Low Computational Cost Crowd Rendering Method ...

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment Mohamed `Adi Bin Mohamed Azahar Department of Computer Graphics and Multimedia, Faculty of Computer Science and Information Systems, Universiti Teknologi Malaysia 81310 UTM Skudai, Johor Darul Takzim, Malaysia [email protected]

Mohd Shahrizal Sunar

Daut Daman

Department of Computer Graphics and Multimedia, Faculty of Computer Science and Information Systems, Universiti Teknologi Malaysia 81310 UTM Skudai, Johor Darul Takzim, Malaysia [email protected]

Department of Computer Graphics and Multimedia, Faculty of Computer Science and Information Systems, Universiti Teknologi Malaysia 81310 UTM Skudai, Johor Darul Takzim, Malaysia [email protected]

Abstract – In recreation of virtual cultural heritage sites and urban environment simulations, the rapid display of densely populated scenery is a common requirement. An empty virtual environment will diminish the immersive experience that the simulations suppose to be presented, that is why simulations of massive virtual characters are needed to deliver sensible immersive experience to the user. In recent years, the growth of research in this area has escalated and there has been many techniques developed in crowd simulations area, both for realtime and non-real time rendering. This paper presents several significant related works in crowd simulation and crowd rendering. Our aim is to address one of the crowd simulation problems, which is crowd rendering in large scale virtual heritage environment. Due to rendering large scale virtual heritage environment were already computational expensive, rendering crowd will add more load to the rendering process. We proposing range detection technique to produce low computational cost real-time rendering of virtual crowd in large scale virtual heritage environment. Range detection technique has been chosen because its ability to produce low computational cost compare to other rendering techniques. Keywords – crowd simulation; crowd rendering; virtual heritage; virtual environment

I.

INTRODUCTION

The wide use of computer graphics in education, entertainment, games, simulation, and virtual heritage applications has led it to become an important area of research. In order to deliver to the user an immersive experience, it is important to create an interactive, complex, and realistic virtual world in any simulation applications [1]. As the size and complexity of the environments in the virtual world increased, it becomes more necessary to populate them with peoples. Nowadays, virtual heritage is one of the latest growing applications in computer graphics industries. One of these paper objectives is to incorporate crowd into virtual heritage system with minimal computational cost for easy delivering on common personal computer. Generally, crowd simulation consists of three important areas. There are realism of behavioural [2], high-quality

visualization [3], and convergence of both areas. Realism of behavioural is mainly used for simple 2D visualizations because most of the attentions are concentrated on simulating the behaviours of the group. High-quality visualization is regularly used for movie productions and computer games, this type of visualization give intention on producing more convincing visual rather than realism of behaviours. The convergences of both areas are mainly used for application like training systems. In order to make the training system more effective, the element of valid replication of the behaviours and high-quality visualization must be added [4]. Virtual heritage is a term being invented for the use of digital techniques to reconstruct and preserve historical monuments and artifacts [5]. The social aspect must not be missed to get an accurate simulation of the cultural heritage site. Crowd is one of important social aspect that can be used to simulate correct situation in the particular historical site. Simulation of heritage site allows us to have better understanding of the selected site without being there. When rendering crowd in complex virtual heritage, there are two things need to be deal with: large-scale static virtual environment and moving crowd and traffic. These two things content thousands of polygons need to be managed to avoid low frame rate and low interactivity, making it uneasy for user to walkthrough in real-time. The limited number of polygons that can be processed in graphics processing pipeline, show that the need of 3D acceleration techniques to manage crowd rendering, especially in large scale virtual heritage environment. The purpose of this paper is to present some of the related work in crowd simulation, crowd rendering issues, a few types of crowd rendering methods that are usually used, and an overview on framework that will be used for developing crowd simulations in Ancient Malacca Virtual Walkthrough project. Afterward, we conclude our paper with some future research directions pertaining with the area. II.

