Title: White balance correcting device
Abstract: A white balance correcting device for correcting white balance by using video signals is arranged to be capable of accurately obtaining white balance even in a case where a picture includes only a small white part therein. The white balance correcting device is arranged to divide a picture into a plurality of blocks, to obtain a mean value of signals by averaging signals of each divided block as representing the block and a peak value of luminance from a signal having the highest luminance among others in each divided block also as representing the block, and to use, for white balance control, either the mean value or the peak value according to the state of the object of shooting.
Patent Number: 6,965,401 Issued on 11/15/2005 to Takei
| Inventors:
|
Takei; Hirofumi (Yokohama, JP)
|
| Assignee:
|
Canon Kabushiki Kaisha (Tokyo, JP)
|
| Appl. No.:
|
583251 |
| Filed:
|
May 31, 2000 |
Foreign Application Priority Data
| Jun 04, 1999[JP] | 11-158564 |
| Current U.S. Class: |
348/225.1 |
| Intern'l Class: |
H04N 009/73 |
| Field of Search: |
348/2231,224.1,225.1,655,656,657
382/162,167
|
References Cited [Referenced By]
U.S. Patent Documents
Primary Examiner: Ometz; David L.
Assistant Examiner: Long; Heather R.
Attorney, Agent or Firm: Cowan, Liebowitz & Latman, P.C.
Claims
1. A white balance correcting device for correcting white balance of a picked-up
image signals, comprising:
an image pickup device which picks-up image signals of an image pick-up plane;
a dividing part which divides a scope of the image pick up plane into a plurality
of blocks;
a peak value acquiring part which acquires a peak value of brightness and color
signal values corresponding to the peak value of brightness obtained in each of
all of the plurality of blocks divided by said dividing part;
an average value calculating part which calculates an average value of brightness
and average values of color signal values obtained in each of all of the plurality
of blocks divided by said dividing part;
a comparison part which makes comparison between brightness information of the
average value and the peak value;
a selection part which selects either of the values obtained by said average
value calculating part or the values obtained by said peak value acquiring part
according to comparison result by said comparison part;
a white balance control part which controls white balance on the basis of the
values selected by said selection part; and
wherein said comparison part computes comparison between a first integral value
obtained by integrating average values obtained by said average value calculating
part and a second integral value obtained by integrating peak values obtained in
the scope by said peak value acquiring part, and,
wherein said selection part selects the values obtained by said peak value acquiring
part if the second integral value is not less than a predetermined number of times
the first integral value, and said selection part selects the value obtained by
said average value calculating part if the second integral value is less than the
predetermined number of times the first value.
2. A white balance correcting device according to claim 1, further comprising:
a white determining part which determines whether the average value of color
signal values calculated by said average value calculating part and the color signal
values corresponding to the peak value of brightness acquired by said peak value
acquiring part exist within a white range,
wherein said comparison part integrates values which have been determined to
exist within the white range by said white determining part, in order to obtain
the first integral value and the second integral value.
3. A white balance correcting device according to claim 1, wherein said peak
value acquiring part acquires peak values of image signals from signals that have
beforehand been subjected to limitation for setting an upper limit to a signal
level of the image signals picked-up by said image pick-up device.
4. A white balance correcting device according to claim 1, wherein said peak
value acquiring part acquires the peak value from signals that have beforehand
been subjected by a low-pass filter to limitation for setting an upper limit to
a signal level of the image signals picked-up by said image pick-up device.
5. A white balance correcting device for correcting white balance of a picked-up
image signal, comprising:
an inputting part which inputs picked-up image signal of an image pick-up plane;
a dividing part which divides a scope of the image pick up plane into a plurality
of blocks;
a peak value acquiring part which acquires a peak value of brightness and color
signal values corresponding to the peak value of brightness obtained in each of
all of the plurality of blocks divided by said dividing part;
an average value calculating part which calculates an average value of brightness
and average values of color signal values obtained in each of all of the plurality
of blocks divided by said dividing part;
a comparison part which makes comparison between brightness information of the
average value and the peak value;
a selection part which selects either of the values obtained by said average
value calculating part or the values obtained by said peak value acquiring part
according to the comparison result by said comparison part;
a white balance control part which controls white balance on the basis of the
values selected by said selection part; and
wherein said selection part selects the values obtained by said peak value acquiring
part if the peak value is not less than a predetermined number of times the average
value, and said selection part selects the values obtained by said average value
calculating part if the peak value is less than the predetermined number of times
the average value.
6. A white balance correcting method for correcting white balance of a picked-up
image signals, comprising:
picking-up image signals of an image pick-up plane;
dividing a scope of the image pick-up plane into a plurality of blocks;
acquiring a peak value of brightness and color signal values corresponding to
the peak value from the image signals obtained in each of all of the plurality
of blocks divided in the dividing step;
calculating an average value of brightness and average values of color signal
values from the image signals obtained in each of all of the plurality of blocks
divided in the dividing step;
making comparison between brightness information of the average value and the
peak value;
selecting either of the values obtained in said average value calculating step
or the values obtained in said peak value acquiring step according to comparison result;
controlling white balance on a basis of the values selected in said selection
step; and
wherein, in making comparison, computing a ratio between first integral value
obtained by integrating average values obtained in said average value calculating
step and a second integral value obtained by integrating peak values obtained in
said peak value acquiring step,
wherein, the values obtained in said peak value acquiring step is selected by
said selection if the second integral value is not less than a predetermined number
of times the first integral value, and the values obtained in said average value
calculating step is selected by said selection if the second integral value is
less than the predetermined number of times the first integral value.
7. A white balance correcting method according to claim 6, further comprising:
determining whether the average value of color signal values calculated in said
color average value calculating step and the color signal values corresponding
to the peak value acquired in said peak value acquiring step exist within a white range,
wherein values which have been determined to exist within the white range in
said white determining step are integrated to obtain the first integral value and
the second integral value.
8. A white balance correcting method according to claim 6, wherein peak values
of the image signals are acquired in said peak value acquiring step from signals
that have beforehand been subjected to limitation for setting an upper limit to
a signal level of the image signals picked-up in said image picking-up step.
9. A white balance correcting method according to claim 6, wherein peak values
of the image signals are acquired in said peak value acquiring step from signals
that have beforehand been subjected by a low-pass filter to limitation for setting
an upper limit to a signal level of the image signals picked-up in said image picking-up step.
10. A white balance correcting method for correcting white balance of a picked-up
image, comprising:
inputting a picked-up image signals of an image pick-up plane;
dividing a scope of the image pick up plane into a plurality of blocks;
acquiring peak value of brightness and color signal values corresponding to the
peak value obtained in each of all of the plurality of blocks divided in the dividing step;
calculating an average value of brightness and average values of color signal
values obtained in each of all of the plurality of blocks divided in the dividing step;
making comparison between information of the average value and the peak value;
selecting either of the values obtained in said average value calculating step
or the values obtained in said peak value acquiring step according to comparison result;
controlling white balance on the basis of the values selected by said selection; and
wherein, the values obtained in acquiring a peak value step is selected by said
selection if the peak value is not less than a predetermined number of times the
average value in making comparison, and the value obtained in calculating an average
value step is selected by said selection if the peak value is less than the predetermined
number of times the average value in making comparison.
11. A storage medium which stores therein a program for executing a process for
correcting white balance of a picked-up image signals, said process comprising:
picking-up image signals of an image pick-up plane;
dividing a scope of the image pick up plane into a plurality of blocks;
acquiring a peak value of brightness and color signal values corresponding to
the peak value from the image signals obtained in each of all of the plurality
of blocks divided in the dividing step;
calculating an average value of brightness and average values of color signal
values from the image signals obtained in each of all of the plurality of blocks
divided in the dividing step;
making comparison between brightness information of the average value and the
peak value;
selecting either of the values of obtained in said average value calculating
step or the values obtained in said peak value acquiring step according to comparison
result; and
controlling white balance on a basis of the values selected in said selection
step; and
wherein, in making comparison, computing a ratio between first integral value
obtained by integrating average values obtained in said average value calculating
step and a second integral value obtained by integrating peak values obtained in
said peak value acquiring step,
wherein, the values obtained in said peak value acquiring step is selected by
said selection if the second integral value is not less than a predetermined number
of times the first integral value, and the values obtained in said average value
calculating step is selected by said selection if the second integral value is
less than the predetermined number of times the first integral value.
12. A storage medium according to claim 11, wherein said process further comprises:
determining whether the average value of color signal values calculated in said
color average value calculating step and the color signal values corresponding
to the peak value acquired in said peak value acquiring step exist within a white range;
wherein values which have been determined to exist within the white range in
said white determining step are integrated to obtain the first integral value and
the second integral value.
13. A storage medium according to claim 11, wherein peak values of the image
signals are acquired in said peak value acquiring step from signals that have beforehand
been subjected to limitation for setting an upper limit to a signal level of the
image signals picked-up in said image picking-up step.
14. A storage medium according to claim 11, wherein peak values of the image
signals are acquired in said peak value acquiring step from signals that have beforehand
been subjected by a low-pass filter to limitation for setting an upper limit to
a signal level of the image signals picked-up in said image picking-up step.
15. A storage medium which stores therein a program for executing a process for
correcting white balance of a picked-up image signals, said process comprising:
inputting a picked-up image signal of an image pick-up plane;
dividing a scope of the image pick-up plane into a plurality of blocks;
acquiring a peak value of brightness and color signal values corresponding to
the peak value of brightness obtained in each of all of the plurality of blocks
divided in the dividing step;
calculating an average value of brightness and average values of color signal
values obtained in each of all of the plurality of blocks divided in the dividing step;
making a comparison between brightness information of the average value and the
peak value;
selecting either of the values obtained in said average value calculating step
or the values obtained in said peak value acquiring step according to the comparison
result in said comparison step; and
controlling white balance control part which controls white balance on the basis
of the values selected by said selection step; and
wherein said selection step selects the values obtained by said peak value acquiring
step if the peak value is not less than a predetermined number of times the average
value, and said selection step selects the values obtained by said average value
calculating step if the peak value is less than the predetermined number of times
the average value.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a white balance correcting device, a white balance
correcting method and a storage medium for correcting white balance, which are
suited for use in a video camera.
