Difference between revisions of "Project3F20"

From Immersive Visualization Lab Wiki
Jump to: navigation, search
(4. Ride Design (25 Points))
 
(100 intermediate revisions by 3 users not shown)
Line 1: Line 1:
=UNDER CONSTRUCTION - DO NOT START YET=
+
=Project 3: Amusement Park Ride=
  
 +
In this project you will need to implement a scene graph to render an amusement park ride which consists of a number of objects that are part of a hierarchy.
  
=Project 3: Hierarchical Modeling=
+
This project implements the following skills from the lectures:
  
In this project you will need to implement a scene graph to render a an animated scene which consists of a number of objects that are part of a hierarchy.
+
* texturing
 +
* sky box
 +
* environment mapping
 +
* scene graph
  
You can pick one of the following types of things to animate:
+
The user of your program should be able to control a variety of options within your simulation using the keyboard. You may choose which keys control which functions, but make sure they are clearly documented in either a text file named <tt>README.txt</tt> which you add to the main directory of your code, or in your video (narration or text overlays). The following interactive controls should be available to the user:
  
* a solar system
+
* Start/stop the animation of each of the levels of your motion hierarchy separately.
* an amusement park ride
+
* Move around within the simulation (by moving the camera): forward, back, up or down, as well as turn the camera left or right (which affects subsequent forward movements).
* an animal or person moving in a coordinated way (no robots!)
+
 +
The total score for this project is 100 points. Additionally, you can obtain up to 10 points of extra credit.
  
The user of your program should be able to control a variety of options within your simulation using the keyboard. You may choose which keys control which functions, but make sure they are clearly documented in a text file named README.txt which you add to the main directory of your code.
+
There will '''not''' be a discussion on November 11th because of Veterans Day. Instead, this project will be introduced in lecture on Tuesday, November 10th.
  
Allow the user the following controls
+
We recommend starting with your code from project 2 and removing the rendering of the 3D models. In this project you are going to render all new objects, using your own implementation of a scene graph.
  
* start/stop the animation
+
==1. Sky Box (20 Points)==
* reset the view of the simulation
+
* move around within the simulation, i.e., move left/right/up/down, turn left/right
+
+
The total score for this project is 100 points. Additionally, you can obtain up to 10 points of extra credit.
+
  
Start with your code from project 2 and remove the rendering of the 3D models. In this project you are going to render all new objects, using your own implementation of a scene graph.
+
Create a sky box for your scene, to give it a nice background. [http://learnopengl.com/#!Advanced-OpenGL/Cubemaps LearnOpenGL] offers a great tutorial for sky boxes and how to implement them in modern OpenGL.
  
==Sky Box (Points)==
+
Make sure the width and height of your sky box textures are powers of two, we recommend 512x512 pixels. You are allowed to create your own sky box if you want, many smartphones can capture 360 degree panorama images which you would need to turn into a cube map. If you would rather use ready-made sky boxes from the internet, here a few pointers:
  
Create a sky box for your scene. A sky box is a large, square box which is shown around your entire scene, to give it a nice background. The camera is located inside of this box. The inside walls of the box usually consist of pictures of a sky and a horizon, as well as the terrain below. Sky boxes are normally cubic, which means that they consist of six square textures for the six sides of a cube. [http://learnopengl.com/#!Advanced-OpenGL/Cubemaps Here] is a great tutorial for sky boxes and how to implement them in modern OpenGL.
+
* http://www.f-lohmueller.de/pov_tut/skyboxer/skyboxer_3.htm ([https://drive.google.com/file/d/1G-lL_eQXxz17ezCIAg5vnE_jHwN2VBl4/view?usp=sharing offline link])
 +
* https://opengameart.org/content/elyvisions-skyboxes ([https://drive.google.com/file/d/1O3WOxZPl2aDQlyDfYnbY42qkFaWGF2Ek/view?usp=sharing offline link])
 +
* [https://www.cleanpng.com/free/skybox.html These are lower resolution and need to be split up into the individual cube faces.]
 +
* [https://hdrihaven.com/hdris/ These are very nice but need to be re-projected into cube maps (see below).]
  