RELATED WORKS

Virtual Heritage system is normally content of complex 3D object as example building, forest and crowd. Due to complexity and densely of virtual environment (VE) in

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

14

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

1. Previous work rendered in crowd simulation. according to researchers and group ofasyears. crowd rendering causes Figure the system usually by high Arrange various steering behaviours goal seeking, obstacle performance computer hardware and almost impossible in avoidance, path following, or fleeing [8], and introduced a lower specification computer system [1]. Crowd rendering simple finite-state machines behaviours controller and spatial can be divided on three parts: geometry, motion/animation queries optimizations for real-time interaction with groups of and behavior [2]. For interactive rendering crowd simulation characters [9]. the geometry part needed to be managed. The geometry part More recent work was studies on group modelling based of crowd rendering can be managed by 3D graphics on hierarchies. Niederberger et al. [10] proposed architecture acceleration techniques: culling, level-of-detail (LOD) and of hierarchical and heterogeneous agents for real-time image-based impostor [3]. Example applied virtual applications. Behaviours were defined through specialization environment is Ancient Malacca Virtual Walkthrough of existing behaviours types and depend on multiple project used this technique to enhance their virtual inheritances to create new types. An approach that has environment rendering speed [1]. These acceleration become more common now was geometry baking. By taking techniques also applied to CHARISMA project for Medieval snapshots of vertex positions and normals, complete mesh Norwich virtual heritage environment [4]. In Virtual Hagia descriptions were stored for each frame of animation as in Sophia, optimization techniques needed to lower amount the work of Ulicny et al. [11]. Another approach was through polygons for interactive simulation [5]. dynamic meshes, which was using systems of caches to reuse skeletal updates. A hybrid of baked meshes and A. Real-Time Crowd Simulation dynamics meshes was found in Yersin et al. [12] where the Real-time crowd simulation is a process of simulating the graphics programming unit (GPU) was used to its fullest. movement of a large numbers of animated characters or A real-time crowd model based on continuum dynamics agents in the real-time virtual environment. Crowd has been proposed by Treuille et al. [13]. In their model, a movement in certain cases requires the agents to coordinate dynamic potential field integrates global navigation with among themselves, follow after one another, walking in line moving obstacles, efficiently solving for the motion of large or dispersing using different directions. All of these actions crowd without the need for explicit collision avoidance. In will contribute to the final collective behaviours of the crowd addition, Mao et al. [14] presented an effective and ready to that must be achieved in real-time. Unlike non-real-time use framework for real-time visualization of large-scale simulations which is able to know the full run of the virtual crowd. Script that describes motion state and position simulated scenario, real-time simulations have to react to the information was use as input, provides convenient interface situation as it unfolds in the moment. Real-time rendering of and makes the framework universal to almost all application a large number of 3D characters is also a challenge, because of crowd simulation [14]. Recently, Berg et al. [15] have it can exhaust the system resources quickly even for a introduced a multi-agent navigation algorithm using a simple powerful system [4]. and effective combination of high-level and low-level Figure 1 shows a timeline of the previous works that methods that model human spatial navigation. A prehave been done in crowd simulation fields. A behavioural computed roadmap was used for global path planning and animation of human crowd was based on foundations of Reciprocal Velocity Obstacles was used for local navigation group simulations of much more simple entities, notably and collision avoidance [15]. flocks of birds [5] and schools of fish [6]. Earliest procedural B. Real-Time Crowd Rendering animation of flocks of virtual birds called Eurhythmy was developed from concept that was presented at The Electronic The complicated part when dealing with thousands of Theater at SIGGRAPH in 1985 and the final version was characters is the quantity of information that needs to be presented at Ars Electronica in 1989 [7]. The flock motion processed. Simple approaches, where virtual humans are was achieved by a global vector force field guiding a flow of processed one after another without specific order will flocks. In his early work, Reynolds describes a distributed produce high computational cost for both the central behavioural model for simulating aggregate motion of a processing unit (CPU) and graphics processing unit (GPU). flock of birds. The idea was that a complex behaviour of a This is the reason why data that flows through the same path group of actors can be obtained by simple local rules for need to be grouped for an efficient use of the available members of the group. The flock was simulated as a complex computing power. Therefore, for the best simulation result, particle system, using the simulated birds, called boids, as characters capable of facial and hand animation are the particles. Reynolds, later extended his work by including simulated in the area near to the camera to improve

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

15

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

believability, bility, while for farther area, less expensive representations are used. Concerning efficiency of storage and data management, database must be used to store all the virtual human-related related data. In this chapter we will discuss one of the major components in crowd simulation development which is rendering.