2. Description of Related Art
White balance correcting devices for video cameras of these days are mainly
arranged to use the outputs of image sensors without using external sensors. Some
of the known white balance correcting devices are arranged to avoid the adverse
influence of chromatic colors as follows. Color-difference signals R-Y and B-Y
and a luminance signal Y obtained from a signal processing circuit are divided
into a number of small blocks corresponding to a picture. The signals within each
of these divided blocks are averaged. Then, color signal components close to white
are extracted from the mean values thus obtained. The white balance is controlled
by bringing the mean values of the extracted color signal components into zero ("0").
FIG. 8 is a block diagram showing, by way of example, the arrangment of an image
pickup apparatus having a conventional white balance correcting device. Referring
to FIG. 8, an object image having passed through a lens
101 and an iris
102 is formed on an image sensor
103. The image sensor
103
outputs signals of primary colors R (red), G (green) and B (blue) obtained by photoelectric
conversion. The R, G and B signals are sent respectively to A/D converters
104,
105 and
106 to be converted into digital signals. The R and B digital
signals are respectively sent to white balance amplifiers
107 and
108
to have their gains controlled on the basis of control signals supplied from a
microcomputer
115. The R and B signals processed by the white balance amplifiers
107 and
108 and the G signal from the A/D converter
105 are
sent to a matrix circuit
109. The matrix circuit
109 is arranged
to form a luminance signal Y and color-difference signals R-Y and B-Y from the
R, G and B signals. The luminance signal Y and the color-difference signals R-Y
and B-Y are sent respectively to D/A converters
110,
111 and
112
to be converted into analog signals. The analog signals thus obtained are sent
to an encoder (not shown) which is arranged to convert these input signals into
standard TV signals. The TV signals from the encoder are sent out from the encoder
either to be displayed on a monitor or to be supplied to a magnetic recording apparatus.
Some of such recording apparatuses are arranged to record these signals in the
form of the digital signals without having them converted into the analog form.
Meanwhile, the signals Y, R-Y and B-Y from the matrix circuit
109
are also supplied to a picture dividing part
113. The picture dividing part
113 is arranged to divide one picture amount of each of the signals Y, R-Y
and B-Y into 8 vertical sections and 8 horizontal sections to give a total of 64
blocks, as shown in FIG. 9. A mean value computing part
114 computes and
obtains the mean value of each of the signals Y, R-Y and B-Y for every divided
block. The 64 sets of the signals Y, R-Y and B-Y are sent from the mean value computing
part
114 to the microcomputer
115. At the microcomputer
115,
only the signals of such blocks that the values of the color-difference signal
and luminance signal are within a certain range are extracted (i.e., the so-called
white extracting process is performed), and only the signals thus extracted are
integrated. This extracting range is as shown, by way of example, in FIG. 10.
FIG. 10 is a diagram showing the variation of vectors of the color-difference
signals taking place as a result of changes in color temperature of an object of
achromatic color with the signal R-Y on the ordinate axis and the signal B-Y on
the abscissa axis. In a case where white balance has been attained when color temperature
is 7000 K which corresponds to an outdoor condition, the color-difference signals
are located at a point P
1 in photo-taking indoors with an incandescent lamp
(at about 3000 K). Conversely, when white balance has been attained indoors with
the incandescent lamp, the color-difference signals are located at a point P
2
in photo-taking outdoors (at about 7000 k). In other words, the changes of color-difference
signals with the color temperature of an achromatic object take place within a
hatched part shown in FIG. 10. Assuming that white balance is to be controlled
within a practicable range of color temperature from 3000 K to 7000 K, the white
balance control can be accomplished by using signals within an area represented
by the hatched part shown in FIG. 10. Hereinafter, this area will be called a white
extracting range. Further, since white balance obtained under the light of a fluorescent
lamp which is tinged with green in spectrum is taken into consideration, a white
extracting area employed generally somewhat spreads in the direction of the color
G. Furthermore, restrictions are sometimes imposed on the luminance signal Y in
addition to the restriction on the color-difference signals. For example, such
a restriction is imposed on the luminance signal Y that the level of the luminance
signal Y is required to be equal to or greater than 50 IRE which is 50% of standard
luminance of the luminance signal Y.
The microcomputer
115 extracts only the signals of blocks in which the
color-difference signals are within the above-stated white extracting area and
the level of the luminance signal is at least 50 IRE, and then computes mean values
of the thus-extracted color-difference signals. Then, the microcomputer
115
corrects the white balance by sending to the white balance amplifiers
107
and
108 such control signals that cause the mean values of the color-difference
signals R-Y and B-Y to become "0".