[http://www.f-lohmueller.de/pov_tut/skyboxer/skyboxer_3.htm Here is is a nice collection of textures for sky boxes], and [http://www.custommapmakers.org/skyboxes.php here is an even bigger one]. Make sure the width and height of your sky box textures are powers of two, such as 512x512 or 1024x1024. You are allowed to create your own sky box if you want.
+
If you created or downloaded a panorama image from the web, you need to convert it into the six cube map faces you need for the sky box. [https://jaxry.github.io/panorama-to-cubemap/ This web tool can do this for you.]
  
Choose a sky box texture. Then create a cubic sky box and make it very large by giving it coordinates from -500 to +500. Check to see if the textures align correctly at the edges - if not you may need to rotate some of the textures. There should not be any visible seams.
+
Once you created or downloaded your sky box textures, use them to create a sky box by texturing the insides of a very large cube (coordinates from about -500 to +500). Make sure that the textures align correctly at the edges - if not you may need to rotate or flip some of the textures.  
  
Make sure single-sided rendering (backface culling) is enabled. If your sky box is defined with the triangles' normals facing outward (like the example in the discussion slides), use the following code lines to ensure that the user will never see the outside of the box:
+
Use single-sided rendering, so that OpenGL will automatically hide the user-facing sides of the cube. If your sky box is defined with the triangles' normals facing outward (like the example in the discussion slides), use the following code lines to ensure that the user will never see the outside of the box:
  
 
<pre>
 
<pre>
Line 62: Line 66:
 
* -3 if backface culling is not enabled, or it is enabled but the wrong faces are culled
 
* -3 if backface culling is not enabled, or it is enabled but the wrong faces are culled
  
==Environment Mapping (20 Points)==
+
==2. Environment Mapping (20 Points)==
  
Use the sphere code you used for the bounding spheres in project 3 to render a solid sphere. Make sure you are parsing for the vertex normals in addition to the vertices. Its surface should look like polished metal, so you need to use environment mapping on it.
+
Create an algorithm to tessellate a disco ball style sphere (ie, create vertices and normals for a sphere with a relatively low number of quads for faces). Its surface should look like polished metal, so you need to use environment mapping on it.
 +
 
 +
This sphere needs to become part of your amusement park ride later, but we recommend implementing it beforehand so that you can test it out separately. Unless you don't get to render the ride, you don't need to show this sphere separately in your demo (although you could).
 +
 
 +
We provide code for the sphere geometry [https://github.com/cbutarbu/Sphere_167 for Windows] and [https://github.com/cbutarbu/Sphere_167/tree/macOS for Mac]. But feel free to write your own.  
  
 
The [http://learnopengl.com/#!Advanced-OpenGL/Cubemaps Cube map tutorial] explains how environment mapping is done with a sky box as the environment. Feel free to follow it and use the code from the site to make it work in your code. The goal is to have the sky box reflect on the sphere, to give the impression of the sphere's surface being polished metal (i.e., act like a perfect mirror).
 
The [http://learnopengl.com/#!Advanced-OpenGL/Cubemaps Cube map tutorial] explains how environment mapping is done with a sky box as the environment. Feel free to follow it and use the code from the site to make it work in your code. The goal is to have the sky box reflect on the sphere, to give the impression of the sphere's surface being polished metal (i.e., act like a perfect mirror).
  