Collision Avoidance

Rendering Algorithms

sources, such as the Sejarah Melayu and the descriptions by Portugese writer; Tome Pires es described the city and the empire as an opulent and prosperous centre of maritime Malay civilizations. As a maritime empire, trading and commercial activities, both local and foreign, became the mainstay and the backbone of her economy. The focus area off visualization is Central Business and Administrative District of the Malacca Empire. The project is visualized in real-time time rendering mode using SGI Onyx 3800 with 16 CPU, 32GB RAM and three Infinite Reality3 graphics pipes.

Motion Planning Issues

Behavior of the Groups

Computing Trajectories

Figure 2.Some of crowd rendering issues. Figure 3. Ancient Malacca Virtual Walkthrough.

Figure 2 shows certain problems that arise when rendering crowd [16]. For instance, collision avoidance problems for a group of peoples in the same place requir required different strategies in comparison with the methods used to avoid collision between individuals. Moreover, motion planning used in a group that walks together requires more information compared to individual motion planning. The trajectories computed forr agents in the same group that walk together with similar speeds must be different even when they share the same environment and goals. In addition, other levels of behaviors can exist when treating crowd in this hierarchical structure. The group behavior behaviors can be used to specify the way a group moves, behaves, and acts in order to fit different group structures (flocking, following, repulsion, attraction, etc). Individual abilities are also required in order to improve the autonomy and intelligence of crowd. However, to render thousands of individuals, these complex behaviors cannot be provided individually due to the hardware constraints and computational time rates. Other problems that arise in real-time time systems are on how to improve the intelligence and to provide autonomy to the scalable crowd. Furthermore, the simulation of large crowd in real time requires many instances of similar characters, that why an algorithms to allow for each individual in the crowd to be unique is needed. C. Project Background Thee Ancient Malacca Virtual Walkthrough [17] is a project that focuses on the modelling and visualization of Malacca city in 15th century. It is based on local and foreign

Figure 3 shows a screen capture of Ancient Malacca Virtual Walkthrough. Currently, there is no crowd simulation done to this walkthrough project. The challenge of this project is to bring the visualization of crowd simulation to the project with low computational cost. III.

CROWD RENDERING OPTIMIZATION METHODS

In recent years, researchers have applied several approaches, either separately or in combination, to develop crowd simulation in various graphics application. In this section, five of the crowd simulation approaches will wi be reviewed as shown in the Figure 4.

Crowd Rendering Methods

Cellular Automata

Social Force

Fluid Dynamics

Particles

Agent Based

Culling

Figure 4. These are crowd rendering methods frequently used in crowd simulation.

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real Real-Time Virtual Heritage Environment

16

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

A. Celular Automata Cellular automata were first proposed by Von Neumann [18]. Cellular automata are discrete dynamic systems consisting of a regular grid of cells. Cellular automata evolve at each discrete time step, with the value of the variable at one cell determined by the values of variables at the neighbouring cells. The variables at each cell are simultaneously updated based on the values of the variables in their neighbourhood at the previous time step and according to a set of local rules [19]. At present, cellular automata have been successfully applied to various complex systems, including traffic models and biological fields. In recent years, cellular automata models have been developed to study crowd evacuation under various situations. These models can be classified into two categories. The first one is based on the interactions between environments and pedestrians. The other is based on the interaction among pedestrians. B. Social Force Helbing and Molnar [20] introduce a ‘Social force model for pedestrian dynamics’. They suggest that the motion of pedestrians can be described as if they are subject to social forces which are a measure of the internal motivation of individuals to perform certain actions or movements. They describe three essential forces: 1.

2. 3.

Acceleration - the velocity of the pedestrian varies over time, as it attempts to reach its optimum speed and as it avoids obstacles. Repulsion - there is a repulsive force from other pedestrians and from obstacles and edges. Attraction - pedestrians are sometimes attracted by other people or by other objects.

