IJRECE VOL.

7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

3D Game based on FPS using Artificial Intelligence

Sanjay Kumar S1, Sagar T Y2, Jayanth C3, Amruta T Bhat4, Prof. Vinayashree5
12345
Vidya Vikas Institute of Engineering and Technology Mysuru,India
list and closed list. Open list: It contains nodes we have
Abstract- A game is a structured form of play. A new solution
for the building a first- person shooter game is by using cross discovered as neighbours during the search but have not yet
platform unity’s game engine. This project uses unity’s processed. It selects one node to process from the open list
mecanium system in conjunction with root motion animation with each iteration of the search loop and remove it from the
and navigation mesh based pathfinding mechanism, and some open list once processed. In the first loop iteration, the open
advanced methods like building the environment using list contains only the start node.
navigation and pathfinding. It also uses animation state Closed list: As A* selects nodes from the open list, it
machines to control the animations of players and non-player removes them from the open list and it adds to the closed list.
characters. Player finds the path to the target in game scene. This allows to keep track of all the nodes that A* have
Another solution is A* algorithm is most widely used to already processed so it won't try to process them again (to
estimate the distance of any point to the target point. 3- avoid infinite loop). The closed list will start off completely
Dimensional rendering of the game objects is the interesting empty.
and attracting key of this project. Open list < 31, 51, 62, 63, 22, 23, 24, 50, 61, 21, 63, 65, 73,
74, 45, 56, 66 >
Keywords- unity’s mecanim, pathfinding, a*algorithm, state Closed list < 42, 43, 53, 33, 41, 52, 32, 64, 55, 46 >
machines, rendering.
I. INTRODUCTION
An understanding of graphs and algorithm is an integral part
of the process. Establishing a clear understanding of object
and how various algorithms affect those objects is crucial for
anyone interested in learning programming. An algorithm is a
procedure of operations for correctly solving a problem in a
finite number of steps, where operations in each step can be
executed in sequential order or in parallel. A graph algorithm,
in turn, is an algorithm for solving problems that can be
formulated in terms of graph and its related data structures.
II. A* ALGORITHM
A* algorithm is a shortest path heuristic search algorithm, and
it is based on “Dijkstra” algorithm which finds the shortest
paths from the source vertex to all other vertices in the graph.
A heuristic is used to determine which neighbour to search
next. This significantly speeds up the search by trying to visit
as few nodes as possible. It does this by scoring neighbours as
they are encountered based on their likeliness to lay on the
cheapest path. It will always find the cheapest path .
Fig.2: A* path finding

For Navigation and Path-finding problem, we have used the

Manhattan Method of estimation to estimate the cost. The
function of current node n is expressed as follows:
F(n)=H(n)+G(n)
Where, F(n) is a cost estimated function, H(n) is the heuristic
value of the shortest path of any node n to the target point,
G(n) is the path which is shortest and also the initial point to
any node n.

Processing of a Neighbour in A* search

 In the iteration of search there is a calculation of score for
the A* terminology and it is called as fScore of the node.
 A* will set its parent node reference to the current node
that is going to be processed.
Fig.1: Navigation graph from a grid  If Neighbour already exists in Closed list the iteration
loop will skip that Neighbour and does not add to Open
list.
A* search is a loop which maintains two lists namely open
INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING
A UNIT OF I2OR 1153 | P a g e
 IJRECE VOL. 7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

 If Neighbour is not in Open or Closed lists the search

iteration will calculate and store fScore in Neighbour. IV. IMPLEMENTATION
And it assigns the reference of Parent Neighbour to A. STATE MACHINES
Current node, then the current Neighbour is added to A device which can be in one of a set number of condition
Open list. dependency on its previous condition and the present value of
 If Neighbour is already in Open list, then the new fScore its input. It starts entry next node only if the given condition is
will be calculated for the Neighbour factoring in current true otherwise it will stay in their current node. Finite state
node. machines are one of the most effective and most frequently
 If new fScore is lower than fScore already stored in the used methods of programming artificial intelligence. Each
Neighbour, then the current fScore will be replaced by state possesses code responsible for the initialization and de-
the new fScore. Also, Parent reference will be updated for initialization of the object. The finite state machine method
the current node. lets us easily divide the implementation of each game object's
behavior into smaller fragments. A long animation
III. UNITY 3D implemented in the update method may be controlled by a
Unity 3D is an incredibly powerful and versatile game and state machine to break the animation sequence into sequential
interactive-experience development tool created by Unity pieces so that draws will not be blocked for long period of
Technologies in 2005. Trendy game engines like unity’s time.
mecanim engine are unbelievably wealthy and powerful
cross-platform. The core game engine includes a rendering
engine for 2D and 3D graphics, a physics engine or collision
detection, sound, scripting advanced animations, audio
support, video support, and powerful UI-development tools.
Any animation designed for Unity 3D must work with the
logic of the sport loop. The Unity 3D game loop consists of a
variable-time update and fixed-time update. Unity 3D
scripting supports co-routines, that permits referred to as
strategies, that permits referred to as strategies to possess
multiple entries interleaved with the Unity 3D system. A
package in Unity 3D may be an instrumentation of scenes and
numerous assets like scripts, models, images, and sound
effects employed by the sport objects inside those scenes.