However, the conventional arrangement described above has presented the
following problems.
While a shooting object having a white portion largely distributed presents
no problem, it is apt to be difficult to accurately extract white from such an
object that has a white part finely distributed on a picture. FIGS. 11 and 12 are
diagrams showing the respective objects on the picture for the purpose of explaining
this problem. FIG. 11 shows the state of a block-divided picture on which a large
image of a person in white clothes appears with a background of chromatic color.
In this case, a white part of the image largely exists within the divided blocks
indicated by arrows in FIG. 11. Therefore, if the color-difference signals are
averaged within each of the divided blocks, white can be accurately extracted.
FIG. 12, on the other hand, shows a small image of the person in white clothes
on the picture. The white part of the image does not much exist within the divided
blocks as indicated by arrows in FIG. 12. Therefore, the white color and the color
of the background are commingled when the signals within each of the divided blocks
are averaged. In that case, the white extraction cannot be accurately accomplished.
Assuming that the background is in a green color of turfs, the white part of the
object and the green of the background mix together to result in a light green
color, which prevents accurate white extraction in the event of the small white
part in the divided blocks as shown in FIG. 12, although the white extraction can
be accurately accomplished in the case of large white parts as shown in FIG. 11.
FIG. 13 is a color-difference signal vector diagram showing how colors are commingling
in the case of the object shown in FIG. 12. The white clothing point Pa and the
green background point Pb are caused to commingle by the averaging of the inside
of each divided block. At the time when the microcomputer
115 reads signals,
the two different colors commingle into a light green color point Pc. Since this
point Pc is located within the above-stated white extracting range, the microcomputer
115 attempts to correct and make this point Pc white. FIG. 14 is a color-difference
signal vector diagram showing a state obtained by the white balance correction.
After the white balance correction, the light green color point Pc has become a
white-balance-corrected point Pc′. The position of the point Pc′
is corrected at the center of vectors. Then, the white clothing point Pa of the
object is erroneously corrected to a point Pa′, and the green background
point Pb is also erroneously corrected to a point Pb′, as shown in FIG.
14. As a result of this white balance correction, the white clothing of the object
comes to be tinged with a purplish color while the green color of the background
becomes a lighter green color. Such an erroneous color correction has presented
a problem.
Another problem of the prior art lies in that, although the above-stated
problem may be mitigated by arranging the picture to be divided into more finely
divided areas, such an arrangement not only causes an increase in size of the circuit
arrangement, but also makes a period of time required for computing processes longer.
BRIEF SUMMARY OF THE INVENTION
In view of the problems of the prior art, it is an object of the invention to
provide a white balance correcting device, a white balance correcting method or
a storage medium for correcting white balance, which are arranged to be capable
of accurately giving apposite white balance even in a case where a picture includes
only a small white part.
To attain the above object, in accordance with an aspect of the invention, there
is provided a white balance correcting device for correcting white balance of a
picked-up image, comprising mean value calculating means for dividing an image
picking-up plane into a plurality of blocks and calculating a mean value of video
signals obtained in each of the plurality of blocks, peak value acquiring means
for acquiring a peak value of video signals obtained in each of the plurality of
blocks, selection means for selecting one of the value obtained by the mean value
calculating means and the value obtained by the peak value acquiring means, and
white balance control means for controlling white balance on the basis of the value
selected by the selection means.
In accordance with another aspect of the invention, there is provided a white
balance correcting device for correcting white balance of a picked-up image, comprising
mean value calculating means for calculating a mean value of inputted video signals,
peak value acquiring means for acquiring a peak value of the inputted video signals,
selection means for selecting one of the value obtained by the mean value calculating
means and the value obtained by the peak value acquiring means, and white balance
control means for controlling white balance on the basis of the value selected
by the selection means.
In accordance with a further aspect of the invention, there is provided a white
balance correcting method for correcting white balance of a picked-up image, comprising
a mean value calculating step of dividing an image picking-up plane into a plurality
of blocks and calculating a mean value of video signals obtained in each of the
plurality of blocks, a peak value acquiring step of acquiring a peak value of video
signals obtained in each of the plurality of blocks, a selection step of selecting
one of the value obtained by the mean value calculating step and the value obtained
by the peak value acquiring step, and a white balance control step of controlling
white balance on the basis of the value selected by the selection step.
In accordance with a still further aspect of the invention, there is provided
a white balance correcting method for correcting white balance of a picked-up image,
comprising a mean value calculating step of calculating a mean value of inputted
video signals, a peak value acquiring step of acquiring a peak value of the inputted
video signals, a selection step of selecting one of the value obtained by the mean
value calculating step and the value obtained by the peak value acquiring step,
and a white balance control step of controlling white balance on the basis of the
value selected by the selection step.
