 |
Overview |
SNVision Library(.dll) is the true core of our technology.
It is structured as a software library with a set of 50 functions
for image processing and analysis. It is delivered as a Win32
DLL for Windows 9x, ME, NT4, 2000 or XP to use with C/C++,
.NET, Visual Basic, LabWindows CVI languages and provided
with "how to use" examples.
SNVision Library is the runtime module which enables you
to include Spikenet Technology into your application. Contact
us for a proposal.
|
Key
Features |
- Real time detection of
objects: extremely fast processing (25 pictures per second)
- Robustness: tolerates rotations
and changes in object scale or resolution
- Great tolerance towards
noise and light conditions
- High speed detection of
multiple targets
- Easily handles large model
database
- Ability to locate contiguous
and overlapping objects
- Lack- and inclusion- proof
adjustable associativity
- Adjustable selectivity
- Small footprint: Compact
core < 550 KB
- Linear and limited memory
use: < 32 M
- Easy-to-parallelize
architecture
SpikeNet Technology works for any kind of image: video images
(single video processing as well as multiple video stream
processing), finger prints, satellite images, x-rays, microscopic
slides.
|
Technical
specifications |
System requirements |
- Windows 9x, ME, NT4, 2000 or XP
- Pentium3 (800 MHz) Processor or higher
- 32 MB RAM
|
|
Performances |
SNVision Library's processing time is extremely fast even
with large images. Processing times provided below were performed
using a Pentium4 2.2 GHz and using one 30x30 pixel model.
Input image
(pixels) |
Elapsed
time (ms) |
Million pixels
per second
|
75x60 |
0.8 |
12 |
150x120 |
2.1 |
30 |
300x240 |
7.3 |
48 |
600x480 |
30.6 |
57 |
1200x960 |
120 |
45 |
2400x1920 |
394 |
49 |
Elapsed time per input image:
SNVision Library handles very large data bases. Processing
time increases linearly with the number of models included
in the data base.
|
Models number |
Elapsed time
(ms) |
1 |
18 |
10 |
44 |
20 |
71 |
30 |
100 |
40 |
128 |
50 |
158 |
60 |
189 |
70 |
217 |
Elapsed time per Models number:
|
