Data Structures
class	reg
	operand of this type can only be an ipu register More...

class	r_int
	operand of this type can be either an ipu register or a constant integer of range [0,255] or [-128,127] More...

class	r_int16
	similar to r_int, but it can be assigned with a constant integer of range [0,65535] More...

class	port
	operand of this type can only be a port address More...

Functions
	_mov (reg d, r_int s)
	move data in to a register More...

	_or (reg d, r_int s)
	bitwise OR operation More...

	_and (reg d, r_int s)
	bitwise AND operation More...

	_add (reg d, r_int s)
	add two data ( d = d + s ) More...

	_sub (reg d, r_int s)
	subtract two data ( d = d - s ) More...

	_jump (label l)
	jump the program to a labeled position More...

	_jump (condition f, label l)
	jump the program conditionally to a labeled position More...

	ipu_delay_50us_x (r_int t)
	time delay with a resolution of 50 us More...

	ipu_wait_frame_trigger ()
	wait until a frame trigger is detected More...

	ipu_pixel_set (dreg b1, r_int x, r_int y)
	set one pixel in a DREG plane More...

dreg	ipu_pixel_select (r_int x, r_int y, r_int mx, r_int my)
	set all pixel a DREG plane of which their coordinates match the rules More...

	ipu_pixel_filter (dreg b1, r_int x, r_int y, r_int mx, r_int my)
	similar to ipu_pixel_select, but the result will be ANDed with the DREG More...

reg	ipu_global_sum (areg a, r_int k)
	sum all pixels in a AREG plane, then sample More...

Detailed Description

Author: Jianing Chen

Function Documentation

_mov	(	reg	d,
		r_int	s
	)

move data in to a register

Parameters

d	the destinated register
s	the data to be moved

Example Usage: _mov(s0,0) _mov(s0,15) _mov(s3,s4)

Flagging:

set Z if s is equal to the data originally stored in d

_or	(	reg	d,
		r_int	s
	)

bitwise OR operation

Parameters

d	the destinated register, LHS of the OR operation
s	RHS of the OR operation

Example Usage: _or(s0,128) _or(s0,s1)

Flagging: Z.

_and	(	reg	d,
		r_int	s
	)

bitwise AND operation

Parameters

d	the destinated register, LHS of the AND operation
s	RHS of the AND operation

Example Usage: _and(s0,15) _and(s0,s1)

Flagging: Z.

_add	(	reg	d,
		r_int	s
	)

add two data ( d = d + s )

Parameters

d	the destinated register, LHS of the addition
s	RHS of the addition

Example Usage: _add(s0,10) _add(s1,s2)

Flagging: Z, C.

_sub	(	reg	d,
		r_int	s
	)

subtract two data ( d = d - s )

Parameters

d	the destinated register, LHS of the subtraction
s	RHS of the subtraction

Example Usage: _sub(s0,1) _sub(s1,s2)

Flagging: Z, C.

_jump ( label l )

jump the program to a labeled position

Parameters

l	a labeled position in the program

Example Usage:

 #start:
 
 ...
 
 _jump(#start)

_jump	(	condition	f,
		label	l
	)

jump the program conditionally to a labeled position

Parameters

f	the condition, can be either of c,nc,z,nz
l	a labeled position in the program

For the condition part, the Z or C flag produced by previous instructions will be checked.

Example Usage:

The following structure is a loop iterating 10 times:

 ...
 
 _mov(s0,10)
 #loop_begin:
 
 ...//loop body
 
 _sub(s0,1)
 _jump(nz,#loop_begin)
 
 ...

ipu_delay_50us_x ( r_int t )

time delay with a resolution of 50 us

Parameters

[in] t (t*50) is the number of micro seconds delayed

Example Usage:

Take a image with an exposure time of 500 us:

 rpix
 ipu_delay_50us_x(10)// wait for 500 us
 rpix(B,C)

ipu_wait_frame_trigger ( )

wait until a frame trigger is detected

The frame trigger signal is a series of pulse with a frequency specified by host_set_frame_rate.