Putting these three forces together, Helbing produces an equation for a pedestrian’s ‘total motivation’ and combining this with a term to allow for fluctuations in behaviour, produces the ‘social force model’. He goes on to describe computer models based on the equations, which have been successful in demonstrating various observed phenomena, for example lane formation. In Helbing et al. [21], the social force model is applied to the simulation of building escape panic, with impressive results. C. Fluids Dynamics Helbing et al. [22] have described that at medium and high densities, the motion of pedestrian crowds shows some striking analogies with the motion of fluids. For example, the footprints of pedestrians in snow look similar to streamlines of fluids or, again, the streams of pedestrians through standing crowds are analogous to riverbeds. Fluid-dynamic models describe how density and velocity change over time with the use of partial differential equations. Colombo and Rosini [23] presented a continuum model for pedestrian flow to describe typical features of this kind of flow such as some effects of panic. In particular, this model describes the possible over-compressions in a crowd and the fall in the outflow through a door of a panicking crowd jam. They

considered the situation where a group of people needs to leave a corridor through a door. If the maximal outflow allowed by the door is low, then the transition to panic in the crowd approaching the door may likely cause a dramatic reduction in the actual outflow, decreasing the outflow even more. D. Particles The majority of pedestrian simulations take this particulate approach, sometimes called the atomic approach. Early influential work was that of Craig Reynolds [5] who worked on simulations of flocks of birds, herds of land animals and schools of fish. Each particle, or boid, was implemented as an individual actor which navigates according to its own perception of the environment, the simulated laws of physics, and a simple set of behavioural patterns. Later work Reynolds [8] extends the concepts to the general idea of ‘autonomous characters’, with an emphasis on animation and games applications. Bouvier et al. [24] describe a generic particle model and apply it to both the problem of pedestrian simulation and to the apparently distinct problem of airbag deployment. They present software allowing the statistical simulation of the dynamic behaviour of a generic particle system. In their system, the particle system was defined in terms of: 1. The particle types - mass, lifetime, diffusion properties, charge, drag, interactions with surfaces, visualisation parameters. 2. The particle sources or generators - size, geometry, rate and direction of emission. 3. The 3D geometry, including obstacles. 4. The evolution of particles within the system. E. Agent Based Agent based are computational models [25] that build social structures from the ‘‘bottom-up’’, by simulating individuals with virtual agents, and creating emergent organizations out of the operation of rules that govern interactions among agents. Bonabeau [26] supported the following point of view: in agent terms, collective panic behaviour is an emergent phenomenon that results from relatively complex individual-level behaviour and interactions among individuals; the agent based seems ideally suited to provide valuable insights into the mechanisms and preconditions for panic and jamming by incoordination. In the last few years, the agent based technique has been used to study crowd evacuation in various situations. Agents based are generally more computationally expensive than cellular automata, social force, fluid-dynamic or particles models. However, their ability to allow each pedestrian to have unique behaviours makes it much easier to model heterogeneous humans, which are groups that contain individual with difference characteristic. F. Culling Culling can be defined as removing portion of scene that no contributes to final output [27][28]. They are collections of techniques that function to discard invisible and not required part of 3D objects to decrease the high computation

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

17

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

cost of real-time computer graphic application. For complex virtual heritage, culling will accelerate rendering process by quickly culling unwanted 3D objects before rendering process [16][27]. There two type culling suitable to be applied in crowd rendering: view frustum culling and occlusion culling. 1) View Frustum Culling There two approaches of view frustum culling (VFC): transform the bounding volume into perspective view [29] and check the bounding volume against view frustum volume by it six planes [30]. Both of them used for occur when object place outside of the view frustum [16]. In crowd simulation, crowds that were laid outside of the view frustum will be removed before the rendering process. The problem of view frustum is when rendering large data set which content many polygons inside frustum to be processed. Usually this view frustum usually be helped a spatial data structure. There are two types of spatial data structure, spatial partitioning and bounding volume hierarchies (BVHs) [31]. Hierarchical grids, portal, octrees, KD-trees are the most popular methods, and KD trees are usually considered the option of choice for massive models. Spatial data structure used attempted to pre-compute visibility relationships within a complex virtual scene [32][33][34]. Figure 5 show the number of object drawn in the screen with and without Range Detection Technique for view frustum culling be culled [16].

