home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
SGI Hot Mix 17
/
Hot Mix 17.iso
/
HM17_SGI
/
research
/
examples
/
demo
/
demotext
/
imagproc.txt
< prev
next >
Wrap
Text File
|
1997-07-08
|
3KB
|
100 lines
WELCOME TO THE IMAGE PROCESSING DEMO
Image processing can improve the way an image
looks. Image-processed data, or a processed image
helps you distinguish and recognize more subtle
characteristics of the original image.
MENU OPTIONS
------------
File Menu:
Select "Quit" to exit the Image Processing
Demo and return to the IDL Demo main screen.
About Menu:
Select "About image processing" for information
about the Image Processing Demo.
FEATURES
--------
IDL offers several ways to process an image,
some of which are introduced below:
o Zooming
Smaller sub-sections of large datasets can be
manipulated. The image shown here is an
aerial view of New York City. A section of
lower Manhattan is enlarged using the REBIN
function on a subscripted section of the
original image array.
To reposition the zooming area, click your
mouse on the area you would like to see more
closely.
o Fourier filtering
IDL's Fast Fourier Transform (FFT) function
can process vectors and 2D array in either
forward or reverse.
<<Filter width>> slider
Selects the filter width. A higher filter
width gives the filtered image higher
resolution.
o Pixel scaling
The BYTSCL (byte scale) command scales pixel
values into the range of available colors.
<<Minimum>> slider
You can specify the minimum byte value
allowed in the reconstructed image.
<<Maximum>> slider
You can specify the maximum byte value
allowed in the reconstructed image.
o Histogram
Histogram equalization can be used to
change the visible contrast of an image.
<<Minimum>> slider
You can specify the minimum byte value
allowed in the reconstructed image.
<<Maximum>> slider
You can specify the maximum byte value
allowed in the reconstructed image.
o Edges
The SOBEL function performs edge enhancement
on an images.
<<Smooth width>> slider
Sets the width of the smoothing function.
A higher width produces a reconstructed
image with lower resolution.
o Convolution
Convolutions may be performed on arrays of
1, 2, and 3 dimensions. This screen shows
the effects of convolving an image with
a kernel.
Use your left mouse button to click on
the squares of the kernel grid, toggling
them on or off.
<<CONVOLVE>> slider
Computes and displays the convolved image.