GPU Splatting  1.0
A deffered render that draws the surface of point cloud data.
Public Member Functions | List of all members
Vis2Application Class Reference

The main class of the renderer - the action is here! More...

#include <Vis2Application.h>

Inherits GLFWApplication.

Public Member Functions

 Vis2Application (int width, int height, int ctx_major, int ctx_minor)
 Creates a new instance with the given parameters. More...
 
void init () override
 Initializes all the OpenGL buffers and loads the bunny-model. More...
 
void update () override
 Does nothing so far.
 
void render () override
 Renders the scene. More...
 
void resize (int width, int height) override
 Reallocates the textures to fit the new screen size.
 
void initUI () override
 Initializes the AntTweakBar.
 
void renderUI () override
 Renders the AntTweakBar.
 
void loadModel (const std::string model_path)
 Loads the specified model. More...
 
void setModel (Models model)
 Sets the active model.
 
Models getModel ()
 Returns the active model.
 
- Public Member Functions inherited from GLFWApplication
 GLFWApplication (int width, int height, int ctx_major, int ctx_minor)
 Creates a new instance with the given parameters. More...
 
void run ()
 This method starts the update-render-loop and returns if the window has been closed. More...
 

Additional Inherited Members

- Protected Member Functions inherited from GLFWApplication
GLFWwindow * getWindow ()
 Returns the GLFW-window-handle.
 
int getWidth ()
 Returns the width of the window.
 
int getHeight ()
 Returns the height of the window.
 
float getFrameTime ()
 Returns the duration of the last frame in seconds.
 
float getAbsTime ()
 
double getMouseX ()
 Returns the X-position of the mouse in the window.
 
double getMouseY ()
 Returns the Y-position of the mouse in the window.
 
double getMouseDx ()
 Returns the difference in X-position of the mouse since last frame.
 
double getMouseDy ()
 Returns the difference in Y-position of the mouse since last frame.
 

Detailed Description

The main class of the renderer - the action is here!

This class inherits from GLFWApplication and implements all its hooks. This is the heart of the renderer - it is capeable of loading and rendering point cloud data.

Constructor & Destructor Documentation

Vis2Application::Vis2Application ( int  width,
int  height,
int  ctx_major,
int  ctx_minor 
)

Creates a new instance with the given parameters.

Parameters
widththe initial width of the window
heightthe initial height of the window
ctx_majorthe requested major version of OpenGL
ctx_minorthe requested minor version of OpenGL

Member Function Documentation

void Vis2Application::init ( )
overridevirtual

Initializes all the OpenGL buffers and loads the bunny-model.

Creates all the buffers needed for storing our point-cloud.

It will also create all the textures needed and attach them to the appropriate Framebuffers.

Finally all shaders are compiled and linked to programs. Then it is time to load the default model - the bunny.

Implements GLFWApplication.

void Vis2Application::loadModel ( const std::string  model_path)

Loads the specified model.

Reads the vertex-data from the file and stores it in memory. It then computes an KD-Tree for the positions.

For each point the 8 closest neighbours are queried and used to compute the eigenvectors and eigenvalues of the current point. With this information at hand one can compute the splat-axes.

Finally all the data is written to the array-buffer of the model.

void Vis2Application::render ( )
overridevirtual

Renders the scene.

First the view and model matrix are computed.

In the first pass(Depth-Pass) all points are rendered into the attached depth-buffer. This is necessary to prevent Vertices that are behind the surface from beeing rendered and create a weird transperency-effect.

In the second Pass(Attribute-Pass) all vertices that are visible will be drawn onto the G-Buffer that consists of 3 textures for color, normals and the accumulated weight. This pass will only use a custom depth-test, because it blends ovelapping splats together.

In the third and final pass(Shading-Pass) the scene is drawn onto a full-screen-quad. It uses the accumulated values from the previous pass to compute shading of the object. We uses the phong-lighting model for our program.

Implements GLFWApplication.

The documentation for this class was generated from the following files: