kd-tree building - check all axes for best split, add additional shape-bbox check
extent Container bounds by Eps to fix invisible triangles on borders
new Camera constructor with more intuitive lookat/up vectors
fix camera axes (mirrored images)
better camera angle-of-view
change capitalization of addShape and addLight
#!/usr/bin/python
# read nff data from standart input and render image to render_nff.png
# see http://tog.acm.org/resources/SPD/
# only spheres and triangles are implemented
from raytracer import Raytracer, Camera, Light, Material, Sphere, NormalVertex, Triangle
from math import pi
import sys, Image
rt = Raytracer()
imagesize = (800, 600)
mat = Material(colour=(1.0, 1.0, 1.0))
f = sys.stdin
while True:
line = f.readline()
if line == "":
break;
ln = line.split()
if ln[0] == 'v': # Viewpoint location
# from
ln = f.readline().split()
assert ln[0] == 'from'
eye = (float(ln[1]), float(ln[2]), float(ln[3]))
# at
ln = f.readline().split()
assert ln[0] == 'at'
lookat = (float(ln[1]), float(ln[2]), float(ln[3]))
# up
ln = f.readline().split()
assert ln[0] == 'up'
up = (float(ln[1]), float(ln[2]), float(ln[3]))
# angle
ln = f.readline().split()
assert ln[0] == 'angle'
angle = float(ln[1])
# hither
ln = f.readline().split()
assert ln[0] == 'hither'
hither = float(ln[1])
# resolution
ln = f.readline().split()
assert ln[0] == 'resolution'
imagesize = (int(ln[1]), int(ln[2]))
# set camera as specified
cam = Camera(eye=eye, lookat=lookat, up=up)
cam.setAngle(angle/180*pi)
rt.setcamera(cam)
elif ln[0] == 'b': # Background color
rt.setbgcolour((float(ln[1]), float(ln[2]), float(ln[3])))
elif ln[0] == 'l': # Light
pos = (float(ln[1]), float(ln[2]), float(ln[3]))
rt.addlight(Light(position=pos))
elif ln[0] == 'f': # Fill color and shading parameters
colour = (float(ln[1]), float(ln[2]), float(ln[3]))
mat = Material(colour=colour)
mat.setPhong(0,float(ln[4]),float(ln[5]),float(ln[6]))
mat.setTransmissivity(float(ln[7]),float(ln[8]))
elif ln[0] == 's': # Sphere
center = (float(ln[1]), float(ln[2]), float(ln[3]))
radius = float(ln[4])
rt.addshape(Sphere(centre=center, radius=radius, material=mat))
elif ln[0] == 'p': # Polygon
vertex_count = int(ln[1])
vertices = []
for i in range(vertex_count):
ln = f.readline().split()
vertex = (float(ln[0]), float(ln[1]), float(ln[2]))
vertices.append(NormalVertex(vertex))
rt.addshape(Triangle(vertices[0], vertices[1], vertices[2], mat))
for i in range(vertex_count)[3:]:
rt.addshape(Triangle(vertices[0], vertices[i-1], vertices[i], mat))
else:
print "Not implemented:", line
f.close()
data = rt.render(imagesize)
img = Image.fromstring("RGB", imagesize, data)
img.save('render_nff.png')