![](http://datasheet.mmic.net.cn/370000/TMC2301_datasheet_16742305/TMC2301_7.png)
PRODUCT SPECIFICATION
TMC2301
7
Transformation Control Parameters
The TMC2301 is a self-sequencing device which requires no
cycle-to-cycle intervention from the host system. To program
the device, the user loads the 16 operating parameters, which
define the transformation to be performed, which sections of
the original and resampled image spaces are to be utilized,
and various control words. Filtering operations are further
defined by the values the user loads into the external coeffi-
cient memory. The transform parameters are described
below. See also Tables 1 through 3.
Note:
1. For each incremental move along the U axis, the starting
point of the new "walk around spiral" is indexed to the
ENDING point of the previous walk around spiral, rather
than to its center. Therefore, the terms dX/dU
0
and
dY/dU
0
must be adjusted accordingly. Since each new line
is referenced back to the previous line's initial spiral
starting point, no similar dX/dV
0
or dY/dV
0
correction is
needed.
XMIN, XMAX,
YMIN, YMAX
These four parameters outline the
"source" rectangular region of the
original image. Whenever the IRS
pair generates an (X, Y) address
within this boundary the CZERO
flags will denote a valid memory
read. In the most common case,
XMIN < XMAX, YMIN < YMAX, 000h
< X < FFFh, and 000h < Y< FFFh. In
this case, addresses out of bounds
cause one or both CZEROs to go
LOW. Refer to Application Note
TP-38 for further information on other
boundary violation cases. Each
parameter is expressed in 12-bit
unsigned binary integer notation.
See Figure 12.
These four parameters outline the
"target" region of the (u, v) plane, into
which the resampled image will be
written. The IRS will generate, line by
line, a scan that fills only this portion
of the plane, permitting the user to
assemble a mosaic of multiple
rectangular subimages. Care must
be taken to ensure that UMAX >
UMIN and VMAX > VMIN. Each
parameter is expressed in 12-bit
unsigned binary integer notation.
See Figure 12.
These are the coordinates of the first
pixel to be read from the original
image. In many applications, this
point will be one of the four corners
of the original image to be
resampled. The pixels near (X
0
, Y
0
)
in the original image will be used to
compute the upper left pixel of the
transformed image. In non-inverting,
non-reversing applications (X
0
, Y
0
)
will be the upper left corner of the
original subimage. Each coordinate
is expressed in 13-bit integer plus
5-bit fraction, two's complement
notation.
UMIN,
UMAX,
VMIN, VMAX
(X
0
, Y
0
)
dX/dU
0
Is the initial horizontal partial first
derivative indicating the
displacement along the X axis which
corresponds to each one pixel
movement along the U axis. Usually,
0 < dX/dU
0
< 1 corresponds to
magnification, whereas dX/dU
0
> 1
represents reduction and dX/dU
0
< 0
denotes reflection about a vertical
axis. The first derivatives are
expressed in 8-bit integer, 12-bit
fraction two's complement notation.
Is the initial horizontal-vertical partial
first derivative. It indicates the
displacement along the X axis
corresponding to each one pixel
movement along the V axis. The
coefficients dX/dV
0
and dX/dU
0
define image rotation and shear.
Is the initial vertical-horizontal partial
first derivative. It indicates the
displacement along the Y axis
corresponding to each one pixel
movement along the U axis.
Is the initial vertical partial first
derivative. It indicates the
displacement along the Y axis
corresponding to each one-pixel step
along the V axis. Since dX/dV
0
and
dY/dV
0
are separate parameters,
vertical magnification and reflection
need not match their horizontal
counterparts.
dX/dV
0
dY/dU
0
dY/dV
0
d
2
X/dU
2
Is the second order horizontal
derivative. It indicates the rate of
change of the horizontal-horizontal
first derivative with each step along a
line in the output image space. All six
second-order derivatives are 4-bit
integer, 20-bit fractional two's
complement parameters.
Is the second order horizontal-
vertical-vertical derivative. It
indicates the rate of change of the
horizontal-vertical first derivative with
each step down a column in the
output image space.
d
2
X/dV
2