~4Dgifts/toolbox/src/exampleCode/video/simpleVideo README
**********************************************************************
SV - Simple Video
"Video so simple, the programs almost write themselves"
**********************************************************************
This library exists to provide the video programmer with a simple
programmatic interface on top of the video library(VL). It provides
functions such as getframe() and putframe() for dealing with single
frame input/output, while requiring as little additional knowledge of
the VL possible. Additional utility routines provide control over some
of the underlying behaviour of the VL. To those familiar with the VL,
SV hides the details of devices, nodes, paths, and ring buffers. It
defines an image data type svImage for dealing with frames of data.
Features
--------
* Simple: No need to know the details of devices, nodes,
paths, and ring buffers;
* Frame I/O: capture frames, send frames to video output,
pastes frames to screen;
* Read/Write compressed images of all types;
* Define simple, safer macros for setting some VL controls;
* Mutiple contexts to support concurrent transfers;
Simple Single Frame Video Input to Video Output
-----------------------------------------------
Suppose we would like to write a simple program to grab a frame
of video and send it back to our video output:
main()
{
getFrame();
putFrame();
}
There, that did it. Except for one a few small details:
* How do we select the video input and output sources, and
* How do we represent the image?
To manipulate the frames of data we're moving around, we define the
following functions to deal with images:
svImage *svNewImage(void);
void svFreeImage(svImage **);
To grab a frame of video or send an image to video output, we
use the functions:
int svGetFrame(svImage *);
int svPutFrame(svImage *);
To select our source an destination, we use the functions:
void svSelectInput(int);
void svSelectOutput(int);
So the full length version of our single frame video
input to video output goes as follows:
main()
{
svImage *image;
svSelectInput(gvINPUT_COMPOSITE_1);
svSelectOutput(gvOUTPUT_ANALOG);
image = svNewImage();
svGetFrame(image);
svPutFrame(image);
svFreeImage(&image);
}
Selecting a Video Input or Output
---------------------------------
As mentioned above, the function svSelectinput(int) is used to select the
video input for svGetFrame() and the function svSelectOutput(int) is
used to selected the video output for svPutFrame(). The list of
valid inputs and outputs are:
/*
* Indy connections -
*/
#define vnINPUT_INDYCAM 1
#define vnINPUT_COMPOSITE 2
#define vnINPUT_SVIDEO 3
/*
* Galileo connections -
*/
#define gvINPUT_COMPOSITE_1 4
#define gvINPUT_SVIDEO 5
#define gvINPUT_COMPOSITE_2 6
#define gvINPUT_DIGITAL_1 7
#define gvINPUT_DIGITAL_2 8
#define gvOUTPUT_ANALOG 9
#define gvOUTPUT_DIGITAL 10
What Else Can You Do With The Image Once You've Got Your Hands On It?
---------------------------------------------------------------------
We've introduced the functions svGetFrame(svImage *) and
svPutFrame(svImage *) for getting images from video input sending
images to video output; Here is a complete list of useful functions
for dealing with images:
svImage *svNewImage(void);
- allocate a new svImage;
void svFreeImage(svImage **);
- free the image;
int svSaveImage(char *filename, svImage *frame);
- save the image to a named file;
int svLoadImage(char *filename, svImage **frame /* return */);
- load in image from a named file;
int svViewImage(svImage *frame, int x, int y);
- paste an rgb image to the screen;
void svCompressedImages(int saveCompressed);
- enable writing of compressed images;
void svSetImagePacking(int packingType);
- set the VL_PACKING control;
What else can I do with SV?
---------------------------
Here are some other utility functions that let you modify the
behaviour of the underlying VL code.
*** Configure the path and control usage mode -
You can control the usage mode using the functions
void svNodeAccessMode(VLUsageType);
void svControlAccessMode(VLUsageType);
To set the usage mode on the underlying path path.
The default is to setup the path with a mode of VL_SHARE.
*** Recover from preemption -
Since by default the path is configured in VL_SHARE mode,
it will be preempted by another video program. Use the
function
svRecoverFromPreemption()
to tell SV to attempt to re-setup the path it is using
when it is preempted.
*** Set the frame count and transfer mode -
void svSetFrameCount(int); /* default 1 */
void svSetTransferMode(int); /* default svTRANSFER_DISCRETE */
I Don't Understand, Where Can I Find Examples?
----------------------------------------------
There is sample source code for capturing frames,
pasting frames to the screen, sending a series of frames
to video out, and continuous transfer of frames from
the vino video input to the galileo video output.
Look in the source code in the examples subdirectory.
Don Bennett
dpb@sgi.com
Digital Media Engineering
Documentation
Reference
Subdirectories
OR, ...