DATAMANIP commands
PROGRAM NAME
kconvolve - Simple convolution and correlation
DESCRIPTION
kconvolve is a simple interface to the more general klinearop
routine in the DATAMANIP toolbox. kconvolve is set up to perform
convolution or correlation on 5D data using a general 5D kernel,
in zero-padded, linear mode (not circular).
The kernel origin, center, or "hotspot" can be set using the -wc, -hc,
-dc, -tc, and -ec arguments; otherwise the KPDS_KERNEL_ORIGIN attribute for
the kernel object is queried for that information.
REQUIRED ARGUMENTS
- -i1
-
type: infile
desc: input object
- -i2
-
type: infile
desc: kernel
- -o
-
type: outfile
desc: output object
OPTIONAL ARGUMENTS
- -corr
-
type: boolean
desc: Select correlation (default) or convolution
default: false
- -upcast
-
type: flag
desc: Upcast data to double or double complex for processing
- -wc
-
type: integer
desc: Width coordinate of kernel hotspot
default: 0
bounds: value >= 0
- -hc
-
type: integer
desc: Height coordinate of kernel hotspot
default: 0
bounds: value >= 0
- -dc
-
type: integer
desc: Depth coordinate of kernel hotspot
default: 0
bounds: value >= 0
- -tc
-
type: integer
desc: Time coordinate of kernel hotspot
default: 0
bounds: value >= 0
- -ec
-
type: integer
desc: Elements coordinate of kernel hotspot
default: 0
bounds: value >= 0
EXAMPLES
SEE ALSO
kfft(1), klinearop(1)
RESTRICTIONS
REFERENCES
COPYRIGHT
Copyright (C) 1993 - 1997, Khoral Research, Inc. ("KRI") All rights reserved.