Developing the Demosaicing - PowerPoint PPT Presentation

About This Presentation
Title:

Developing the Demosaicing

Description:

Demosaicing algorithm is to reconstruct a full color image ... Dual Core AMD Opteron(tm) 2212 Frequency 2.0GHz. Experiment Results ... – PowerPoint PPT presentation

Number of Views:60
Avg rating:3.0/5.0
Slides: 21
Provided by: winstonvs
Learn more at: http://coachk.cs.ucf.edu
Category:

less

Transcript and Presenter's Notes

Title: Developing the Demosaicing


1
Developing the Demosaicing Algorithm
in GPGPU
Ping Xiang Electrical engineering and computer
science
2
Outline
Background Algorithm Implementation Experiment
Results Future Work
3
Background
  • Color Filter Array.
  • A mosaic of color filters in front of the
    image sensor

4
Background
  • Demosaicing algorithm is to reconstruct a full
    color image
  • from the data collected by the color filtering
    array.

5
Algorithm
  • Bilinear interpolation
  • The red value of a non-red pixel is computed by
    the
  • average of the two or four adjacent red pixels,
    and
  • similarly for blue and green.

6
Algorithm
7
Algorithm
8
Algorithm
For Green Channels
9
Algorithm
For Red or Blue Channels
10
Implementation
Optimization 1. Vectorize the pixel data to be
processed 2. use shared
memory to reduce the data transfer
11
Implementation
1. Vectorize the pixel data to be processed
12
Implementation
2. Use shared memory to reduce the data transfer
13
Experiment Results
Platform ATI Radeon HD 4870
Brook 1.4 Nvidia GeForce 8800
GTX CUDA 2.1 Dual Core AMD
Opteron(tm) 2212 Frequency 2.0GHz
14
Experiment Results
For small data size, GPU is not always a good
choice a. Memory transfer time dominates
the kernel execution time b. Computation is
not that complex enough
15
Experiment Results
When the data size is small, CUDA has better
performance. When the data size increases to
4K, the brook performance catches up with
CUDA
16
Experiment Results
Explanation ?

Memory Speed Memory Speed Stream processing Units Stream processing Units Stream processing Units
HD 4870 800 85102
GTX 8800 GTX 8800 128 (168)

17
Experiment Results
Shared Register Usage
Read data into shared register and try to reuse
the data
18
Experiment Results
19
Future Work
1. Shared memory usage for further optimization
2. Integrate the code with proper interface to
import image data and export pixel data
3. Report
20
Reference
1. High-Quality linear interpolation for
Demosaicing of Bayer-patterned color images,
Henrique S. Malvar, Li-wei He, and Ross Cutler
2. An Improved Demosaicing Algorithm Alexey
Lukin, Denis Kubasov
Questions?
Write a Comment
User Comments (0)
About PowerShow.com