2) Occlusion Culling Occlusion culling removes 3D objects hidden by groups of other objects [16]. This technique is not suitable to be applied in rendering crowd simulation. Occlusion culling needs distinct organization of the whole geometry database. Also, occlusion culling is not appropriate for individual, deforming and animating objects because the scenes needed to be divided into smaller units or cells. Dynamic culling introduced as new occlusion culling based on binary tree merging with KD-tree [35]. This method used discretization and properties of urban scene for fast cull away invisible static object and also avatar. Figure 6 shows dynamic culling introduced to improve occlusion culling for crowd rendering [35].

(a)

(a) (b) Figure 6. Dynamic culling used in crowd rendering [15].

(b) Figure 5. The difference virtual walkthrough executed between with (a) and without (b) Range Detection Technique to improve view frustum culling [16].

G. Level of Detail Level of detail (LOD) involves reducing the complexity of a 3D object representation as it moves away from the 3D viewer in VE for faster rendering process [36]. This technique improves the efficiency of rendering by reducing the data structures on graphics pipeline stages, usually vertex transformations. LOD used to improve system performance and quality of graphic system [37]. LOD helps culling technique in reducing the polygon to be rendered. The criterion always be used in LOD is the size and quality of object relative to distant of the viewer.

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

18

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

15123 Triangles 6526 Triangles 4814 Triangles 814 Triangles Figure 7. Multi-resolutions from LOD for crowd rendering [38].

Figure 7 show LOD used for crowd rendering as the object became far, number of polygons reduced [38]. LOD is common techniques applied to crowd simulation to speed up rendering process. Aubel et al. [38] review that LOD is one of acceleration techniques that traditionally used in crowd rendering and automatically generate multiple LOD. Adaptive Level-of Detail for Human Animation (ALOHA) introduced to be applied LOD in behavior apart than geometry and motion [39]. Three rendering techniques applied with LOD [40]. Geometry, adapted layered imposter and flat imposter contain sub LOD to create a continuous hierarchy of LOD. LOD applied in KD-tree to improve existing crowd simulation [41]. Two components required in LOD: a method of approximating the desired result and a metric for determining how much approximation is allowed. IV.

FRAMEWORK

Propose of this research project is to add crowd simulation into the Ancient Malacca Virtual Walkthrough. Ulicny et al. [11] has proposed Crowdbrush as an approach to create complex scenes involving thousands of animated individuals in a simple and intuitive way. By employing a brush metaphor, analogous to the tools used in image manipulation programs, they can distribute, modify and control crowd members in real-time with immediate visual feedback. They define concepts of operators and instance properties that allow creating and managing variety in populations of virtual humans. An efficient technique allowing rendering up to several thousands of fully threedimensional polygonal characters with keyframed animations at interactive framerates is presented. Millan et al. [42] present another technique suitable to render large crowds of characters that takes advantage of existing programmable graphics hardware. Impostors are used for low-detail representation, while pseudo-instancing is used for higher detail. Their technique also used a LOD map to select between both representations, based on a customizable threshold. Figure 8 show the overview of our framework for crowd rendering in Ancient Malacca Virtual Walkthrough [16]. In the beginning, user will run the walkthrough using user interface provided. User interface is design according to functionality added to the system. It will display several

functions that can be choose by the user, they either can enable or disable view frustum culling. Database will store all the models and objects, either static or dynamic in database manager. Before rendering start, view frustum culling will be used to cull any unneeded objects from the walkthrough system. View frustum culling that will be used here is a new improved technique that is developed to reduce the computational cost of virtual heritage application. This technique is named as Range Detection Technique (RDT) [43]. It is based on the View Frustum Culling (VFC) method. The conventional VFC method tested the intersection of six planes using the plane equation to determine the visibility scope. Unlike conventional VFC technique, RDT is based on camera referential points and test the 3D objects whether they are in the viewing range or not. RDT execute the testing with the combination of bounding volume. Ancient Malacca Virtual Walkthrough

View Frustum Culling

Crowd Rendering

User Interface

Particle System Static Objects (Building)

Collision Detection

Dynamic Objects (Crowd)

Heightmap

Simulation Update

Figure 8. Crowd rendering framework.

