Bojian Wu
ustcbjwu [AT] gmail.com
Chinese Academy of Sciences (CAS)
Week 13
Ray tracing of some real B-rep models
| Model | Raytracing (with frame) |
Raytracing | nirt | ||
|---|---|---|---|---|---|
| Top | Front | Left | |||
| Bench |  |
 |
 |
 |
 |
| Basin |  |
 |
 |
 |
 |
| Toy |  |
 |
 |
MISS |   |
A comparison of solid BoT/B-rep sphere ray tracing results
| Sphere(radius=10) | BoT | B-rep | ||
|---|---|---|---|---|
| Top |  |
 |
 |
 |
 |
 |
 |
 |
|
 |
 |
 |
 |
|
| Front |  |
 |
 |
 |
 |
 |
 |
 |
|
 |
 |
 |
 |
|
| Left |  |
 |
 |
 |
 |
 |
 |
 |
|
 |
 |
 |
 |
|
A comparison of plate-mode BoT/B-rep sphere ray tracing results
The difference here is LOS. I find where is the problem. For BoT, it supports two plate modes: *plate* and *plate nocos*. But in my implementation of plate mode B-rep, I only consider the *plate nocos* but not *plate* as discussed previously. I think it should influence the result of *gqa*.
| Sphere(radius=1.0, thickness=0.5) | BoT | B-rep | ||
|---|---|---|---|---|
| Top |  |
 |
 |
 |
 |
 |
 |
 |
|
 |
 |
 |
 |
|
| Front |  |
 |
 |
 |
 |
 |
 |
 |
|
 |
 |
 |
 |
|
| Left |  |
 |
 |
 |
 |
 |
 |
 |
|
 |
 |
 |
 |
|