In accordance with a still further aspect of the invention, there is provided
a storage medium which stores therein a program for executing a process for correcting
white balance of a picked-up image, the process comprising dividing an image picking-up
plane into a plurality of blocks, calculating a mean value of video signals obtained
in each of the plurality of blocks, acquiring a peak value of video signals obtained
in each of the plurality of blocks, selecting one of the calculated mean value
and the acquired peak value, and controlling white balance on the basis of the
selected one of the calculated mean value and the acquired peak value.
In accordance with a still further aspect of the invention, there is provided
a storage medium which stores therein a program for executing a process for correcting
white balance of a picked-up image, the process comprising calculating a mean value
of inputted video signals, acquiring a peak value of the inputted video signals,
selecting one of the calculated mean value and the acquired peak value, and controlling
white balance on the basis of the selected one of the calculated mean value and
the acquired peak value.
The above and other objects and features of the invention will become apparent
from the following detailed description of a preferred embodiment thereof taken
in connection with the accompanying drawings.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING
FIG. 1 is a block diagram showing the arrangement of an image pickup apparatus
having a White balance correcting device according to an embodiment of the invention.
FIG. 2 is a block diagram showing the arrangement of a picture dividing part
and the arrangement of a mean value computing part.
FIG. 3A is a block diagram showing the arrangement of a peak value computing part.
FIG. 3B is a block diagram showing the arrangement of a peak value computing
part in which low-pass filters are installed.
FIG. 4 is a block diagram showing the arrangement of a luminance peak detector.
FIG. 5 is a flow chart showing the operation of a microcomputer included in
the image pickup apparatus shown in FIG. 1.
FIG. 6 is an enlarged view showing one block, illustrating a part of an object
image, obtained by dividing a picture into a plurality of blocks.
FIG. 7 is a color-difference signal vector diagram showing color-difference
signals included in a white extracting range.
FIG. 8 is a block diagram showing the arrangement of a conventional image pickup apparatus.
FIG. 9 is a diagram showing a picture divided into a plurality of blocks.
FIG. 10 is a color-difference signal vector diagram showing a white extracting range.
FIG. 11 is a diagram showing an object image and the block-divided picture.
FIG. 12 is a diagram showing another object image and the block-divided picture.
FIG. 13 is a color-difference signal vector diagram showing color-difference
signals to be inputted into a microcomputer of the conventional image pickup apparatus.
FIG. 14 is a color-difference signal vector diagram showing color-difference
signals obtained after the white balance correction in the conventional image pickup apparatus.
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a preferred embodiment of the invention will be described
in detail with reference to the drawings.
FIG. 1 is a block diagram showing the arrangement of an image pickup apparatus
having a white balance correcting device according to the embodiment of the invention.
In FIG. 1, all parts that function in the same manner as the parts of the conventional
apparatus shown in FIG. 8 are denoted by the same reference numerals as in FIG.
8, and the details of them are omitted from the following description. The actions
of a picture dividing part
117 shown in FIG. 1 are first described below.
FIG. 2 is a block diagram showing the arrangement of the picture dividing part
117.
Referring to FIG. 2, a Y selector
203 acts to output a Y signal
input from its terminal Y
0 when one effective horizontal scanning period
begins. A horizontal counter (H counter)
201 is arranged to count the pulses
of a horizontal clock signal and to send a control signal to the Y selector
203
after the lapse of ⅛ of one effective horizontal scanning period. The control
signal causes the Y selector
203 to output another Y signal input from a
terminal Y
1. The Y selector
203 is thus arranged to output the Y
signal inputs one after another by changing use of its output terminals from one
over to another, in such a way as to finish the Y signal output from its terminal
7 at the end of one effective horizontal scanning period.
Upon commencement of the next effective horizontal scanning period, the Y selector
203 again selects the terminal Y
0. In the meantime, a vertical counter
(V counter)
202 counts horizontal scanning lines and outputs, after the
lapse of ⅛ of one effective vertical scanning period, a data output trigger
signal to cause the signals of integrators and registers to be sent to a microcomputer
115 and also a register reset signal for resetting the integrators and the
registers after outputting data.
After the above-stated action, the Y selector
203 acts again to output
a Y signal input from the terminal Y
0 upon commencement of the next one
effective horizontal scanning period. Then, after the lapse of the next ⅛
of the effective vertical scanning period, the data output trigger signal and the
register reset signal are sent out. This action, i.e., a dividing action, is likewise
performed also by an R-Y selector
204 and a B-Y selector
205. With
the signals changed from one signal over to another for one picture in the above
manner, data of one picture is divided into 8×8 blocks, i.e., 64 blocks.
Integrators
210 to
237 shown in FIG. 2 constitute a mean
value computing part
114 shown in FIG. 1. The mean value computing part
114 integrates one block amount of signals supplied from each of the selectors
203,
204 and
205. After the integrated signals are sent as
mean data to the microcomputer
119 in response to the data output trigger
signal, the mean value computing part
114 is reset by the register reset signal.
