In this page are represented the results obtained, for each one of the 3D objects,
in each step of the pipeline and some demos obtained to prove the proper working of our framework.
Results of the Pipeline
The first result is for the triangulation step. In this step we decided to do the segmentation of the virtual object in small triangles to neglect the distortion's effects in the mirror.
For the parallelepiped it was used the "Delanauy" algorithm (45k 3D triangles), to obtain the 3D triangles that define the object. For the "bunny" and the "buddha" it was used the triangles' mesh
provided by the Stanford repository (approximately 72k 3D triangles).
The second result represents the solution for the texturization step. For the parallelepiped we used detailed textures for each face of the object and for
the "bunny" and the "buddha" we used a single colour texture, since it was not an objective of our work to map a texture to an irregular object.
The third result shows an example of the well known problem of occlusions, in Computer Graphics, when dealing with the parallelepiped object. In the first image is represented the final shape of
the object without considering this problem and, in the second image is represented the result obtained using our solution. There are several methods that can be used to solve the occlusions' problem, in our
framework we decided to make a reformulation of the "Painter's algorithm" since it is a simple an efficient methodology.
The fourth result is the representation of the three objects, taking into account all the stages of the pipeline. In these images we can notice the huge influence of the illumination step. In this step it was applied
a virtual spotlight, with a white colour for the parallelepiped the and "bunny" and a gold colour for the "buddha". To solve this problem it was made a reformulation of the "Phong's reflection model", since we are not
using a conventional perspective system, to find the colour of each triangle and, then, a flat shading technique to give the entire colour to the all the respective triangle.
Demos
Four different tests were applied to our framework to test its robustness.
To the bunny and to the cube was made an experiment with a moving spotlight showing the proper working of
our frame work.
Considering the "Phong's reflection model" more than one light source can be applied to a scene. So, an experiment was made using three spotlights with different movements
and, each one, with different RGB properties associated.
A final experiment was made using a moving robot. In this experiment it was used a spotlight positioned on the top of the robot pointing to the center of the arena. The 3D object is static and it is generated on the
center of the arena.