User`s guide

Version 2.5 (R2008a ) Video and Image Processing Blockset™
Increased N-Dim
ensional Support
The following bl
ocks now have support for N-D signals:
• Variance
• Standard Deviat
ion
• Mean
• Histogram
• Median
New and Enhanced
Demos
The Video and Ima
ge Processing Blockset contain new and enhanced demos.
New Chroma-Based Road Track ing Demo
This demo uses ch
roma information to detect and track road edges set in
primarily resid
ential settings where lane markings may not be present.
The Chroma-bas
ed Tracking demo illustrates h ow to use the Color Space
Conversion blo
ck, the Hough Transform block, and the Kalman Filter block to
detect and trac
k information using hue and saturation. For more information
see the Chroma
-based Road Tracking Demo.
Video Mosaicking Enhanced
The enhanced V
ideo Mosaicking demo incorporates the new Estimate
Geometric Tra
nsformation block, along with the C orner Detection block,
the Projectiv
e Transformation b lock, a nd the C ompositing block to create a
mosaic image
from a video sequence. The demo accumulates transformation
matrices bet
ween cons ecutiv e video frames. The subsys te m calculates the
transformat
ion matrix between the current and the first video frame, and
then overla
ys the current video frame on to the output image. By repeating
this proces
s, the subsystem generates a mosaic image.
Disparity Estimation for Stereo Vision
The enhance
d D isparity E stimation for Stereo Vision demo uses the Corner
Detection b
lock to find image features. If y ou w ould like to feed your own data
into the de
mo you may wish to e xplo re the various corner detection methods
and other s
ettings of the Corner Detection block.
15