Use the camera controls from project 2, which allow the user to rotate the view from a fixed view point, as well as zoom in. It is important that your camera controls allow you to zoom in on the sphere, so that you can check to see if your environment mapping algorithm works correctly.
+
Here is an example of a sphere with environment mapping in a sky box:
 +
 
 +
[[Image:DiscoballF20.jpg|x200px]]
  
 
'''Grading:'''
 
'''Grading:'''
 
* -10 if reflection looks somewhat like the skybox but is off by a lot
 
* -10 if reflection looks somewhat like the skybox but is off by a lot
* -5  if reflection is upside down but code looks good
+
* -5  if reflection is upside down but otherwise correct
  
==Scene Graph Engine (Points)==
+
==3. Scene Graph Engine (25 Points)==
  
To create a robot with multiple moving body parts (head, torso, limbs, eyes, antennae), we need to first implement a simple scene graph structure for our rendering engine. This scene graph should consist of at least three nodes: <tt>Node</tt> (5 points), <tt>Transform</tt> (10 points) and <tt>Geometry</tt> (10 points). You are free to add more scene graph node types as you see fit.
+
To create a ride with multiple moving parts, we need to first implement a simple scene graph structure for our rendering engine. This scene graph should consist of at least three nodes: <tt>Node</tt> (5 points), <tt>Transform</tt> (10 points) and <tt>Geometry</tt> (10 points). You are free to add more scene graph node types as you see fit.
  
 
* Class <tt>Node</tt> should be abstract and serve as the common base class. It should implement the following class methods:
 
* Class <tt>Node</tt> should be abstract and serve as the common base class. It should implement the following class methods:
Line 92: Line 102:
 
** have a class method which draws the 3D model associated with this node.
 
** have a class method which draws the 3D model associated with this node.
  
==Robot (Points)==
+
==4. Ride Design (25 Points)==
  
Now that we have the scene graph classes, it is time to put them to work. Build your own robot using the <tt>addChild</tt> methods. Use at least 3 different types of parts for your robot (e.g., body, head and limb). In total, your robot needs to consist of at least 4 parts. (15 points)
+
Now that we have the scene graph classes, it is time to put them to work. Build your own amusement park ride using the <tt>addChild</tt> methods.
  
Start with code that uses your trackball code, and modify it so that trackball rotations control the camera's direction vector instead - the camera should stay in place. (If you don't have trackball rotation code, keyboard controls will suffice, but the camera still needs to be stationary and pivot about its location.) Also allow zooming in and out (i.e., changing the field of view). (5 points)
+
Your design needs to include the following features:
  
Thanks to our tutors Weichen and former tutor Yining, you have the following robot components to choose from: head, body (torso), limb, eye, antenna. You will find the OBJ files in [[Media:Robot-parts-2018.zip |this ZIP file]]. In these OBJ files, each vertex has not only a 3D coordinate and a normal associated with it, but also a texture coordinate - you can ignore the latter. Note that '''unlike the previous obj files''' in the course, each face has different indices for v/vt/vn. (You many want to check [http://en.wikipedia.org/wiki/Wavefront_.obj_file Wikipedia] for more information about *.obj format) One way to deal with the different indices is to re-order (and duplicate) the v//vn data when parsing so that their indices align. The following code might be helpful for this:
+
* You need to have a ground plane for the ride.
<pre>
+
* The ride must be anchored to the ground plane (can't be floating in space).  
// Assume that you parse indices of v//vn into different std::vector (vertex_indices_, normal_indices_)
+
* Your disco ball sphere must be part of the design. It can occur once or multiple times.  
// input_vertices and input_normals are raw input data from *.obj files
+
* Your hierarchy needs to be at least '''three''' levels deep. For instance a ferris wheel which spins around its center with cars that rotate so they are always aligned with gravity, with the cars rotating about their vertical axis (see first picture below).  
// vertices_ and normals_ are aligned data
+
* Drawing the same object multiple times in different positions, orientations, and/or with different scaling factors.
 +
* Moving objects over time, including rotations not centered at the origin, linear motion (changing direction at the end points) and compound motions.
 +
* Use regular Phong shading as in project 2 for the surfaces of all your ride geometries. This means that you need to have correct normals for all your 3D objects.  
 +
* Implement at least one directional or point light source for your scene, and make sure that all surfaces are illuminated correctly. (Ideally, you use your lighting environment from project 2 which would also allow you to move your light source around with the mouse, but this is not a requirement.)
  
