Lecture 9: Intro to Ray-Tracing & Accelerating Ray-Scene Intersection (8)
geos98
Just to test my understanding here. So assume if we are performing an exhaustive test, we will have a ray that originate from each pixel in the screen. How do we get the direction of each ray? Do we assume that the eye point is in the center of the screen? If so, would it cause the pixel at the edge of the screen to shoot in very sharp angles that basically cannot hit anything?
yzliu567
Eye ray starts at eye and goes through each pixel. For your last question, I think the distance between the eye point and the image plane is well designed to simulate real eyes. (Recall fov in the pinhole camera model)
adham-elarabawy
This formulation makes sense, but I am a bit confused about where the eye point is assumed to be, since the eye point can change physical location when looking at a screen, which means that the computations done to render the screen are not accurate to the true location of the eye point.
kkoujah
In the pinhole camera model, the eye point is typically assumed to be fixed at a single location, usually, at the center of the projection. To address the issue of the eye point moving as a person looks at a screen from different distances or angles, more advanced rendering techniques may take into account the movement of the eye point, as well as other factors such as the viewer's interpupillary distance and the curvature of the cornea. These techniques can provide more accurate and realistic renderings.
prannaypradeep999
Appel's shading/coloring algorithm specifically utilized math of the angle between the surface normal of the intersected object and the direction of the light source. This link goes into brief introductions of his efforts: https://www.semanticscholar.org/paper/Some-techniques-for-shading-machine-renderings-of-Appel/14a97553cf2d5414ec94b14bf22700b1b3c93a0d
Just to test my understanding here. So assume if we are performing an exhaustive test, we will have a ray that originate from each pixel in the screen. How do we get the direction of each ray? Do we assume that the eye point is in the center of the screen? If so, would it cause the pixel at the edge of the screen to shoot in very sharp angles that basically cannot hit anything?
Eye ray starts at eye and goes through each pixel. For your last question, I think the distance between the eye point and the image plane is well designed to simulate real eyes. (Recall fov in the pinhole camera model)
This formulation makes sense, but I am a bit confused about where the eye point is assumed to be, since the eye point can change physical location when looking at a screen, which means that the computations done to render the screen are not accurate to the true location of the eye point.
In the pinhole camera model, the eye point is typically assumed to be fixed at a single location, usually, at the center of the projection. To address the issue of the eye point moving as a person looks at a screen from different distances or angles, more advanced rendering techniques may take into account the movement of the eye point, as well as other factors such as the viewer's interpupillary distance and the curvature of the cornea. These techniques can provide more accurate and realistic renderings.
Appel's shading/coloring algorithm specifically utilized math of the angle between the surface normal of the intersected object and the direction of the light source. This link goes into brief introductions of his efforts: https://www.semanticscholar.org/paper/Some-techniques-for-shading-machine-renderings-of-Appel/14a97553cf2d5414ec94b14bf22700b1b3c93a0d