Block mean value signals to be outputted from the mean value computing part
114 include Y signals AVR(Y(0, n)) to AVR(Y(7, n)), R-Y signals AVR(R-Y(0,
n) to AVR(R-Y(7, n)) and B-Y signals AVR(B-Y(0, n)) to AVR(B-Y(7, n)). In these
signal symbols, "n" represents line numbers in the vertical direction and numerals
represent column numbers in the horizontal direction. A total of 64 block mean
values of one picture are obtained for each of the Y, R-Y and B-Y signals.
FIG. 3A is a block diagram showing the arrangement of a block-luminance-peak-signal
extracting part, which is the peak value computing part
116 shown in FIG.
1. Referring to FIG. 3A, peak detectors
300 to
307 are arranged to
detect the luminance peak of each of the divided blocks. The details of each of
the peak detectors
300 to
307 are shown in FIG. 4. Referring to FIG.
4, a signal inputted to the peak detector is supplied to a luminance limiter
400.
The luminance limiter
400 allows to pass only signals of such luminance
levels that are between levels HiLim and LoLim indicative respectively of the upper
limit and lower limit of the luminance level, which are set under the control of
the microcomputer
119. The upper limit and the lower limit of the luminance
level are set for the purpose of removing such an abnormally high luminance that
color-difference signals cannot be correctly obtained and such a low luminance
that there is little probability of having any white object.
A signal Yin′ having passed through the luminance limiter
400 is
sent to a comparator
401 and a switch
402. The comparator
401
is arranged to receive the output of a register
403 in which the Y signal
which has passed the switch
402 up to the current time is stored. The comparator
401 compares the currently inputted Y signal Yin′ with the level
PK(Y) of the Y signal stored in the register
403. If the level of the Y
signal Yin′ is found to be higher than the level PK(Y), the comparator
401
outputs a switch control signal SPK at a high level. When the level of the switch
control signal SPK becomes high, the switch
402 allows the Y signal Yin′
to pass there and to enter the register
403. Then, the content of the register
403 becomes the same as the value of the signal Yin′. This action
is performed on all luminance signals within the block inputted. As a result, the
register
403 comes to store the largest of luminance signals within each
of the divided blocks when the last horizontal scanning line of the block is obtained.
Meanwhile, the switch control signal SPK is sent also to the outside. The signal
having the highest (largest) luminance among others within each of the divided
blocks is obtained when the level of the switch control signal SPK last becomes
high between the first horizontal scanning line and the last horizontal scanning
line of the same block.
For the Y signal, the peak luminance value of each block is detected by the peak
detectors
300 to
307. For the color-difference signal R-Y, peak detectors
which are similar to the peak detector shown in FIG. 4 are arranged also to output
switch control signals SPK. When the level of the switch control signal SPK
0
becomes high, a switch
310 is caused to allow the R-Y signal R-Y
0
to pass there to be stored in a register
320. Then, at a point of time when
the data output trigger signal is inputted to the register
320, an R-Y signal
of the same part where the largest of luminance values between the first and last
horizontal scanning lines is obtained comes to be stored in the register
320.
This action is performed on each of the R-Y signals R-Y
0 to R-Y
7.
Further, an action similar to the above action is also performed on the color-difference
signal B-Y. The values stored at these registers are sent as peak value data to
the microcomputer
119 according to the data output trigger signals in the
same manner as the mean value computing part
114 mentioned in the foregoing.
After that, the peak value computing part
116 (block-luminance-peak-signal
extracting part) is reset by the register reset signal.
The block peak signals to be outputted from the peak value computing part
116
shown in FIG. 3A include Y signals PK(Y(0, n)) to PK(Y(7, n)), R-Y signals PK(R-Y(0,
n) to PK(R-Y(7, n)) and B-Y signals PK(B-Y(0, n)) to PK(B-Y(7, n)). In these signal
symbols, "n" represents line numbers in the vertical direction and numerals represent
column numbers in the horizontal direction. A total of 64 sets of peak value data
(Y, R-Y and B-Y) are thus obtained per picture.
As mentioned above, with one picture divided into 64 blocks, 64 sets of mean
value
signals and 64 sets of peak value signals are supplied to the microcomputer
119.
Incidentally, as shown in FIG. 3B, low-pass filters
350 to
377
may be inserted in detecting the peak values of the luminance signals and the color-difference
signals. The use of the low-pass filters
350 to
377 as shown in FIG.
3B lessens adverse effects of abnormal luminance and color-difference peak signals
resulting from a malfunction of the image sensor, abnormal reflection, etc.
A flow of processes to be executed within the microcomputer
119 is next
described. FIG. 5 is a flow chart showing the flow of processes in the microcomputer
119.
Referring to FIG. 5, at step S
500, 64 sets of mean value data of
the luminance and color-difference signals and 64 sets of peak value data of the
luminance and color-difference signals are first taken in a memory of the microcomputer
119. At the next step S
501, a block horizontal coordinate value x
and a block vertical coordinate value y for serially processing the 64 sets of
data one by one are reset.