for (unsigned i = 0; i < vertex_indices_.size(); i++) {
+
Adequate results can be achieved with simple cubes, cones, cylinders and spheres. More complex objects are not required for this simulation, but you can get extra credit for them. Below are a few links to basic 3D shapes you can use. You are also allowed to use the OBJ files from projects 1 and 2.
  vertices_.push_back(input_vertices[vertex_indices_[i]]);
+
  normals_.push_back(input_normals[normal_indices_[i]]);
+
  indices_.push_back(i);
+
}
+
</pre>
+
  
Here is an example of a robot with two antennas, two eyeballs, one head, one torso and 4 limbs (2 legs and 2 arms):
+
* [[Media:CubeOBJ.zip|Cube]]
 +
* [[Media:ConeOBJ.zip|Cone]]
 +
* [[Media:CylinderOBJ.zip|Cylinder]]
 +
* Sphere: [[Media:TriSphereOBJ.zip|triangles, no texture coordinates)]], [[Media:TexSphereOBJ.zip|quads with texture coordinates]]
 +
* Torus: [[Media:TorusLR_OBJ.zip|low polygon count]], [[Media:TorusHR_OBJ.zip|high polygon count]], [[Media:TorusThin_OBJ.zip|thin torus]]
 +
* [[Media:TeapotOBJ.zip|Utah Teapot]]
  
[[Image:robot.png]]
+
Here a few examples of rides (but yours needs to have a sky box and a disco ball also):
  
Use your creativity to build the most creative robot in class! The 5 most creative robots in class are going to get extra credit, after a vote on Canvas.
+
[[Image:Ride3F20.jpg|x200px]]
 +
[[Image:FerrisWheel.png|x200px]]
 +
[[Image:RideF20.jpg|x200px]]
 +
[[Image:Ride2F20.jpg|x200px]]
 +
[[Image:Carousel.png|x200px]]
  
 
Once you've created your scene graph, you need to get your rendering engine ready to recursively traverse the scene graph for rendering by creating a root node of type <tt>Transform</tt> and calling its <tt>draw()</tt> function with the identity matrix as its parameter.
 
Once you've created your scene graph, you need to get your rendering engine ready to recursively traverse the scene graph for rendering by creating a root node of type <tt>Transform</tt> and calling its <tt>draw()</tt> function with the identity matrix as its parameter.
  
<!-- require normal based lighting for the robot surface -->
+
'''Grading:'''
 +
* -3 if no ground plane or ride not anchored to ground
 +
* -5 if disco ball is missing
 +
* -3 per missing hierarchy level (3 levels required)
 +
* -5 for lighting problems (no light source, no Phong shading, incorrect normals, etc)
  
==3. Animated Robot (15 Points)==
+
==5. Animation and Interaction (10 Points)==
 +
 
 +
* Animate your amusement park ride to make it look like it would when it is running, by changing the matrices in the <tt>Transform</tt> nodes.
 +
* Allow the user to start/stop the animation of each of the levels of your motion hierarchy separately. (For example: if you're making a ferris wheel: the '1' key starts/stops the rotation of the ferris wheel itself, the '2' key starts/stops the rotation of the cars around the horizontal axis to keep them level with the ground, the '3' key starts/stops the rotation of the cars about their vertical axis.)
 +
* Give the user keyboard or mouse control to move around within the simulation (by moving the camera): the user needs to be able to move forward, back, up or down, as well as turn left or right. The up vector of the camera should always point straight up in the world (no banking or pitching the view).
 +
 
 +
'''Grading:'''
 +
* -5 if ride is not animated
 +
* -5 if motion on hierarchy levels can't independently be started/stopped
 +
* -5 if user can't move around the simulation (-2 if only rotations are missing)
  
Animate the robot to make it look like it is walking, by changing the matrices in the <tt>Transform</tt> nodes. Walking in place is fine, the robot does not need to actually move forward.
+
==6. Extra Credit (Up to 10 Points Max.)==
  
In your robot, at least 3 of its parts need to be moving independently from one another and they need to be connected to the 4th part
+
Choose from the following options. The points add up to a maximum of 10 points.
  