The database manager will be called to render all objects after crowd rendering starts. A fragment program will read the character pixel buffer as a texture, and then will update the attributes for characters. The new position and heading will be then copied to a new character pixel buffer. Both character pixels buffers can be swapped after the rendering procedure has finished. This pass can also include the update of the animation frame for characters. This framework will utilized particle system technique for simulating crowd behavior in crowd rendering part. Particle system [44] will calculate and recalculate each character destination during the simulation. From this

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

19

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

calculation,, the crowd can move randomly throughout the Ancient Malacca terrain within the boundaries that has been set earlier in the particle system. Using collision detection, each character that collides with other character will turn around randomly into another direction and continue walking to a new destination that has been calculated in particle system. Collision detection and height-map map are added into the crowd rendering to add realism of the crowd movement. Using height-map, map, the crowd can move ups and downs along the terrain provided in the Ancient Malacca Virtual Walkthrough. Lastly, the system will update the simulation for user visualization. TESTING

In this section we describe testing result of crowd simulation system that we have done. We test our crowd problems using Horde3D rendering engine.

Numbers of character

V.

Figure 9 show our two screen shot from the crowd simulation system testing. In this test, we were comparing frame per second (FPS) of the system with a few character rendered compared to immense numbers num of character rendered. First we test the crowd simulation system with just 50 characters, and then we add more characters in addition of 50 for every test phase. We stop our test when the characters reach up to 500. 500 450 400 350 300 250 200 150 100 50 0

Crowd FPS

1 2 3 4 5 6 7 8 9 10 Numbers of try Figure 10. Graph numbers of character haracter in the crowd rendered vs. vs FPS.

Figure 10 show comparison between numbers of character rendered by the crowd simulation system with frame per second of the system. From the graph, we can conclude that more character rendered can produce higher computational mputational load thus lowering FPS of the system. (a) VI.

CONCLUSION

In this paper we have presented some of the related work in crowd simulation, crowd rendering issues, a few types of common crowd rendering methods, and an overview of the framework that will be used for developing crowd simulations in Ancient Malacca Virtual Walkthrough project. For the future work, there are rooms of improvement to the framework developed. As the framework will be using CPU for the entire processing task, we can also make use the t power of nowadays GPU to take part in the processing task. This will make the application not only rely with the CPU but also distribute the task to GPU. As conclusion, we hope that this research is useful for the virtual walkthrough system developer. Consequently, onsequently, it can also benefit the virtual heritage community. ACKNOWLEDGMENT (b) Figure 9. The difference between numbers of character rendered in crowd simulation system, with just 50 characters in (a) and 500 characters in (b).

We would like express our appreciation to Malaysian Ministry Of Science, Technology and Innovation under eScienceFund grant (01-01-06-SF0472) SF0472) for financial support of this research. arch. We also like to thanks Multimedia Development Corporation, Cyberjaya, Malaysia for allowing

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real Real-Time Virtual Heritage Environment

20

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

us to use the 3d model taken from Ancient Malacca project and permit us to explore the potential improvement of the project. REFERENCES [1] [2] [3]

[4] [5]

[6]

[7]

[8]

[9] [10]

[11]

[12]

[13] [14]

[15]

[16]

[17]

[18]

[19]

