PROPOSAL OF EXPERIMENTS FOR EVALUATION OF …...Pavel Hájek, Karel Jedlička, Václav Čada Ing. Pavel Hájek* +420 377 639 208, [email protected] ... (such as map composition, a symbolization,

PROPOSAL OF EXPERIMENTS FOR EVALUATION OF CARTOGRAPHIC PRINCIPLES ON VIRTUAL 3D MAPS OF

URBAN AREAS

Pavel Hájek, Karel Jedlička, Václav Čada

Ing. Pavel Hájek* +420 377 639 208, [email protected]

Ing. Karel Jedlička, Ph.D.* +420 377 639 210, [email protected]

Doc. Ing. Václav Čada, CSc.* +420 377 639 205, [email protected] * NTIS – New Technologies for the Information Society – Research Center, Faculty of Applied Sciences, University of West Bohemia, Univerzitni 8, Plzen, 306 14, Czech Republic ACKNOWLEDGEMENT This paper was supported by the Project LO1506 of the Czech Ministry of Education, Youth and Sports and the Project SGS-2016-004 Application of Mathematics and Informatics in Geomatics III.

Abstract Traditional cartographic principles (such as a map composition, a symbolization, etc.) have been known for decades so far. Thanks to the evolution of ICT and cartographic products as well, so called (virtual) 3D maps are currently in the forefront of interest for both professionals and general public. This contribution is focused on a proposal of an evaluation of symbolization techniques (as one of traditional cartographic principles) used for virtual 3D maps of urban areas. It tents to investigate, which kinds of principles are still valid for 3D objects displayed within a 3D scene using a monitor, which need to be altered and which need to be newly defined. The evaluation itself is proposed to be done based on an online questionnaire. That questionnaire should be composed mainly of interactive 3D web scenes, complemented with related 2D representations of the same situation. These 3D and 2D representations serve as a background for representative tasks and questions to be answered by participants. The chosen tasks are based on dealing with principles used for visual variables, map symbols, means of map representation and on a dimensionality of displayed data. The results of the online questionnaire (answers of the tasks) are gathered, compiled and evaluated to provide a report of the examined cartographic principles describing their usability in the field of virtual 3D maps of urban areas. The paper’s aim is to boost a discussion based on presented preliminary results describing which cartographic principles apply for the virtual 3D maps of urban areas seamlessly and which have to be adapted.

Keywords: Proposal, cartography, experiment, virtual 3D map, urban area

INTRODUCTION

Representation of real objects in their native (three-dimensional) form using ICT is very popular these days. This trend is presented even in the field of cartography and cartographic products, because 3D objects are presented on maps more often than ever. Thus this contribution tends to present a cartographic product called a virtual 3D map of urban areas as an example of maps showing 3D objects natively. This map is a digital map presenting geographic and thematic features of an urban area (such as DTM, buildings, transport infrastructure, vegetation and so on) in a virtual 3D scene using a monitor.

A definition, a concept of such maps and a workflow for a creation of virtual 3D maps are not stated comprehensively yet, even though the International Cartographic Association is aware of “increasing influence … of new technology on map production and use”, ICA (2011). The Strategic plan for the ICA 2011 - 2019 (see ICA (2011)) states, among others, one threat which is “definitions may be dated and open to different interpretations”. This paper is based on such

Proceedings, 7th International Conference on Cartography and GIS, 18-23 June 2018, Sozopol, Bulgaria ISSN: 1314-0604, Eds: Bandrova T., Konečný M.

481

an idea. Thus this contribution continues with a proposal of a definition and of concepts applicable to virtual 3D map of urban areas. Concepts consist of approaches how to create a map and what to choose to be as a content of a map, rules of interactivity with the virtual 3D map and principles of cartographic representations used for the map as well as represented features. These concepts are built on traditional cartographic concepts for (digital) 2D maps.