<!-- next time: specify how exactly update routine must be called -->
+
# Create a camera node and attach it to one of the leaf nodes of your scene graph, for instance showing a rider's view. Allow switching between this view and the default observer view with a keyboard key. (3 points)
 +
# Create a custom cube map for your sky box using your smartphone in 360 degree panorama capture mode, or a dedicated 360 degree panorama camera. [https://aerotwist.com/tutorials/create-your-own-environment-maps/ This tutorial] might be useful. (5 points)
 +
# Create a custom, non-trivial object (not a basic mathematical shape or derived from one) with a 3D modeling tool (e.g., Blender), or find a free model on the internet which you add to your amusement park scene. It needs to be textured, have normal vectors, and you need to render it with Phong shading. (5 points)
 +
# Dynanmic environment mapping: have the disco ball reflect not just the sky box, but also all the components of your ride. (10 points)
  
==Robot Party (Points)==
 
  
* Construct a scene which consists of a large amount of robots, at least 100. The robots can all be identical clones. (5 points)
+
=='''Submission Instructions'''==
* Distribute the robots on a 2D grid (i.e., place them on a plane with uniform spacing). For 100 robots, use a 10x10 grid. (5 points)
+
* Enable the animation for all the robots so that they look like they are dancing. (5 points)
+
* Make sure your camera can be rotated and zoomed so that you can change between having all robots on screen all the way to none and back.
+
  
This image illustrates the grid layout of the robots (background image not required):
+
Once you are done with the homework project, record a video demonstrating all the functionality you have implemented.
  
[[Image:robot-army.png]]
+
The video should be '''no longer than 5 minutes''', and can be substantially shorter.
  
==Culling (Points)==
+
Ideally, you will talk over the video to explain what you are showing. Alternatively, you can display text for this purpose.
  
Implement object level culling, to allow the existence of thousands of instances of your robot, without having to render them all at once.
+
To create the video you don't need to use video editing software, but you should use software to capture your screen to a video file.  
  
* Determine the parameters for a bounding sphere (Vector3 for its center point, and a radius) for each of your robots, which contains all parts of the robot. You do not need to find the tightest possible bounding spheres. Just make them as tight as you reasonably can. Estimating the bounding sphere size is fine. You don't need to have a separate bounding sphere for each animation step - one that encloses all steps is fine.
+
* For Windows users we recommend the free [https://obsproject.com/ OBS Studio].  
* Add the optional ability to render the bounding spheres. Support a keyboard key to toggle the rendering of the bounding spheres on or off. You can render the spheres by using the sphere OBJ file from project 2. Other ways of rendering spheres are also acceptable. We recommend rendering a wireframe sphere by rendering OpenGL lines instead of triangles. (5 points)
+
* On Macs you can use Quicktime to record off the screen.
* Add view frustum culling, and support a keyboard key to turn it on or off.  (10 points)
+
* Display the number of robots that are visible (=not culled) on standard out or in the window's title bar. Make sure that by turning culling on this number goes down when part of the robot array is off screen. (5 points)
+
* Debug Mode: Implement a demo mode in which you zoom the camera out (increase the FOV by a few degrees) but do the culling with the original FOV, so that one can see when the robots get culled. Allow toggling between demo mode and normal mode with a keyboard key. (5 points)
+
  
'''Notes:'''
+
If this does not work for you, you can hold your phone in front of the screen and record with a steady hand.
* [http://www.lighthouse3d.com/tutorials/view-frustum-culling/geometric-approach-extracting-the-planes/ Lighthouse3D] has an excellent description of how to calculate the view frustum plane equations. Note that this tutorial and the discussion slides assume that the frustum plane normals point away from the view volume, whereas the lecture slides have them point into the view volume. Either way can work, you just need to be consistent.
+
  
==Extra Credit (Up to 10 Points)==
+
Upload this video at the Assignment link on Canvas, and comment which functionality you have or have not implemented and what extra credit you have implemented. If you couldn't implement something in its entirety, please state which parts you did implement and hope to get points for.
  
* Hierarchical Culling: Create a hierarchical culling algorithm by storing bounding sphere information at every level of the scene graph, so that you can cull an entire branch of the scene graph at once. Structure your scene graph so that you have multiple levels (for instance, by subdividing your robot array into four quarters, and create a separate node for each quarter. Show how many times you were able to cull based on the higher level bounding spheres by displaying the number in the text window. Use differently colored spheres for the different levels of your culling hierarchy. (5 points)
+
* Example 1: I've done the base project with no issues. No extra credit.
 +
* Example 2: Everything works except an issue with '''x''', I couldn't get '''y''' to work properly.
 +
* Example 3: Sections 1-2 and 4 work.
 +
* Example 4: The base project is complete and I did '''z''' for extra credit.
  
* Creativity Contest (up to 5 Points): We're going to have a contest for the top 5 most creative robot creations. Submit a JPEG image or GIF animation of your robot to Canvas by adding an entry to the respective discussion thread. To create a GIF animation you can use [http://gifmaker.me/ this converter]. To convert any image file format to a GIF image, we recommend [http://www.irfanview.com/ IrfanView.] The top five most voted for robots are going to get extra credit. Voting starts Friday 11/1 at 11:59pm and ends on the following Wednesday 11/6 at 11:59pm. The winner of the contest will get 5 points of extra credit. The second place will get 4 points, third gets 3, fourth 2, fifth 1 point.
+
Finally, zip up your source code (.cpp, .h, executable and shader files, as well as '''any .obj files which were not provided to you''') and submit the zip file at the Assignment link on Canvas as well.

Latest revision as of 20:15, 13 December 2020

Contents

Project 3: Amusement Park Ride

In this project you will need to implement a scene graph to render an amusement park ride which consists of a number of objects that are part of a hierarchy.

This project implements the following skills from the lectures:

  • texturing
  • sky box
  • environment mapping
  • scene graph

The user of your program should be able to control a variety of options within your simulation using the keyboard. You may choose which keys control which functions, but make sure they are clearly documented in either a text file named README.txt which you add to the main directory of your code, or in your video (narration or text overlays). The following interactive controls should be available to the user:

  • Start/stop the animation of each of the levels of your motion hierarchy separately.
  • Move around within the simulation (by moving the camera): forward, back, up or down, as well as turn the camera left or right (which affects subsequent forward movements).

The total score for this project is 100 points. Additionally, you can obtain up to 10 points of extra credit.

There will not be a discussion on November 11th because of Veterans Day. Instead, this project will be introduced in lecture on Tuesday, November 10th.

We recommend starting with your code from project 2 and removing the rendering of the 3D models. In this project you are going to render all new objects, using your own implementation of a scene graph.

1. Sky Box (20 Points)

Create a sky box for your scene, to give it a nice background. LearnOpenGL offers a great tutorial for sky boxes and how to implement them in modern OpenGL.

Make sure the width and height of your sky box textures are powers of two, we recommend 512x512 pixels. You are allowed to create your own sky box if you want, many smartphones can capture 360 degree panorama images which you would need to turn into a cube map. If you would rather use ready-made sky boxes from the internet, here a few pointers:

If you created or downloaded a panorama image from the web, you need to convert it into the six cube map faces you need for the sky box. This web tool can do this for you.

Once you created or downloaded your sky box textures, use them to create a sky box by texturing the insides of a very large cube (coordinates from about -500 to +500). Make sure that the textures align correctly at the edges - if not you may need to rotate or flip some of the textures.