Tecchia, F., Loscos, C., & Chrysanthou, Y.: Visualizing Crowd in Real-Time. Computer Graphics Forums. (2002) Thompson P., Marchant E.: Testing and application of the computer model ’simulex.’ Fire Safety Journal 24, 2, 149–166 (1995) Dobbyn S., Hamill J., O’Conor K., O’Sullivan C.: Geopostors: A real-time geometry/impostor crowd rendering system. In SI3D ’05: Proceedings of the 2005 Symposium on Interactive 3D Graphics and Games (New York, NY, USA), ACM Press, pp. 95–102 (2005) Thalmann, D. and Musse, S. R.: Crowd Simulation. Springer-Verlag, London (2007) Reynolds C. W.: Flocks, herds, and schools: A distributed behavioural model. In Computer Graphics (ACM SIGGRAPH ’87 Conference Proceedings) (Anaheim, CA, USA), Vol. 21, ACM, pp. 25–34 (1987) Tu X., Terzopoulos D.: Artificial fishes: Physics, locomotion, perception, behaviour. In Computer Graphics (ACM SIGGRAPH ’94 Conference Proceedings) (Orlando, FL, USA), Vol. 28, ACM, pp. 43–50 (1994) Amkraut S., Girard M., Karl G.: Motion studies for a work in progress entitled ”Eurnythmy.” SIGGRAPH Video Review, 21. (second item, time code 3:58 to 7:35) (1985) Reynolds C. W.: Steering behaviours for autonomous characters. In Proceedings of Game Developers Conference 1999, pp. 763–782 (1999) Reynolds C. W.: Interaction with groups of autonomous characters. In Proc. Game Developers Conference ’00, pp. 449–460 (2000) Niederberger C., Gross M.: Hierarchical and heterogeneous reactive agents for real-time applications. Computer Graphics Forum 22, 3 (Proc. Eurographics’03) (2003) Ulicny B., de Heras Ciechomski P., Thalmann D.: Crowdbrush: Interactive authoring of real-time crowd scenes. In Proc. ACM SIGGRAPH/Eurographics Symposium on Computer Animation (SCA’04) (2004), pp. 243–252. Yersin B., Ma¨ım J., de Heras Ciechomski P., Schertenleib S., Thalmann D.: Steering a virtual crowd based on a semantically augmented navigation graph. In First International Workshop on Crowd Simulation (VCROWD’05) (2005). pp. 169-178 Treuille A., Cooper S., Popovic Z.: Continuum crowd. ACM Transactions on Graphics 25 (3), 1160–1168 (2006) Mao, T., Shu, B., Xu, W., Xia, S., Wang, Z.: CrowdViewer: from simple script to large-scale virtual crowd. In Proc. of the 2007 ACM Symposium on Virtual Reality Software and Technology. (2007), pp. 113-116. Berg, J. v. d., Patil, S., Sewall, J., Manocha, D., Lin, M.: Interactive Navigation of Multiple Agents in Crowded Environments. In Proc. of the 2008 Symposium on Interactive 3D Graphics and Games. (2008), pp. 139-147. Azahar. M.AM., Sunar. M.S, Daman. D, Bade. A, “Real-Time Crowd Simulations for Interactive Virtual Environments,” in Proc. of Computer Games & Allied Technology in Animation, Multimedia, IPTV & Entertainment, Singapore, pp. 52 - 59, 28-30 (2008) Sunar, M. S., A. M. Zin , "Improved View Frustum Culling Technique for Real-Time Virtual Heritage Application." IJVR 7(3): 43-48 (2008) Xiaoping, Z., Tingkuan, Z., Mengting, L.: Modeling crowd evacuation of a building based on seven methodological approaches. In: Building and Environment, 44 (2009), pp. 437–445 Wolfram S.: Statistical mechanics of cellular automata. In: Reviews of Modern Physics 55:601–44 (1983)

