Real-time Edge-Aware Image Processing with the Bilateral Grid: Difference between revisions
No edit summary |
(→Method) |
||
Line 6: | Line 6: | ||
==Method== | ==Method== | ||
# Create a low-resolution 3D grid with axis (x, y, intensity). | # Create a low-resolution 3D grid with axis (x, y, intensity). | ||
#* For an 8MP image, they claim a \(70 \times 70 \times 10\) grid is sufficient. | |||
#* This is called a ''bilateral grid''. | #* This is called a ''bilateral grid''. | ||
# Place pixels for your 2D image into bins within the bilateral grid based on x, y | # Place pixels for your 2D image into bins within the bilateral grid based on x, y position and intensity. | ||
# When performing image effects, only change pixels within ''slices'' of the grid: pixels within a certain intensity. | # When performing image effects, only change pixels within ''slices'' of the grid: pixels within a certain intensity. |
Latest revision as of 15:19, 26 August 2020
Real-time Edge-Aware Image Processing with the Bilateral Grid (SIGGRAPH 2007)
Method
- Create a low-resolution 3D grid with axis (x, y, intensity).
- For an 8MP image, they claim a \(70 \times 70 \times 10\) grid is sufficient.
- This is called a bilateral grid.
- Place pixels for your 2D image into bins within the bilateral grid based on x, y position and intensity.
- When performing image effects, only change pixels within slices of the grid: pixels within a certain intensity.