Use single-sided rendering, so that OpenGL will automatically hide the user-facing sides of the cube. If your sky box is defined with the triangles' normals facing outward (like the example in the discussion slides), use the following code lines to ensure that the user will never see the outside of the box:

glEnable(GL_CULL_FACE); 
glCullFace(GL_FRONT); 

If your normals point inward, you need to use GL_BACK instead of GL_FRONT in the above code.

Use the following settings for your texture after your first glBindTexture(GL_TEXTURE_CUBE_MAP, id) for correct lighting and filtering settings:


  // Make sure no bytes are padded:
  glPixelStorei(GL_UNPACK_ALIGNMENT, 1);

  // Use bilinear interpolation:
  glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
  glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_MAG_FILTER, GL_LINEAR);

  // Use clamp to edge to hide skybox edges:
  glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
  glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);

Grading:

  • -5 if sky box works but is not seamless
  • -3 if backface culling is not enabled, or it is enabled but the wrong faces are culled

2. Environment Mapping (20 Points)

Create an algorithm to tessellate a disco ball style sphere (ie, create vertices and normals for a sphere with a relatively low number of quads for faces). Its surface should look like polished metal, so you need to use environment mapping on it.

This sphere needs to become part of your amusement park ride later, but we recommend implementing it beforehand so that you can test it out separately. Unless you don't get to render the ride, you don't need to show this sphere separately in your demo (although you could).

We provide code for the sphere geometry for Windows and for Mac. But feel free to write your own.

The Cube map tutorial explains how environment mapping is done with a sky box as the environment. Feel free to follow it and use the code from the site to make it work in your code. The goal is to have the sky box reflect on the sphere, to give the impression of the sphere's surface being polished metal (i.e., act like a perfect mirror).

Here is an example of a sphere with environment mapping in a sky box:

DiscoballF20.jpg

Grading:

  • -10 if reflection looks somewhat like the skybox but is off by a lot
  • -5 if reflection is upside down but otherwise correct

3. Scene Graph Engine (25 Points)

To create a ride with multiple moving parts, we need to first implement a simple scene graph structure for our rendering engine. This scene graph should consist of at least three nodes: Node (5 points), Transform (10 points) and Geometry (10 points). You are free to add more scene graph node types as you see fit.

  • Class Node should be abstract and serve as the common base class. It should implement the following class methods:
    • an abstract draw method: virtual void draw(Matrix4 C)=0
    • an abstract virtual void update()=0 method to separate bounding sphere updates from rendering
  • Transform should be derived from Node and have the following features:
    • store a 4x4 transformation matrix M
    • store a list of pointers to child nodes (std::list<Node*>)
    • provide a class methods to add a child node (addChild()) to the list
    • its draw method needs to traverse the list of children and call each child node's draw function
    • when draw(C) is called, multiply matrix M with matrix C.
  • Geometry should be derived from Node and have the following features:
    • set the modelview matrix to the current C matrix
    • an initialization method to load a 3D model (OBJ file) whose filename is passed to it (init(string filename). Your OBJ loader from project 2 should work.
    • have a class method which draws the 3D model associated with this node.

4. Ride Design (25 Points)

Now that we have the scene graph classes, it is time to put them to work. Build your own amusement park ride using the addChild methods.

Your design needs to include the following features:

  • You need to have a ground plane for the ride.
  • The ride must be anchored to the ground plane (can't be floating in space).
  • Your disco ball sphere must be part of the design. It can occur once or multiple times.
  • Your hierarchy needs to be at least three levels deep. For instance a ferris wheel which spins around its center with cars that rotate so they are always aligned with gravity, with the cars rotating about their vertical axis (see first picture below).
  • Drawing the same object multiple times in different positions, orientations, and/or with different scaling factors.
  • Moving objects over time, including rotations not centered at the origin, linear motion (changing direction at the end points) and compound motions.
  • Use regular Phong shading as in project 2 for the surfaces of all your ride geometries. This means that you need to have correct normals for all your 3D objects.
  • Implement at least one directional or point light source for your scene, and make sure that all surfaces are illuminated correctly. (Ideally, you use your lighting environment from project 2 which would also allow you to move your light source around with the mouse, but this is not a requirement.)