After that, some of the above mentioned concepts and approaches are selected and investigated using an online questionnaire. That questionnaire consists of 2D pictures and interactive 3D web scenes showing the same objects. Such representations are used for solving representative tasks. On the base of such pictures and 3D scenes, questions of the questionnaire are answered by participants.

The benefits of this contribution are in a clarification of a definition of the term “virtual 3D map (of urban areas)”, in a presentation of some specific, but not only cartographic principles suitable for such maps and in opening of investigation and discussion about differences between 2D and 3D representations of specific geographical data.

A BRIEF OVERVIEW OF VIRTUAL 3D MAPS

Contributions such as Kraak (1988), Bandrova (2001, 2005), ICA (2003), Moellering (2007), Häberling et al. (2008), and Petrovič (2003) were dealing with definitions and characterization of 3D maps. Based on the mentioned papers, a virtual 3D map can be briefly defined as a “virtual 3D interactive cartographic representation of geographical features and phenomena”. Unlike the papers of authors mentioned above, the prefix “virtual” was added to the term “3D map” for the purposes of this paper to point out that this kind of a map is visible, but not tangible. The steps of designing such a virtual 3D map are described in Häberling et al. (2008) or Hájek et al. (2016).

Huge advantages of such virtual 3D maps are in an opportunity to display another dimension (1D/2D/3D models or symbols of objects can be presented via such a map), in a natively presented interactivity of changing a view of a user on the displayed 3D scene and in representing world in a more familiar way. Disadvantages can be seen in a possible information overload of a user (graphical map load), in an amount of data to be transferred/load for displaying a virtual 3D map (due to complex 3D objects presented) and in an eventual loss of perceiving a location/position of a user in a map due to the freedom of movement in a 3D scene.

Specifics of virtual 3D map of urban areas

Virtual 3D maps of urban areas are specific due to the displayed objects (features represented in urban areas, i.e. the content of the map) and the extent of the shown territory (smaller area, i.e. analogy to large or medium-scale maps). Bandrova (2001) or Biljecki et al. (2015) use a term “3D city map” to represent thought-like cartographic product as well. But virtual 3D maps of urban areas are not limited only to show cities understood as large human settlements.

The content of such maps, i.e. what is shown on them, can be described according to the categories stated in Bandrova (2005), such as “main content (terrain, roads, buildings), secondary content (city furniture, vegetation, information and other signs) and additional content (qualitative and quantitative information about objects).” More branched list of the displayed objects on the virtual 3D maps of urban areas is described in OGC (2012) within the definition of CityGML standard.

Regarding the mentioned content of the virtual 3D map of urban areas, the cartographic representation of the content (i.e. showed features) is also a specific issue for this kind of maps. Concepts of classic cartography are obviously still valid for virtual 3D maps, because the virtual 3D map is a subcategory of maps. Therefore cartographic principles such as graphical/visual variables (see Bertin (1983), MacEachren & Kraak (2001) or Halik (2012)) or ways of creation of maps and cartographic symbols (see Robinson et al. (1995), Slocum et al. (2010) or Kraak & Ormeling (2011)) must be used for virtual 3D maps as well. Selected cartographic principles for virtual 3D maps of urban areas are in more detail way described in Hájek et al. (2016). The particular topic regarding a design of cartographic 3D symbols of objects displayed on virtual 3D maps of urban areas is also specific for such maps and it is described further.


482

Cartographic principles for virtual 3D maps of urban areas

Based on the sorting of objects listed in the previous chapter, as well as on the research in Bandrova (2005), Petrovič & Mašera (2005) or Jedlička et al. (2012), four main categories were chosen as the most representing objects of urban areas’ landscape: buildings, roads, vegetation and city furniture.

A comprehensive research on suitable cartographic principles used for representations of such objects on virtual 3D maps of urban areas was done, see e.g. Hájek et al. (2016). A few of the outcomes from Hájek et al. (2016) are mentioned in the bullets list below. Some of the outcomes tends to be evaluated via experiments, which are proposed and described in the next chapter.

