LIBRARY ROUTINE
ligradient - perform image differentiation using gradient operator
LIBRARY CALL
int ligradient (
kobject src,
int gradient_type,
int operation_type,
kobject dest )
INPUT
- src - the source data object to be processed
- gradient_type - type of gradient calculation
- operation_type - absolute values (default) or squares and
square root operations
OUTPUT
- dest - the destination data object
RETURN VALUE
TRUE (1) on success, FALSE (0) on failure
DESCRIPTION
takes an image and performs the differentiation (high pass
spatial filtering or edge enhancement) operation by using
standard gradient operators. The standard gradient operators
supported by ligradient are: Robert's, Sobel, Prewitt and
Isotropic. The Robert's operators use a 2x2 kernel, whereas
the other three use a 3x3 kernel. Each operator uses two
templates to enhance edges in the horizontal and vertical
directions only. For edge enhancements in other directions
other operators should be used.
There are two approximations to the true gradient. The user
can select between calculating the magnitude of the gradient
by using squares and square roots (Euclidean distance
evaluation), or by using absolute differences. The latter
is computational faster.
Mask data is ignored.
ADDITIONAL INFORMATION
none
EXAMPLES
none
SIDE EFFECTS
none
RESTRICTIONS
COMPLEX data processing is not supported
MODIFICATION
none
FILES
$IMAGE/objects/library/kimage_proc/src/ligradient.c
SEE ALSO
kimage_proc(3)
COPYRIGHT
Copyright (C) 1993 - 1997, Khoral Research, Inc. ("KRI") All rights reserved.