2. INTRODUCTION
The main objective of use of the parallel
computers is to speed up computations and
improve efficiency and response time by using
multiple CPUs, or to perform larger
computations which are not possible on single
processor system.
3. IMAGE PROCESSING
• Digital image processing is to improve the
pictorial information in order to perform other
tasks such as feature extraction or pattern
recognition.
• Image processing operations are done at
following levels:
Low level image processing
Intermediate image processing
High level image processing
4. PARALLEL PROCESSING
We differentiate between three models of
parallel processing:
• Pipeline Processing
• Asynchronous Processing
• Synchronous Processing
6. TASK PARALLEL SYSTEM
• Low level operations are grouped into tasks
and each task is assigned to a different
computing unit.
• An image processing application consists of
many different operations.
• The main challenge in task parallel approach is
efficient data decomposition and result
composition
8. PIXEL-PARALLEL MODE
Mean filter is a simple filter used to smooth
images, i.e. reduce the intensity variation
between one pixel and its neighbours (reduce
noise).
9. Given the assumption that the CPA used here is
4-neighbour connected, each pixel can directly
get its four direct neighbours indicated by
PE,PW,PS and PN. On the other hand, for every
pixel, obtaining each diagonal neighbour
represented by P SE , P SW, P NE and P NW,
requires an extra instruction.
10. • Processing speed of the most low-level image
processing algorithms (image enhancement,
filters, feature detections and so on) can be
accelerated significantly when running on
CPAs.
11. PARALLEL BACKGROUND DETECTION
• Detection process consists of two steps.
Firstly, the distance between the current
image and the background model is
calculated, pixel by pixel.
Secondly, the pixels where that distance is
larger than a pre-set decision threshold are
marked as foreground pixels.
12.
13. CONCLUSION
• A pixel-parallel self-tuning multi-resolution
background detection algorithm has been
presented. This algorithm adapts to the
changing background through a number of
background model updating and parameter
adaptation mechanisms, and implements
simple but effective noise suppression
strategies