Selection of outcomes from the above mentioned paper is as follows:

• Symbols located by point/line/polygon have not only their 0D/1D/2D representations, but also 2,5D and 3D as well;

• Interactivity is the key attribute of virtual 3D maps due to free view on the displayed objects in the virtual 3D map resolving occlusion culling (obscuring objects);

• But due to interactivity, the dimensions of displayed objects can vary, e.g. when a 2D polygon is displayed as a 1D line in a certain view angle. That means values of visual variables used for cartographic representation of objects in the map can be altered, such as shape, size, orientation, resolution or perspective height (for the list of visual variables see Halik (2012));

• The influence of atmospheric phenomena (if they are presented in the map during the phase of its visualization) is considerable and it changes values of visual variables, such as lightness or saturation;

• 3D representation of objects is more complex (from a data size point of view), but more native for the user (from a recognition of objects point of view).

METHODOLOGY

Research methods used for investigating symbolization techniques suitable for virtual 3D maps of urban areas are described in this chapter. There are stated conditions to be fulfilled by such a research, which are profound for obtaining desirable data from respondents of the research.

To introduce research methods, Rohrer (2014) offers very comprehensive overview of instructions for using different User-Experience Research Methods (UERM). Figure 1 below describes a framework for the UERMs based on attributes of a research: attitudinal vs. behavioral (subjective vs. objective), qualitative vs. quantitative and division of the research based on a context of use of such a research.

Figure 1. An overview of User-Experience Research Methods (source: Rohrer (2014)).


483

The key question for this chapter is: “What kind of research methods to use to get as many information from respondents of the research as possible in a short period of time?”. In other words, what kind of research methods proposed in Fig. 1 to choose to fulfill following conditions:

1. Reach as many respondents as possible;

2. Get as many information from respondents as possible;

3. Use a research method which doesn’t take a long time to go through for a respondent;

4. Allow to respondent to take a part of the research anytime he/she wants;

5. Allow researcher not to interfere with ongoing research, i.e. no need to actively interfere into the ongoing research;

6. The research enables comparing between 2D and 3D representations of objects.

Assembling the conditions above with the context of research methods in Fig. 1 resolves in the following outcome. The chosen research should be both qualitative (get a certain commentary from respondents) and quantitative (to be able to quantify the answers of the respondents), subjective (based on the experiences of respondents) and objective (have a quantifying variables for statistical evaluation of answers) and of course easy to promote, i.e. using web technologies to allow higher amount of respondents to participate, without a need to take respondents to a lab. According to the list of research methods in Rohrer (2014) the combination of Unmoderated UX Studies, Desirability Studies and Customer Feedback fulfill the conditions above.

The research itself is based on a comparative case study described in detail in Goodrik (2014). That says: “Comparative case studies may be selected when … there is a need to understand and explain how features within the context influence the success of programme or initiatives. Comparative case studies involve the analysis and synthesis of the similarities, differences and patterns across two or more cases that share a common focus or goal.”

That means the executed research is based on a comparison of 2D and 3D representations of objects used for acquiring outcomes, which going to be profound information for an investigation, which kinds of cartographic principles are still valid for 3D objects displayed within a virtual 3D scene using a monitor. The description of the particular research is described in the following chapter.

RESEARCH

The combination of research methods fulfilling certain conditions was proposed in the previous chapter. This chapter tends to state research questions for the comparative study dealing with comparing 2D and 3D representations of objects on maps.

Key evaluation questions (KEQ) defined within the comparative study are based on comparing 2D and 3D representations of objects, on used cartographic principles for 3D representations of such objects respectively. The KEQs defined for such a comparative study are as follows:

1. What cartographic principles are applicable for 3D representations without modifying?

2. What cartographic principles are applicable for 3D representations with modifying?

3. What cartographic principles are specific for 3D representations?

4. Is there any relation between principles for 3D representations of objects on virtual 3D maps?

The key evaluation questions stated above are suitable for the comparative study itself. Moreover they serve as rudiments for research questions (RQ) regarding virtual 3D maps in more general way. The research questions are defined as follows:

A. Are there any tasks which are solved more simply using a 2D map or 3D form of a map?

B. Are the symbolic representations of objects as they are used on traditional 2D maps suitable for using within virtual 3D maps?