The following structure provide a frame of which the exposure time is the interval between two frame triggers:

 ipu.wait_frame_trigger
 rpix
 ipu.wait_frame_trigger
 rpix(B,C)

ipu_pixel_set	(	dreg	b1,
		r_int	x,
		r_int	y
	)

set one pixel in a DREG plane

Parameters

[in,out]	b1	the DREG plane
[in]	x	x coordinate of the pixel
[in]	y	y coordinate of the pixel

Example Usage:

Make (30,50) and (50,50) in R5 DREG plane to be 1:

 R5 = 0 // do not forget to clear the plane before putting new points
 ipu.pixel_set(R5,30,50)
 ipu.pixel_set(R5,50,50)

dreg ipu_pixel_select	(	r_int	x,
		r_int	y,
		r_int	mx,
		r_int	my
	)

set all pixel a DREG plane of which their coordinates match the rules

Parameters

[in]	x	x coordinate match of the pixels
[in]	y	y coordinate match of the pixels
[in]	mx	x coordinate mask of the pixels
[in]	my	y coordinate mask of the pixels

Returns: the result in DREG plane

Example Usage:

For R5 DREG plane, make all pixel with x and y coordinates in in the format "xxxxx101" (binary form) to be 1:

1 R5 = ipu.pixel_select(5,5,248,248)

Note:

Binary form of 248 is 11111000, which is used as the coordinate mask.

Binary form of 5 is 00000101. which is used as the matching part.

ipu_pixel_filter	(	dreg	b1,
		r_int	x,
		r_int	y,
		r_int	mx,
		r_int	my
	)

similar to ipu_pixel_select, but the result will be ANDed with the DREG

Parameters

[in]	b1	the selection result will be ANDed with this DREG
[in]	x	x coordinate match of the pixels
[in]	y	y coordinate match of the pixels
[in]	mx	x coordinate mask of the pixels
[in]	my	y coordinate mask of the pixels

Returns: the result in DREG plane

Example Usage:

Make a rectangular with bottom left corner on (64,64) and top right corner (127,127) in R5:

 R5 = 1 // becuase the result will be ANDed so the initial value should be 1
 ipu.pixel_filter(R5,0,0,191,191)
 ipu.pixel_filter(R5,128,128,127,127)

Notice that it is very difficult to make up these parameters in mind. Instead, it is more feasible to try some combinations using a program like Example 10: DREG Selecting.

Another example:

Compose a rectangular of size 128x128 in the center of R5 DREG plane:

 R5 = 0
 
 R6 = 1
 ipu.pixel_filter(R6,0,0,191,191)
 ipu.pixel_filter(R6,128,128,127,127)
 R7 = R5
 R5 = or(R6,R7)
 
 R6 = 1
 ipu.pixel_filter(R6,64,0,191,191)
 ipu.pixel_filter(R6,0,128,127,127)
 R7 = R5
 R5 = or(R6,R7)
 
 R6 = 1
 ipu.pixel_filter(R6,64,64,191,191)
 ipu.pixel_filter(R6,0,0,127,127)
 R7 = R5
 R5 = or(R6,R7)
 
 R6 = 1
 ipu.pixel_filter(R6,0,64,191,191)
 ipu.pixel_filter(R6,128,0,127,127)
 R7 = R5
 R5 = or(R6,R7)

reg ipu_global_sum	(	areg	a,
		r_int	k
	)

sum all pixels in a AREG plane, then sample

Parameters

[in]	a	the AREG plane
[in]	k	the ADC gain setting

Returns: the sampling result

The ADC gain value need to be determined through testing.

Example 11: Analogue Regional Summation demonstrates how to sum only a specific region in a AREG plane.

Example Usage:

Sum all pixels in A with a ADC gain of 15, store the sampling result in s0:

1 s0 = ipu.global_sum(A,15)

Data Structures

Functions

Detailed Description

Function Documentation