Lecture 5: Texture Mapping (5)
ryazlee

does this account for rigid/nonflat surfaces? it seems like irl when the chocolate wrapper wraps the surface is smooth, but how would you model a rigid surface with a foil wrapper?

didvi

Great observation! This is one of the many difficulties that come up when texture mapping. In line with the chocolate wrapper analogy, you may still be able to wrap a rigid surface with this chocolate wrapper, however this may cause distortions or other un-intended effects. Dealing with these effects is a rich area of research that we will sadly not be covering in this course.

alexkassil

One thing that I think is interesting to note is distortion. At the center of the image (face/front of body) but as you move out, it gets more distorted (gloves/had/hat top/pants/back of body).

I think this is because the face/front of body is relatively flat, and it curves around the hat/side of the body, so flatness in the 3d model looks normalish but big curves look distorted in the texture image.

tlswoo

Great observation, Alex! I noticed that as well, the process behind mapping a flat texture image to something 3D, and accounting for the wrapping factor as well is honestly pretty cool. Something I was also wondering about was the integrity of the flattened texture based on different application contexts, or using different materials to form the texture image (e.g. hydrodipping). What might be the benefits of this method over others, and what properties of this method make it better suited to certain situations?

jeremymui

@alexkassil Interesting to note that even if we are projecting the globe onto a map, where every part is equally "relatively flat", there tends to be more distortion on the edges than near the center, as seen through Tissot's indicatrix.

You must be enrolled in the course to comment