C. What kind of representations of objects are suitable for virtual 3D maps regarding dimension of such representations (0D/1D/2D/3D)?

D. Are there any alternatives of a legend for virtual 3D maps?


484

The list of such research questions can be extended rapidly, the above stated RQs is just a sample of them. Nevertheless the next chapter tries to propose experiments, which could be eligible as a source of knowledge for answering the RQs.

A proposal of experiments

To answer research questions mentioned in the previous chapter, concrete experiments should be proposed. A frame of such experiments used within the comparative study is described in this chapter.

Several conditions for proposed experiments were stated in the Methodology chapter. This chapter describes a particular implementation of experiments, which serve as a background for fulfilling the mentioned conditions and answering research questions. The experiments are incorporated into an online questionnaire, which is used for delivering the experiments to respondents via web browser. The responses of respondents from this questionnaire serve as a source for the comparative study, which deals with comparing 2D and 3D representations of objects on maps from cartographic point of view.

The created tool for such a study combines an online questionnaire together with experiments based on showing 2D and 3D representations of same objects separately. This tool is based on one, which was created by RNDr. Lukáš Herman (© 2016 - 2018 Lukáš Herman, Masaryk University in Brno, Faculty of Science, Department of Geography; http://olli.wz.cz/webtest). This tool was used or described by the creator for example in Herman & Řezník (2015), or Herman & Stachoň (2016). Technologies used for creating such a tool are:

• HyperText Markup Language (HTML) for creation of web pages;

• JavaScript, which is an interpreted programming language for web pages;

• Hypertext Preprocessor (PHP), which is a server-side scripting language designed also for web development;

• X3DOM library, which is based on JavaScript and enables embedding virtual 3D models in X3D data format directly into a web page.

The above mentioned technologies were used for implementing the online questionnaire and experiments into web pages. An example of a web page showing a 3D part of an experiment is in Fig. 2 on the next page. Description of particular parts of the web page depicted in Fig. 2 (each part is framed with a rectangle of different color) is in the bullets list below:

• Blue part - Title and a topic of an experiment;

• Green part - question for a respondent;

• Yellow part - the key part of the web page showing either 2D or 3D representation of same objects. Each experiment has two variations, the first showing a 2D representation of objects as a static picture and the second showing a 3D representation of the same objects within an interactive virtual 3D scene;

• Red part - proposed answers for the question asked;

• Magenta part - save and continue button.

The particular situation in Fig. 2 shows an experiment titled as “Experiment A” dealing with a topic of finding a particular type of buildings, showing a 3D representation of displayed objects.


485

Figure 2. An example showing a structure of a web page for a 3D variation of an experiment.

Another proposed experiment deals with a potential advantage of 3D representation of objects, especially with a decision about visibility between certain locations on a map represented as yellow dots (2D) or spheres (3D). An example of that experiment is depicted in Fig. 3 on the next page. Upper part of this figure shows the 2D version, the lower part shows the 3D version of the experiment.


486

Figure 3. Examples for 2D and 3D versions of an experiment dealing with visibility.


487

PRELIMINARY RESULTS AND FUTURE PLANS

The selected cartographic principles mentioned in chapter Cartographic principles for virtual 3D maps of urban areas are going to be evaluated via the online questionnaire. Several experiments will be prepared, each regarding to a specific cartographic principle transferred to virtual 3D maps, such as possible representations of point, linear, polygonal or three dimensional symbols. The experiments will display objects presented on a virtual 3D maps of urban areas as a particular example of using such principles within a concrete topic of a virtual 3D map. Based on results of statistical and empirical evaluation of such experiments (i.e. correctness of respondents’ responses, time of responses, screen logging of camera movement within a virtual 3D scene and commentaries of respondents), the KEQs and the RQs are going to be answered. That means an analysis of used cartographic principles applied for 3D representations will be done, i.e. which principles can be used for a 3D representations of objects within the virtual 3D maps, which have to be adopted and which are specific for 3D representations.