Fig.4: State machine

B. SCENE BUILDING, NAVIGATION AND PATH

FINDING
A Scene contains the environments and menus of your game.
Contemplate each distinctive Scene file as a singular level. In
each Scene, you place your environments, obstacles, and
decorations, essentially turning out with and building your
game in things.

Fig.3: CSI interpreter design Different aspects of development

process

 Rendering: Unity’s graphics engine uses its own APIs

like OpenGL, direct3D and also supports other file
formats from other software’s like Blender Adobe
Photoshop etc.
 Scripting: Programmers write Unity Script similar to Fig.5: A new empty Scene, with the default 3D objects
JavaScript on mono-an open source platform for .net
framework. a. Navigation mesh
 Physics: Provides built-in support for PhysX [10] physics NavMesh (short for Navigation Mesh) is a data structure which
engine with real time simulations on skinned meshes, describes the walkable surfaces of the game world and allows
thick ray casts etc. to find path from one walkable location to another in the game
INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING
A UNIT OF I2OR 1154 | P a g e
 IJRECE VOL. 7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

world. The data structure is built, or baked, automatically from deterrents as opposed to simply different specialists. These
your level geometry. could be the standard containers and barrels in a shooter
amusement, or vehicles. The obstructions can be taken care of
utilizing neighbourhood hindrance evasion or global
pathfinding.
When associate obstacle is moving, it's best handled
exploitation native obstacles shunning. this fashion the agent
will predictively avoid the obstacle. once the obstacle becomes
stationary, and might be thought-about to dam the trail of all
agents, the obstacles ought to have an effect on the worldwide
navigation, that is, the navigation mesh.
dynamic objects. Multiple agents and other dynamic objects
also need to be able to happily co-exists. As the navigation
graph is compiled at development time, path searches do not
take into account objects which can moves at run time. A local
avoidance system is employed to push dynamic objects always
away from each other during agent steering. Unity can alter the
navigation mesh at run time by carving objects into it. This is
useful for when we know a dynamic object will never move
again.
Fig.6: Game scene with NavMesh

NavMesh (short for Navigation Mesh) is a data structure which

describes the walkable surfaces of the game world and allows
to find path from one walkable location to another in the game
world. The data structure is built, or baked, automatically from
your level geometry.
NavMesh Agent component help you to create characters
which avoid each other while moving towards their goal.
Agents reason about the game world using the NavMesh and Fig.8: Local avoidance system
they know how to avoid each other as well as moving
obstacles. c. Path corridor
A path search in unity returns to the agent a corridor. A
corridor is a list of polygons that must be traversed. Corridors
are useful for supplying the agent with surrounding
information so run-time path diversions can be safely
computed. It is actually the vertices of the corridor that form
the waypoint list the agent must pass through.

Fig.9: Run-time path diversions using Path corridor.

C. CHARACTER DESIGN
Changing the NavMesh is termed carving. the method detects
Fig.7: Avoiding obstacles in navigation
that elements of the obstacle touches the NavMesh and carves
holes into the NavMesh. this can be computationally
The controlling rationale takes the situation of the following
overpriced operation, that is one more compelling reason,
corner and dependent on that makes sense of an ideal heading
why moving obstacles ought to be handled exploitation
and speed (or speed) expected to achieve the goal. Utilizing the
collision shunning.
ideal speed to move the specialist can prompt impact with
different operators.
b. Local avoidance system
Hindrance evasion picks another speed which adjusts between
As other dynamic objects will not be represented in the
moving in the ideal bearing and avoiding future crashes with
navigation graph a separate system must exist that moderates
different specialists and edges of the route work. Solidarity is
the queried path to perform on-the-fly adjustment to the
utilizing complementary speed deterrents (RVO) to anticipate
agent’s velocity to avoid nearby dynamic objects. Only static
and counteract impacts.
objects are compiled into the navigation graph. Without the
Numerous utilizations of navigation require different kinds of
addition of a local avoidance system for dynamic objects,
INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING
A UNIT OF I2OR 1155 | P a g e
 IJRECE VOL. 7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

