computer-graphics

Some experiments about computer graphics.

It implements a Camera, Object, Scene, and some rendering and transformation functions. The Camera windowing system is based on wxPython.

Demo using two scenes, Scene 1 with two cameras and Scene 2 with just one camera.

Dependencies

numpy 1.20: used np.array, np.full, np.matmul, np.dot, np.cross and np.linalg.norm;
webcolors 1.11: converts HEX code to RGB;
wxPython 4.1: windowing system which provides a bitmap canvas to freely draw into, besides a plenty of useful features.

The easiest method to install the requirements is thought a Python environment managed by Conda or Anaconda. If you don't have one setup, we recommend using miniconda because it's a minimal installer for conda that includes only conda, Python and a few other utility packages.

After setting up your Conda installation, run the following commands:

# create a Python environment called `cg`
conda create -n cg

# entering the newly created environment
conda activate cg

# installing the requirements
conda install -c conda-forge numpy=1.20 webcolors=1.11 wxpython=4.1

💡 You can install each dependency by hand using pip, but it's a more complicated process involving compiling wxPython specifically for your platform. That's out of the scope of this guide, so read this page for more information.

Usage

Usage as in src/main.py:

from Camera import Camera
from Object import Object
from Scene import Scene
from LightSource import LightSource

if __name__ == '__main__':
    s1 = Scene(.1)
    s2 = Scene(.4)

    ls1 = LightSource([5_000., 5_000., -5_000.], 1.)
    s1.add_light_source(ls1)

    ls2 = LightSource([5_000., 5_000., -5_000.], 1.)
    ls3 = LightSource([-5_000., -5_000., -5_000.], .3)
    s2.add_light_source(ls2)
    s2.add_light_source(ls3)

    obj1 = Object('./3d-obj-examples/coarseTri.fandiskAuto.obj', '#4070a0')
    obj1.translate(-0.5, 0, 0)
    obj1.scale(200)
    s1.add_object(obj1)

    obj2 = Object('./3d-obj-examples/coarseTri.egea1.obj', '#e74c3c', .7, 1., 5)
    obj2.translate(0.5, 0, 0)
    obj2.scale(200)
    s1.add_object(obj2)

    obj3 = Object('./3d-obj-examples/coarseTri.rockerArm.obj', '#208050', .3, 1., 5)
    obj3.scale(200)
    obj3.rotate_y(90)
    obj3.rotate_x(30)
    s2.add_object(obj3)

    cam1 = Camera(s1)
    cam1.rotate_y(-50)
    cam1.rotate_x(20)

    cam2 = Camera(s1)
    cam2.rotate_x(20)

    cam3 = Camera(s2)

    Scene.main_loop()

Running

After installing the requirements, you need to cd to the project root then activate your Python environment:

conda activate cg

If you see (cg) at the beginning of the command line, you're ready to go. Run the project using the command:

python3 src/main.py

You can interact with each Camera using the following commands:

RIGHT: right (translate(C, 0, 0))
LEFT: left (translate(-C, 0, 0))
UP: front (translate(0, 0, C))
DOWN: back (translate(0, 0, -C))
SHIFT + UP: up (translate(0, C, 0))
SHIFT + DOWN: down (translate(0, -C, 0))
CTRL + RIGHT: positive yaw (rotate_y(-D))
CTRL + LEFT: negative yaw (rotate_y(D))
CTRL + UP: positive pitch (rotate_x(-D))
CTRL + DOWN: negative pitch (rotate_x(D))
CTRL + SHIFT + RIGHT: positive roll (rotate_z(-D))
CTRL + SHIFT + LEFT: negative roll (rotate_z(D))

To terminate the execution you can close all windows one by one or hit CTRL + C on the terminal.

Troubleshooting

If you encounter any error, it probably might be related to wxPython and it's dependencies:

ImportError: /path/to/miniconda3/envs/cg/bin/../lib/libstdc++.so.6: version `CXXABI_1.3.13' not found (required by /path/to/python3.9/site-packages/wx/libwx_baseu-3.1.so.5): it means that there's a wxPython installed within your main Python install. You can uninstall it (pip uninstall wxpython) and try to run again;
Gtk-Message: 18:21:48.703: Failed to load module "canberra-gtk-module" and Gtk-Message: 18:21:48.703: Failed to load module "pk-gtk-module": as this project doesn't use features provided by these modules, they're not needed. This error can be safely ignored.

arthursimas1 / computer-graphics Goto Github PK