The preliminary results imply that the vast majority of traditional cartographic principles can be applied to the virtual 3D maps of urban areas seamlessly, minority has to be adapted and some principles, which are specific for the environment of virtual 3D maps need to be stated. Another outcome of the preliminary results speaks about the fact, that using the 3D representations of objects within the virtual 3D maps are not always better for task solving, than using 2D representations of objects on the traditional 2D maps for solving the same tasks.

CONCLUSION

This paper presented a proposal of experiments for evaluation of cartographic principles on a particular example of virtual 3D maps showing urban areas. At the beginning of this paper a brief overview of virtual 3D maps was described, including advantages and disadvantages of such maps. It was followed by specifics of virtual 3D maps of urban areas, as an example of use of virtual 3D maps for a particular topic. Especially a definition and a content of such maps was described there, together with some outcomes from a research on cartographic principles suitable for virtual 3D maps of urban area done in Hájek et al. (2016). Some of such cartographic principles tends to be evaluated using a research method, which was depicted in the chapter regarding methodology. A comparative study for a comparison of 2D and 3D representations of objects was chosen to be deployed as an instrument for investigation of applying cartographic principles on virtual 3D maps. The chapter about the research itself followed, including definitions of key evaluation questions for the comparative study dealing with questions about usability of traditional cartographic principles for virtual 3D maps and research questions dealing with general properties of maps, which can be answered based on outcomes from answering key evaluation questions. Preliminary results of the research and future plans about particular experiments for deploying were covered in the last chapter of this paper.

REFERENCES

Bandrova (2001) Designing of Symbol System for 3D City Maps. In Proceedings of the 20th International Cartographic Conference, Beijing, Vol. 2, pp. 1002-1010, ICA, Beijing, China.

Bandrova (2005) Innovative Technology of the Creation of 3D Maps. In Data Science Journal, Volume 4, 31 August 2005.

Bertin, J. (1983) Semiology of Graphics: Diagrams, Networks, Maps (1st ed). ESRI Press, California, USA, ISBN 978-1589482616.

Biljecki, F., Stoter, J., Ledoux, H., Zlatanova, S., Çöltekin, A. (2015). Applications of 3D City Models: State of the Art Review. In ISPRS International Journal of Geo-Information, 4(4), 2842-2889. doi:10.3390/ijgi4042842.

Goodrick, D. (2014) Comparative Case Studies, Methodological Briefs – Impact Evaluation No. 9, UNICEF, Office for Research. Available at <http://devinfolive.info/impact_evaluation/ie/img/downloads/Comparative_Case_Studies_ENG.pdf>, [cit. 2017-04-12].

Häberling, Ch., Bär, H., Hurni, L. (2008) Proposed Cartographic Design Principles for 3D Maps: A Contribution to an Extended Cartographic Theory. Cartographica 43 (3): 175-188.

Hájek, P., Jedlička, K., Čada, V. (2016) Principles of Cartographic Design for 3D Maps – Focused on Urban Areas, 6th International Conference on Cartography and GIS 2016 Proceedings, Vol. 1 and Vol. 2. Sofia, Bulgaria: Bulgarian Cartographic Association, 2016. s. 297-307, ISSN: 1314-0604.

Halik, L. (2012) The analysis of visual variables for use in the cartographic design of point symbols for mobile Augmented Reality applications. Geodesy and Cartography, 61(1): 19–30. doi:10.2478/v10277-012-0019-4.

Herman, L., Řezník, T. (2015) 3D web visualization of environmental information - Integration of heterogeneous data sources when providing navigation and interaction, In International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives 40(3W3), pp. 479-485.


488

Herman, L., Stachoň, Z. (2016) Comparison for User Performance with Interactive and Static 3D Visualization – Pilot Study, In The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLI-B2, 2016 XXIII ISPRS Congress, 12–19 July 2016, Prague, Czech Republic.

ICA (2003) A Strategic Plan for the International Cartographic Association 2003-2011 As adopted by the ICA General Assembly 2003-08-16. Available at <http://icaci.org/files/documents/reference_docs/ICA_Strategic_Plan_2003-2011.pdf>, [cit. 2017-10-03]

ICA (2011) A Strategic Plan for the International Cartographic Association 2011-2019. Available at <https://icaci.org/files/documents/reference_docs/ICA_Strategic_Plan_2011-2019.pdf>, [cit. 2017-10-03]

Jedlička, K., Hájek, P., Čerba, O. (2012) Experience with methods of 3D cartography gained during visualization of detailed geographic data for purposes of documenting cultural heritage (case study at the castle Kozel). In GeoCart'2012 and ICA Regional Symposium on Cartography for Australasia and Oceania. Auckland: New Zealand Cartographic Society Inc., s. 89-100. ISBN: 978-0-473-22313-7.

Kraak, M. J. (1988) Computer-assisted cartographical three-dimensional imaging techniques, Dissertation (PhD Thesis), Delft: Delft University Press. ISBN 90-6275-463-5.

Kraak, M. J., F. J. Ormeling (2011). Cartography visualization of spatial data. New York, Guildford Press, ISBN 978-1-60918-193-2

MacEachren, A. M., Kraak, M. J. (2001) Research Challenges in Geovisualization. Available at <http://schiewe-online.de/geovis_refs/MacEachren_Kraak.pdf> , [cit. 2017-12-31].

Moellering, H., (2007) Expanding the ICA conceptual definition of a map. In Proceedings: of the 23th International Cartographic Conference. Moscow.

OGC (2012) OGC City Geography Markup Language (CityGML) Encoding Standard, Open Geospatial Consortium, version 2.0.0, Publication Date: 2012-04-04. Available at <http://www.opengeospatial.org/standards/citygml>, [cit. 2017-04-11].

Petrovič, D. (2003) Cartographic design in 3D Maps, In International Cartographic Conference 2010 Proceedings, DOI 10.1.1.655.7874. Available at <http://icaci.org/files/documents/ICC_proceedings/ICC2003/Papers/248.pdf>, [cit. 2016-11-22].

Petrovič, D., Mašera, P. (2005) ANALYSIS OF USER’S RESPONSE ON 3D CARTOGRAPHIC PRESENTATIONS. Available at <http://www.mountaincartography.org/publications/papers/papers_bohinj_06/16_Petrovic_Masera.pdf>, [cit. 2017-11-22]

Robinson, A., Morrison, J., Muehrcke, P., Kimerling, A., Guptill, S. (1995) Elements of Cartography, 6th Edition, ISBN 978-0471555797.

Rohrer, C. (2014) When to Use Which User-Experience Research Methods, Nielsen Norman Group. Available at <https://www.nngroup.com/articles/which-ux-research-methods/>, [cit. 2018-01-02].

Slocum, T. A., McMaster, R. B., Kessler, F. C., Howard, H. H. (2010) Thematic Cartography and Geovisualization (3rd ed.). London: Pearson Education LTD., 561 p.

BIOGRAPHY

Pavel HÁJEK

Karel JEDLIČKA

Václav ČADA

The team of people from the Department of Geomatics at the University of West Bohemia consists of the experts in fields of Geographic Information Systems, database systems, 3D modeling, 3D cartography and geospatial data gathering. The members of the team are researchers and/or academic employees and have extended experiences from the long time research and from numerous projects, where they have been involved. More information about the members and the projects in which they have been involved can be found at http://kgm.zcu.cz/.


489

Documents

PROPOSAL OF EXPERIMENTS FOR EVALUATION OF …...Pavel Hájek, Karel Jedlička, Václav Čada Ing. Pavel Hájek* +420 377 639 208, [email protected] ... (such as map composition, a symbolization,