Adequate results can be achieved with simple cubes, cones, cylinders and spheres. More complex objects are not required for this simulation, but you can get extra credit for them. Below are a few links to basic 3D shapes you can use. You are also allowed to use the OBJ files from projects 1 and 2.

Here a few examples of rides (but yours needs to have a sky box and a disco ball also):

Ride3F20.jpg FerrisWheel.png RideF20.jpg Ride2F20.jpg Carousel.png

Once you've created your scene graph, you need to get your rendering engine ready to recursively traverse the scene graph for rendering by creating a root node of type Transform and calling its draw() function with the identity matrix as its parameter.

Grading:

  • -3 if no ground plane or ride not anchored to ground
  • -5 if disco ball is missing
  • -3 per missing hierarchy level (3 levels required)
  • -5 for lighting problems (no light source, no Phong shading, incorrect normals, etc)

5. Animation and Interaction (10 Points)

  • Animate your amusement park ride to make it look like it would when it is running, by changing the matrices in the Transform nodes.
  • Allow the user to start/stop the animation of each of the levels of your motion hierarchy separately. (For example: if you're making a ferris wheel: the '1' key starts/stops the rotation of the ferris wheel itself, the '2' key starts/stops the rotation of the cars around the horizontal axis to keep them level with the ground, the '3' key starts/stops the rotation of the cars about their vertical axis.)
  • Give the user keyboard or mouse control to move around within the simulation (by moving the camera): the user needs to be able to move forward, back, up or down, as well as turn left or right. The up vector of the camera should always point straight up in the world (no banking or pitching the view).

Grading:

  • -5 if ride is not animated
  • -5 if motion on hierarchy levels can't independently be started/stopped
  • -5 if user can't move around the simulation (-2 if only rotations are missing)

6. Extra Credit (Up to 10 Points Max.)

Choose from the following options. The points add up to a maximum of 10 points.

  1. Create a camera node and attach it to one of the leaf nodes of your scene graph, for instance showing a rider's view. Allow switching between this view and the default observer view with a keyboard key. (3 points)
  2. Create a custom cube map for your sky box using your smartphone in 360 degree panorama capture mode, or a dedicated 360 degree panorama camera. This tutorial might be useful. (5 points)
  3. Create a custom, non-trivial object (not a basic mathematical shape or derived from one) with a 3D modeling tool (e.g., Blender), or find a free model on the internet which you add to your amusement park scene. It needs to be textured, have normal vectors, and you need to render it with Phong shading. (5 points)
  4. Dynanmic environment mapping: have the disco ball reflect not just the sky box, but also all the components of your ride. (10 points)


Submission Instructions

Once you are done with the homework project, record a video demonstrating all the functionality you have implemented.

The video should be no longer than 5 minutes, and can be substantially shorter.

Ideally, you will talk over the video to explain what you are showing. Alternatively, you can display text for this purpose.

To create the video you don't need to use video editing software, but you should use software to capture your screen to a video file.

  • For Windows users we recommend the free OBS Studio.
  • On Macs you can use Quicktime to record off the screen.

If this does not work for you, you can hold your phone in front of the screen and record with a steady hand.

Upload this video at the Assignment link on Canvas, and comment which functionality you have or have not implemented and what extra credit you have implemented. If you couldn't implement something in its entirety, please state which parts you did implement and hope to get points for.

  • Example 1: I've done the base project with no issues. No extra credit.
  • Example 2: Everything works except an issue with x, I couldn't get y to work properly.
  • Example 3: Sections 1-2 and 4 work.
  • Example 4: The base project is complete and I did z for extra credit.

Finally, zip up your source code (.cpp, .h, executable and shader files, as well as any .obj files which were not provided to you) and submit the zip file at the Assignment link on Canvas as well.