Volume RayCasting on Graphics Hardware CDA 6938 Spring
- Slides: 22
Volume Ray-Casting on Graphics Hardware CDA 6938 : Spring 2009 Mangesh Nijasure
Algorithm n Ray-Casting n Sampling n Shading n Compositing
Ray-Casting
Setup in hardware
Ray-Plane intersection t = -eye_pos. x / ray_dir. x curr_point = eye_pos + (t * ray_dir) if(curr_point. y>=0. 0 f && curr_point. y<=1. 0 f && curr_point. z>=0. 0 f && curr_point. z<=1. 0 f) { } if(t<tmin) tmin = t if(t>tmax) tmax = t
Visualizing tmin and tmax tmin tmax
Sampling n n At each sample interval, fetch from the 3 D volume texture Using this value fetch the color from the 1 D transfer function
Tri-Linear interpolation
Shading n Use the Center-Differences method for gradient computation (pre-computation step) n Fetch from the 3 D gradient texture n Compute light direction n Evaluate the N. L dot product to get the lighting factor
Compositing n alpha = 1 – old_alpha n new_color = (shaded_col*alpha) + old_color n old_color = new_color
Demo
Results n Software Solution n n Intel Pentium IV 3. 8 GHz 1 GB Ram GPU Hardware Solution n n AMD Radeon HD 4670 - mid range part (~ $100) Core Clock 750 MHz 320 Stream Processors 512 MB On board memory Direct. X 10 Graphics API implementation
Performance No Lighting at 512 x 512 Data Set Software Milisec/Frame Hardware Milisec/Frame Hydrogen Atom 1, 375 11 Bucky Ball 1, 394 11 Bonsai 1, 451 12 CT Head 1, 426 11 64 samples
Performance With Lighting at 512 x 512 Data Set Software Milisec/Frame Hardware Milisec/Frame Hydrogen Atom 6, 346 16 Bucky Ball 6, 322 16 Bonsai 7, 002 28 CT Head 6, 953 31 * Results are not visually accurate, but timings are exact 64 samples
Performance with Hydrogen Data Set Samples & Resolution No Lighting Times in miliseconds / frame 512 x 512 Samples 1024 x 1024 S/W H/W 32 777 6 3, 129 22 64 1, 375 11 5, 796 36 128 2, 614 22 10, 516 77 256 5, 034 36 20, 220 114
Performance with Hydrogen Data Set Interpolation Effect No Lighting 512 x 512 Times in miliseconds / frame Point Sampled Samples Tri-Linear S/W H/W 32 426 6 777 6 64 740 11 1, 375 11 128 1340 22 2, 614 22 256 2545 42 5, 034 36
Keys to performance gain n Data level parallelism (embarrassingly parallel) n Ray coherence (texture cache efficiency) n Specialized interpolation hardware (texture filtering) n Minimized bus traffic for data sets (large onchip memory)
Hydrogen Data Set
Bonsai Data Set
Bucky. Ball Data Set
Head Data Set
Questions ?
- Direct volume rendering ray casting
- Dot matrix display ppt
- Hardware 1990 cda
- Internal hardware
- Spring, summer, fall, winter... and spring (2003)
- Months of spring summer autumn winter
- Vbo1
- What is computer
- Graphics monitors and workstations and input devices
- Real time volume graphics
- Lung capacity
- Stroke volume units
- Example of large volume parenterals
- Volume kerucut = .....x volume tabung *
- Stroke volume normal
- Solute vs solvent
- Cda competency statement 5 examples
- Gangs of new york cda
- Cda3101
- Rrrrrrr cda
- Lady macbeth cda
- Mispeculation
- Lusi cda