Title: Hitachi Camera Signal Processing LSI HD49815TF
1Hitachi Camera Signal Processing LSI (HD49815TF)
AWB Data Detection Report
2System Block Diagram
- Hitachi Camera System Configuration
(1)
(2)
We emphasize on AWB!
(3)(4)
3System Block Diagram(1)
- Color Signal Processing Block Diagram
4System Block Diagram(2)
- Luminance Signal Processing Block Diagram
5System Block Diagram(3)
6System Block Diagram(4)
- AWB detection circuit block diagram
7AWB data obtain
- For AWB detection,data is obtained from axes R-B
and Mg-G.The detection range is an area other
than around the origin on the vector scope and
can be moved in accordance with photography
conditions. - By controlling the white convergent point,AWB
suitable for various photography conditions can
be implemented.Input signals are a color
difference signal (8 bits offset binary,D range
-128127)and a luminance signal(for reference,5
bits,after gamma,negative limit).
8Color difference signals(1)
- Color-difference Matrix Circuit Block Diagram
- This block transform R-G B-G signal to R-Y
B-Y form.
9Color difference signals(2)
- k1Coefficient k1 depends on the balance of the
three primary colors. First we calculate from
Y0.59G0.3R-0.11B.Therefore R-YR(0.59G0.3R0
.11B)0.7(RG)0.11(B-G)B-YB (0.59G0.3R0.11B)
0.89(B-G)-0.3(R-G) - k2The signal level of B-Y becomes higher than
that of R-Y.To prevent this,the broadcasting
standard specifies 0.493x(B-Y) and
0.877x(R-Y).The chroma signal saturation balance
is uniform.
k1
k2
10White detection range concept
11Color-difference Plane (Vector Diagram)
12Offset Data
- Indicates the detection center of each axis.The
counter decides to count up or down depending on
whether data is greater or less than the offset
value.Offset R-B and Mg-G vary independently. - In the color difference plane,an offset can be
applied to the R-B Mg-G axes.The offset range
is 32 to 31 with color difference signals 128
to 127,and white convergent pointcan be changed
by changing this data.In MV-SF20,it is moved
according to the color temperature.
13Slope of Offset to Luminance
- Y moves parallel to the offset Y axis.The basic
expression is as follows (OFS_DEC) Y ( offset
) OFFSET - Data format of OFS_DEC 1(sign)6(absolute
value) 7 bits
14Two extreme conditions of Offset
- (1) only OFS_DEC is used (offset is 20H)(2) only
offset is used (OFS_DEC is 00H)
15White Detection Range(1)
- If Y exceeds 100 IRE,the Y signal is
limited,holding the white detection range
constant. - White Detection Range a(slope) x Y0 /32
b(constant parameter) - Parameters a and b can be individually set on the
positive and negative sides of R-B and Mg-G.
16White Detection Range(2)
- Slope A of the White Detection RangeData format
1(sign)7(absolute value)
Center point A of C_LEVEL is the origin when Y is
0
17White Detection Range(3)
- Base B of White Detection RangeData format
1(sign)7(absolute value)
a0
Settings combining a and b cause an overflow to
occur due to the setting range extension of
C_LEVEL b.Therefore,C_LEVEL is limited to
FFH(11111111) on the side and 00H(00000000) on
the negative side.
18White Balance Data of Input Output
- Input DataR-B Mg-G DataOutput Data
indicating where the majority of the data
lies.11-bit data counts for R-B and Mg-G are
output every field (output data22bits).The
output data is transferred to the output latch
and cleared.
?????????,???????