At step S
502, a check is made for the values of the color-difference signal
components AVR(R-Y(x, y)) and AVR(B-Y(x, y)) of the mean value data to find if
they are within the white extracting range. If so, the flow proceeds to step S
503.
If not, the flow proceeds to step S
507.
At the step S
503, a check is made to find if the luminance signal mean
value AVR(Y(x, y)) of the same block is equal to or larger than a predetermined
level. This step is provided for excluding a low luminance value, because any object
that has excessively low luminance is not likely to be white in color. Therefore,
the predetermined level is, for example, set at 20 IRE.
When the results of both the checks made at the steps S
502 and S
503
are YES, the data is considered to satisfy the white extracting condition, and
the flow proceeds to step S
504. At the step S
504, the mean value
AVR(R-Y(x, y)) of the color-difference signal R-Y which has passed the checks is
integrated as integral data. At step S
505, the mean value AVR(B-Y(x, y))
of the color-difference signal B-Y which has been found to meet the white extracting
condition is also integrated as integral data. At step S
506, the mean value
AVR(Y(x, y)) of the luminance signal Y found to meet the white extracting condition
is also integrated as integral data.
Meanwhile, all data found not to meet the white extracting condition at
the steps S
502 and S
503 are not subjected to the integrating processes
of the steps S
504, S
505 and S
506.
At the step S
507, the coordinates of blocks to be taken in for the next
processes are serially incremented one after another. The coordinates are first
incremented, for example, as (x, y)=(0, 0), (1, 0), (2, 0) - - - (7, 0), for the
uppermost line in the block-divided picture shown in FIG. 9. The increment is next
made as (x, y)=(0, 1), (1, 1), (2, 1) - - - (7, 1), for the second line in the
block-divided picture shown in FIG. 9. The process is carried on to eventually
increment the coordinates as (x, y)=(0, 7), (1, 7), (2, 7) - - - (7, 7), up to
the last block coordinates. The steps S
502 to S
507 are executed until
all data of 64 blocks are found by a check made at step S
508 to have been processed.
With data of all the blocks found at the step S
508 to have been completely
processed, the flow proceeds from the step S
508 to step S
509. At
the step S
509, the block coordinates are again reset for processing the
groups of peak value data (x=0 and y=0). At step S
510, a check is made to
find if the data of color-difference signals PK(R-Y(x, y)) and PK(B-Y(x, y) which
correspond to the luminance (Y) peak of each of the blocks are within the white
extracting range shown in FIG. 10. If so, the flow proceeds from the step S
510
to step S
511.
At the step S
511, the peak value PK(R-Y(x, y)) of the color-difference
signal R-Y which has passed the check is integrated as integral data. At step S
512,
the peak value PK(B-Y(x, y)) of the color-difference signal B-Y found to meet the
white extracting condition is integrated as integral data. At step S
513,
the peak value PK(Y(x, y)) of the luminance signal Y found to meet the white extracting
condition is integrated also as integral data. In checking the peak values for
white extraction, the luminance level is not checked to find if it is equal to
or larger than a predetermined level, unlike at the step S
503, because the
luminance level of the Y signal has been restricted to a range between the lower
limit LoLim and the upper limit HiLim by the luminance limiter
400 as mentioned
above (see FIG. 4).
Upon completion of the integrating processes on the peak values of the luminance
signal and color-difference signals which have passed the above-stated check, the
flow proceeds to step S
514. At the step S
514, the block coordinates
are incremented for a block to be next processed in the same manner as the mean
value processing step S
507. The white extracting and integrating processes
are repeated until all the peak value data of 64 blocks are found at step S
515
to have been processed. Upon completion of the processes on all the 64 blocks,
the flow proceeds from the step S
515 to step S
516 for the next process.
At the step S
516, a luminance ratio SR of the integral value AVR
—Y
—P
of peak values of the luminance signal to the integral value AVR
—Y
of mean values of the luminance signal obtained through the white extraction is
computed. At the next step S
517, a check is made to find if the luminance
ratio SR is larger than a reference value SR
—Ref. If so, i.e.,
if the integral value of peak values of the luminance signal is not less than a
predetermined number of times the integral value of mean values of the luminance
signal, the flow proceeds to step S
518. At the step S
518, the integral
values AVR
—RY
—P and AVR
—BY
—P
of peak values of the color-difference signals are substituted for white balance
computing color-difference data RY
—WB and BY
—WB,
respectively. Further, the integral value AVR
—Y
—P
of peak values of the luminance signal is substituted for white balance computing
luminance data Y
—WB.