agents may collide with each other or other When making a modifier, every bone can should be adjusted with envelopes
personality for a game there's typically a predefined work flow and weights to attach the proper vertices to the bone for that
or pipeline which has plenty of steps. a number of the foremost specific part. rather than creating the character style method
common steps embody idea style, base modelling, high poly manually, there are tools to come up with a completely rigged
modelling, low poly modelling, unwrap, texturing and and abraded low two- dimensional figure character with
rigging/skinning. the primary step of making a personality is to corresponding maps. this may provide artists an extremely
come back up with an idea. an idea may be a painting or high freedom of exploitation the creations from build Human
sketch, illustrating the specified look and theme of the finished in any manner imaginary.
character.
The next step is base modelling. during this step, the bottom
model ought to include a rough model, with the proportions
and shapes fairly correct whereas still keeping the pure
mathematics straightforward. This step is sometimes wiped out
a 3D special effects software package like liquidizer and
Autodesk 3ds GHB. liquidizer could be a free and open supply
software package used for modelling and making animated
films, visual effects, interactive 3D applications or video
games. once the bottom model has been created, a high poly
model is to be created. this will be done by sculpting the

Fig.11: Rigging, Skinning and humanoid avatars

D. MECANIM
bottom model to feature additional detail to the character, for An important a part of 3D game development is animations,
example wrinkles. thanks to the introduction of high level and additional specifically character animations. 2 of the best
details, this step may be terribly long. The sculpting method difficulties once desegregation the animations in a very game
may be delineated as shaping a coffee plane figure virtual clay engine are generating diagrammatically swish transitions
mesh into a high polygon elaborated clay mesh. between different animations and additionally the advanced
Fig.10: A high polygon model scripting that's needed for the various animation state
transitions. However, Unity recently free Unity four. Unity
Rigging and Skinning four comes with the new animation system referred to as
When the look and modelling of the character is complete, the Mecanim.
character must be rigged and abraded so as to be able to use Mecanim provides the user practicality to setup and
with animations within a game engine. Character animation is management animations for a personality within the Unity
that the most advanced sort of animation. not like a tree, a editor. There are 2 styles of character varieties supported by
personality not solely must be able to move, it conjointly Mecanim, particularly mechanical man and generic characters.
should be able to categorical itself and show emotions. To rig A mechanical man may be a character that resembles the looks
the character, a skeleton designed along by completely of a person's being whereas a generic character has no
different bones must be additional to the character. The bones predefined bone structure, center of mass or orientation. If a
must be connected to the character employing a skin modifier. generic character is to be used with Mecanim, a number of the
This method is to make sure that the pure mathematics close elementary options provided by Mecanim won't be on the
the bones deforms properly once the skeleton moves. In market. a straightforward example of a generic character may
liquidizer a skeleton might either be created by with either a be a dog. once a mechanical man character is to be foreign to
separate bone affiliation approach or with Blender’s Unity and used with Mecanim, Mecanim creates Associate in
equivalence of Biped, the Rigify system. once a skeleton has Nursing Avatar. Associate in Nursing Avatar is an interface
been created and placed so as to suit with the proportions and that interprets the characters bone structure to the bone
anatomy of the character, the bones should be abraded to the structure understood by Mecanim. Fig.9 shows a translation
character. For skinning a Skin modifier is employed with this mapping between the Mecanim bone structure and therefore
INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING
A UNIT OF I2OR 1156 | P a g e
 IJRECE VOL. 7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

the actual bones within the character.

INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING

A UNIT OF I2OR 1157 | P a g e
 IJRECE VOL. 7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

different animation states and the transitions between

them.
Blend Tree
Another powerful feature of the Animator part is that the mix
Trees. a mix Tree may be created within a state to
interchange this animation clip therein state. for instance, the
‘locomotion’ state’s walking motion clip may be replaced by
a mix Tree, during which the recently created mix Tree might
then embody a multiple variety of motion clips i.e., ‘walk’,
and ’Run’. once the ‘locomotion’ state is active, the mix Tree
plays one among its motion clips betting on a user outlined
variable. during this example such a variable might represent
the speed of the run so as to let the character walk or run
consequently. If the worth of the speed variable is lesser than
given threshold value it'll walk or if its larger than threshold
value then it will begin running and also the animations will
mix into a dynamic animation that interpolates between the 2
motions. This feature ends up in sleek and dynamic transition
motions between completely different animation clips.

Fig.12: Avatar Configuration, mapping the character’s bone

structure to Mecanim’s bone structure.

When a character has been successfully created with an

Avatar, Mecanim allows the user to tell the game engine
how and when different character animation clips should be
played. A powerful tool within Mecanim is the Animator
component.

E. ANIMATOR
In the Animator part, completely different states and layers
are outlined and controlled. A state within the Animator part
corresponds to the precise set of actions and therefore the
surroundings of the character in the game. for instance, a
state outlined as ’Idle’ might correspond to a situation within Fig.14: Blend Tree including three different animation clips.
the game wherever the character isn't moving in any respect.
during this specific state, a whorled associate degree Imation V. CONCLUSION
clip of an idle motion may be applied to the state, forcing the In this paper, we have discussed various methods and
character to play that animation whereas within the ’Idle’ algorithm that are being used in the field of first person
state. so as for the character to perform another action of shooter games. However, the combination of various
movement, another state should be outlined. for instance, if evolutionary algorithm along with the finite state machine
the character is meant to run, a state known as ’Running’ mechanisms have proven to be useful in generating AI bots
may be created, holding a clip of a whorled running motion. for first person shooter.
Also, transitions between the 2 states should be outlined, First person shooters are new phase in gaming world. With
permitting the character to modify between the idle motion their advanced technology and sophisticated environment,
and a running motion. A transition is controlled by one or a FP shooter gives a player thrilling experience. In every
lot of conditions that need to be met so as to modify aspect it offers a great excitement with fast pace while
between the states. A condition is controlled by a user playing. They give you the complete freedom to configure
outlined variable that successively is controlled by a script. everything just the way you want it.

VI. REFERENCES
[1]. Geldenhuys, T. (2013) C# Interpreter Console for Unity 3D.
[ONLINE] <http://blog.tiaan.comilink!201
0/03115/csharpinterpreter-unity-plugin-console-debugger>.
[2]. Unity Asset Store: https://www.assetstore.unity3d.com.
[3]. Unity 3D: http://unity3d.com
[4]. S. M. LaValle, Planning algorithms, Cambridge University
Press, 2006. [2] T. K. Whangbo, “Efficient Modified
Bidirectional A * Algorithm for Optimal Route- Finding,”
Fig.13: Animator component, showing an example of New Trends in Applied Artificial Intelligence, Japan, vol.

INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING

A UNIT OF I2OR 1158 | P a g e
 IJRECE VOL. 7 ISSUE 2 (APRIL- JUNE 2019) ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE)

4570, pp. 344–353, June 2007.

[5]. Y. Lang, “Research on collision detection method in unity,”
Software Guide, China, vol. 13, pp. 24–25, July 2014.
[6]. Kennedy J, Eberhart R C. Particle swarm optimization[A].In:
Proceedings of IEEE International conference on Neural
Networks[C] . Perth, Australia: [s. n.], 1995. 1942- 1948.
[7]. Fuellerer G, Doemer K F, Hartl R F, et al. Metaheuristics for
vehicle routing problems with three- dimensional loading
constraints[J]. European Journal of Operational Research,
2010, 201(3): 751-759.
[8]. McFarlane, A., Sparrowhawk, A. & Heald, y. 2013. Report on
the educational use of games. [ONLINE] Available at:
http://www . pewinternet.org/-/media/ /F ileslReportsl2008/PI
P _Teens_ Games_and_ Civics_Report_FINAL.
pdf. pdf. [Accessed 11 April 2013]. affiliations as succinct as
possible (for example, do not differentiate among departments
of the same organization).
[9]. Ismail Buyuksalih, Serdar Bayburt, Gurcan
Buyuksalih, A P Baskaraca, Hairi KAlias AbdulRahman.#d
modeling and visualization based on the Unity game engine-
advantages and challenges.4th International GeoAdvance
Workshop,2017.

INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING

A UNIT OF I2OR 1159 | P a g e