[20] Helbing D, Molnar P.: Social force model for pedestrian dynamics. In: Physical Review E 51(5):4282–6 (1995) [21] Helbing D, Farkas I, Vicsek T.: Simulating dynamical features of escape panic. In: Nature 407:487–90 (2000) [22] Helbing D, Farkas IJ, Molnar P, Vicsek T.: Simulation of pedestrian crowds in normal and evacuation situations. In: Schreckenberg M, Sharma SD, editors. Pedestrian and evacuation dynamics. Berlin: Springer; p. 21–58 (2002) [23] Colombo RM, Rosini MD.: Pedestrian flows and non-classical shocks. In: Mathematical Methods in the Applied Sciences 28:1553– 67 (2005) [24] Bouvier E., Cohen E., Najman L.: From crowd simulation to airbag deployment: Particle systems, a new paradigm of simulation. Journal of Electrical Imaging, 6 (1), 94–107 (1997) [25] Goldstone RL, Janssen MA.: Computational models of collective behavior. In: Trends in Cognitive Sciences 9(9):424–30 (2005) [26] Bonabeau E.: Agent-based modeling: methods and techniques for simulatinghuman systems. In: Proceedings of the National Academy of Sciences of the USA(PNAS) 99(Suppl. 3):7280–7 (2002) [27] Tomas, A.-M., M. Tomas, (2008). Real-Time Rendering 3rd Edition, A. K. Peters, Ltd. [28] Sunar, M. S., A. M. Zin , "Visual Culling Techniques for Geometric Processing in Walkthrough Application." Jurnal of Computational Information Systems 2(1): 421-427 (2006) [29] L. Bishop, D. Eberly, T. Whitted, M. Finch and M. Shantz. Designing a PC Game Engine, Computer Graphics in Entertainment, pp. 46-53, January/February 1998 2 [30] Ned, G. (1994). Detecting Intersection of a Rectangular Solid and a Convex Polyhedron. Graphics gems IV, Academic Press Professional, Inc.: 74-82. [31] Gobbetti, E., Kasik, D. & Yoon, S., Technical strategies for massive model visualization, Proceedings of the 2008 ACM symposium on Solid and physical modeling, ACM, (2008), pp. 405-415 [32] Airey, John M., Rohlf, John H., and Brooks, Jr., Frederick P. Towards Image Realism with Interactive Update Rates in Complex Virtual Building Environments. ACM SIGGRAPH Special Issue on 1990 Symposium on Interactive 3D Graphics, 24, 2 (1990), pp. 41-50 [33] Seth, J. T., H. S. Carlo, (1991). "Visibility preprocessing for interactive walkthroughs." SIGGRAPH Comput. Graph. 25(4): 61-70. [34] D. Luebke and C. Georges (1995). Portals and Mirrors: Simple, Fast Evaluation of Potentially Visible Sets. Proceedings of the 1995 symposium on Interactive 3D graphics. Monterey, California, United States, ACM. pp. 105-106 [35] F. Tecchia, C. Loscos, and Y. Chrysanthou. Real Time rendering of Populated Urban Environments. In ACM Siggraph Technical Sketch, August 2001. [36] Luebke D.,Reddy M., Cohen J, Varshney A, Watson B., and Huebner R., 2000, Level of Detail for 3D Graphics, Morgan Kaufmann, USA. [37] Reddy, M. (1998). "Specification and Evaluation of Level of Detail Selection Criteria." Virtual Reality 3(2): 132-143. [38] Aubel, A., R. Boulic (2000). "Real-time display of virtual humans: levels of details and impostors." Circuits and Systems for Video Technology, IEEE Transactions on 10(2): 207-217. [39] C. O’Sullivan, J. Cassell, H. Vilhjalmsson, J. Dingliana, S. Dobbyn, B. McNamee, C. Peters,and T. Giang. 2002. Levels of detail for crowds and groups. Computer Graphics Forum, 21(4). [40] Coic, J.-M., C. Loscos, (2005). Three LOD for the Realistic and Real-Time Rendering of Crowds with Dynamic Lighting, LIRIS UMR 5205 CNRS/INSA de Lyon/Université Claude Bernard Lyon 1/Université Lumière Lyon 2/Ecole Centrale de Lyon. [41] Stephen, J. G., L. Ming, (2008). Level of Detail Techniques for Accelerating Crowd Simulations [42] Millan, E., & Rudomin, I.: Impostors and pseudo-instancing for GPU crowd rendering. In: Proceedings of the 4th international

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

21

electronic Journal of Computer Science and Information Technology (eJCSIT), Vol. 2, No. 1, 2010

conference on Computer graphics and interactive techniques in Australasia and Southeast Asia, 49-55 (2006) [43] Azahar, M. A. M., Sunar, M. S., Bade, A., and Daman, D. “Crowd Simulation for Ancient Malacca Virtual Walkthrough”. The 4th International Conference on Information & Communication Technology and Systems. Institut Teknologi Sepuluh Nopember (ITS), Surabaya, Indonesia, 511 - 516. (2008)

[44] Azahar, M. A. M., Sunar, M. S., Daman, D., Bade, A., "Survey on Real-Time Crowds Simulation," in Proc of the 3rd International Conference on E-Learning and Games, Nanjing, China, SpringerVerlag Berlin Heidelberg, Edutainment 2008, LNCS 5093, pp. 573– 580, 2008

M. A. M. Azahar, M. S. Sunar and D. Daman, Low Computational Cost Crowd Rendering Method for Real-Time Virtual Heritage Environment

22