If the luminance ratio SR is found at the step S
517 to be smaller than
the reference value SR
—Ref, i.e., if the integral value of peak
values of the luminance signal is less than the predetermined number of times the
integral value of mean values of the luminance signal, the flow proceeds from the
step S
517 to step S
519. At the step S
519, the integral values
AVR
—RY and AVR
—BY of mean values of the color-difference
signals are substituted for the white balance computing color-difference data RY
—WB
and BY
—WB, respectively. Further, the integral value AVR
—Y
of mean values of the luminance signal is substituted for the white balance computing
luminance data Y
—WB. With either the mean value data or peak value
data thus stored as the white balance computing color-difference and luminance
data, the flow proceeds to step S
520. At the step S
520, a computing
operation is performed on the data RY
—WB, BY
—WB
and Y
—WB to obtain white balance correcting data. The flow of
processes then comes to an end.
The following describes cases where the use of peak value data is preferable
and where the use of mean value data is preferable. FIG. 6 is an enlarged view
of one block which is indicated with an arrow in FIG. 12 and described as presenting
a problem in the description of the prior art given in the foregoing. In this case,
the ratio of an area occupied within one block by a white object to the whole area
of the block is small. Therefore, the mean luminance value of the block becomes
a low value which is nearly equal to the luminance value YG of green of the background.
The peak luminance value of the block is, however, nearly equal to the luminance
value YW of the white object. In cases where objects of white and chromatic colors
coexist within one and the same picture, the luminance of the object of chromatic
color tends to be lower than that of the white object in general. Particularly,
the luminance of an object of a green or brown color or the like is much lower
than that of a white object. In other words, the ratio of the value YW to the value
YG becomes large in such a case. Hence, when the ratio of the peak luminance to
the mean luminance is at a level equal to or larger than a certain level, it is
highly probable that a part where a peak luminance value is obtained is white.
Therefore, it is preferable that the white balance control is performed by using
signals of a part where the peak luminance is obtained.
On the other hand, in the event of such an object that does not have much difference
between a means luminance value and a peak luminance value, it is assumable that
the divided blocks of the picture do not much vary in color. In such a case, the
white balance control can be more accurately accomplished by using the mean value
of each divided block than by using the peak value which is obtained using a part
of the block.
When this signal selecting method is employed in performing the white extracting
process on such an object that would cause the problem mentioned in the foregoing
in accordance with the conventional method, the integral value AVR
—Y
—P
of peak values of the luminance signal becomes considerably larger than the integral
value AVR
—Y of mean values of the luminance signal. The peak value
signals are, therefore, selected for such an object. The signals to be taken in
the microcomputer
119 under this condition become as shown in FIG. 7. As
shown in FIG. 7, only the signal component Pa for the white object is taken in,
while the green signal component Pb for the background is removed. Therefore, the
white balance control can be correctly accomplished. Incidentally, the advantageous
result can be attained not only for a green background but also for any other background
color such as brown, red or the like, so long as the background has low luminance.
In addition, the threshold value SR
—Ref for the ratio of the
peak value to the mean value does not have to be fixedly set but may be arranged
to be variable.
Further, in the above-described embodiment, the picture of a picked-up image
is divided into a plurality of blocks and the mean value and the peak value of
video signals are obtained for every one of the divided blocks. This arrangement
of the embodiment, however, may be changed to obtain a mean value and a peak value
of one picture without dividing the picture into a plurality of blocks. In the
case of such a modification, a ratio of the peak value to the mean value in one
picture is obtained. Then, the white balance control is performed by selecting
and using the peak value, if the peak value is found to be not less than a predetermined
number of times the mean value. If the peak value is found to be less than the
predetermined number of times the mean value, the mean value is selected and used
for the white balance control.
It goes without saying that the object of the embodiment of the invention is
attainable
also by supplying a system or an apparatus with a storage medium (or a recording
medium) in which software program codes for carrying out the functions of the embodiment
are recorded, and having the program codes read out from the storage medium and
executed by a computer (or a CPU or an MPU) of the system or the apparatus. In
such a case, the functions of the embodiment are carried out by the program codes
thus read out from the storage medium. Then, the storage medium which stores the
program codes therein is to be considered to constitute the invention. Further,
the program codes read out are not only executed by the computer to carry out the
functions of the embodiment but also may be executed either in part or in their
entirety on the basis of instructions included in the program codes by an operating
system (OS) or the like working on the computer.
In addition, the scope of the invention of course includes also a case where
the
functions of the embodiment are carried out by writing the program codes read out
from the storage medium into a memory provided in a function extending card inserted
into a computer or a function extending unit connected to the computer, and, after
that, executing actual processes either in part or in their entirety on the basis
of instructions of the program codes with a CPU or the like provided in the function
extending card or the function extending unit.
In a case where the arrangement of the embodiment is to be applied to a storage
medium in the above-stated manner, the storage medium is arranged to store therein
program codes corresponding to the processes described in the foregoing with reference
to the flow chart of FIG. 5.
According to the arrangement of the embodiment as described above, a white
component can be correctly and reliably extracted without being affected by the
state of a background, even in a case where an object of shooting includes only
a small white part, so that white balance can be accurately corrected. Besides,
in a case where the object of shooting includes a large white part, the embodiment
is of course capable of carrying out white balance correction in the same manner
as the conventional white balance correcting arrangement.
*
