geometric shapes of models and text is helpful to keep the file sizes lower. VI.2. Limitations of Viewpoint There are some limitations or problems that I faced while demonstrating the 3D models in Viewpoint. The text get lost when looked at from beyond a certain distance. In Viewpoint, the scale of text is fixed, and cannot be changed or controlled. But, the text need to be clearly readable from all distance. While creating the model, I fixed the size of the text relative to the size of the 3D models, so that it looks nice when looked at from a certain distance. Increasing the size of text, affects the overall look. It becomes too large and disproportionate with the rest of the model. I also experimented with the spacing of the text, but the effect is too small and legibility problem still remains. The chosen style of text matches with the hand-drawn style of the map. Some other fonts, such as Arial, or Times Roman, are sharper and more clear to read, but not suitable with the overall look of the map. Changing the color of the text was another option. But, I wanted to use the color white for text, so that it does not interfere with the wide variety of colorful models. Since tourists navigate the map visually based on highly recognizable landmarks, I wanted the 3D buildings and natural landscape to be more in emphasis. Changing the color of text has the risk of making the model look too crowded. It needs to blend in with the background, and yet be clearly legible from all distance. Also, the width of the outline of the text is not changeable in Viewpoint. To solve this problem, there should be further provisions in Viewpoint Technology to make the text scalable. Further research is possible to try to control the scale of 48 text, where the text size will increase when the camera zooms-out. This can solve the problem of text legibility from any distance. In Viewpoint, the user can zoom out to infinity. But, the scale or distance for zooming needs to be fixed, so it cannot be zoomed out beyond a certain distance. That will also allow the text to be always readable. Another problem faced while using billboard text is that sometimes the billboard text collide with the buildings, as the camera rotates. It becomes very difficult to position all of the text in such locations, that they can move freely, and at the same time look aesthetically pleasant. The pivot point of the text cannot be controlled as the angle of view is changing. VI.3. Future Research Possibilities Once a method has been established to create these 3D pictorial maps, it will allow for further researches or possibilities of development of programs or software that will create pictorial maps out of any 2D map outlines, especially statistical maps. A template may be created with generic symbols where the user can create pictorial maps out of any 2D maps with the use of templates of symbols. For statistical maps, more information can be shown with uses of 3D models, such as, houses or people, based on user input and what kind of data they represent. Mel Scripts can be generated accordingly. Incase of interactivity, further research can be made on how a lot more information can be provided without increasing the file size. This technique is applied in 2D zoom-able tourist maps, but it needs to be explored further in 3D interactive maps. For example, when creating a map of a continent, it is not possible to show every important building of all countries of that continent. It will take more memory space 49 and that will increase the download time of maps on the Web. The map can be made with few selected information and then certain regions of that continent can be made clickable (hotspots). This may allow gradual zooming-in and show more details of that area. There can be movement, in terms animated time-series displays, and fly-through animation. The prospect of showing the growth or change with time, as in density, population or urban growth is under consideration. It can offer a base for further researches, where the entire surface of the earth can be created and revolved into a 3D sphere. The creation of such pictorial maps can be derived from a database of integrated Geographic Information System and tourism information. Further experiment can be done on how the creation of these maps can be successfully based on a logical integration geographic information stored in a GIS system and tourism information stored in database management system. 50 REFERENCES [1] N. Holmes, Pictorial Maps: History, Design, Ideas, Sources, New York, Watson-Guptill Publications, 1991. [2] S.A. Jankovic, “Image of a Pictorial Map of Italy,” Ars-Cartae.com, (available at http://www.ars-cartae.com/), accessed in 2003. [3] G. Glazer and H. Glazer, “Collection of Pictorial Maps,” George Glazer Gallery, (available at http://www.georgeglazer.com/), accessed in 2003. [4] Osher Map Library, “Collection of Pictorial Maps in Different Media,” Carto-Maine-ia: Puzzles and Wraps and Oddball Maps, (available at http://www.usm.maine.edu/maps/exhibit6/), accessed in 2003. [5] The Library of Congress, “Images of Pictorial Maps,” Zoom into Maps, (available at http://www.loc.gov/loc/lcib/9909/litmap.html), accessed in 2003. [6] USGS, I. Bortins, S. Demers, and K. Clarke, “Images and Information about Statistical Maps,” Cartogram Types, (available at http://www.ncgia.ucsb.edu/projects/Cartogram Central/types.html), accessed in 2003. [7] S. Demers, “Mural of a Population Map,” Mural Cartography: Combining the Art and Science of Map Making, (available at http://www.geog.ucsb.edu/muralcartography/muralcartography.html), accessed in 2003. [8] D. House and C.J. Kocmoud, “Images and Information about Continuous Statistical Maps,” Continuous Cartogram Construction, (available at 51 http://davinci.informatik.uni-kl.de/vis98/archive/tp/papers/Vis98.PDF), Texas A&M University, accessed in 2003. [9] J.A. Dykes, K.M. Moore, and D. Fairbairn, “Images of Virtual Pictorial Maps,” From Chernoff to Imhof and Beyond: VRML and Cartography, (available at http://www.geog.le.ac.uk/jad7/VRML99/paper.html), accessed in 2003. [10] C. Haeberling, “Image of a Graphical Pictorial Map in 3D,” Symbolization and Visualization of Topographic 3D-Maps, (available at http://www.karto.ethz.ch/research/research13.html), Institute of Cartography, accessed in 2003. [11] C. Mendel, “Tourist Maps of Cities,” Carol Mendel Maps, (available at http://www.carolmendelmaps.com), accessed in 2003. [12] G. Palsky, “Information about Statistical Maps,” The Debate on the Standardization of Statistical Maps and Diagrams (1857 - 1901), (available at http://www.cybergeo.presse.fr/cartogrf/texte1/palsky.htm), accessed in 2004. [13] G. Palsky, “Information about Reading Statistical Maps,” The Process of Reading Statistical Maps: The Effect of Color, (available at http://lib.stat.cmu.edu/scgn/v51/section1 3 0 1.html), accessed in 2004. 52 VITA Asma Naz 1211 Vine Street, Apt. 204 Denver, CO-80206 anaz03@gmail.com Education M.S. in visualization sciences B. Arch Texas A&M University, December 2004 Bangladesh University of Engineering & Technology Bangladesh, August 1997 Interests 3D Modeling Web Design Graphic Design Employment 3D Modeler & Graphic Designer H+L Architecture 1621 18th Street, Suite 110, Denver, CO-80202 September 2004 - Present The Academy Arts Texas A&M University August 2003 - August 2004 Texas A&M University August 2002 - December 2002 Web & Graphic Designer Teaching Assistant