JP2007082174A - White balance adjustment circuit, its adjustment method and camera apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform appropriately AWB adjustment at the time of switching between a high color temperature side and a low color temperature side in a feeding forward method. <P>SOLUTION: There are provided a primary color separation circuit 15, an integrating means (PD17), a controller 18 and variable gain amplifiers 16r, 16g and 16b. For example, the controller 18 comprises: a means for carrying out the coordinate conversion of the integral value for every color; a means for detecting color temperature transition of pictures between the high color temperature side and the low color temperature side based on information about brightness or color temperature of two or more pictures; and a gain setting means for setting up a gain value to every primary color signal by drawing a coordinate value or a presumed color temperature value in a predetermined desired value. The gain setting means makes it possible to set up a dead zone within a frame wherein color temperature adjustment is inhibited. The position, range and validity of the dead zone can be set up independently from the high color temperature side and the low color temperature side on the color coordinates. Based on the detection result of color temperature transition, the validity or invalidity of the dead zone is determined by the color temperature transition side. According to the result of the determination, the drawing judgement is performed only in a case of the validity of the dead zone at the color temperature transition side. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a white balance adjustment circuit, an adjustment method thereof, and a camera device for correcting a white portion with a tint to an achromatic white color for an image signal obtained by processing a signal generated by an image sensor. .

  Shooting by the camera device is performed under various light sources such as natural light (sunlight), incandescent light, and artificial light such as fluorescent light. At this time, when the type of light that illuminates the subject changes, the color of the subject slightly changes. Therefore, the color of the same subject is different under the imaging environment. White balance adjustment is color balance correction that matches white that appears colored by the difference in light source to achromatic white.

  As a white balance control method, an integral value is obtained for each field from R (red), G (green), and B (blue) primary color signals, and the integral value data is converted (RB). ) / G, (R + B−2G) / G, for example, a method for estimating the color temperature of natural light on the high color temperature side and setting the white balance gain by drawing the estimated color temperature into the target value (For example, refer to Patent Document 1).

Here, “estimation” means that the integrated value data is converted into color coordinate integrated value data with (R−B) / G and (R + B−2G) / G as axes, and the converted coordinate points. (Color difference data) is moved on a black body radiation curve according to a predetermined rule, and this is estimated as the color temperature of the light source and used as the starting point of the pull-in. “Color temperature” is, according to a general definition, “When a virtual object (black body) that does not reflect any light is heated, the color of the black body is expressed by the temperature (K) of the black body” However, in the field of image processing, chromaticity in which white changes from an achromatic white under a test light source in an image sensor or camera device is expressed in correspondence with the color temperature of a black body. The “black body radiation curve” is obtained by drawing a change curve of the color temperature of natural light on predetermined color coordinates.
“Pulling” refers to an operation of extracting a color balance adjustment amount (gain ratio) from the amount of change of each color component obtained when the estimated color temperature is moved to a predetermined target value. Regarding the reason for obtaining the color temperature of the imaging screen from the integral value for each field in the above method, there are various color components on the entire screen when imaging a general subject, but all the color components of the entire screen are integrated. When one color difference data on the color coordinate is obtained, the color difference data reflects the balance of RGB of the entire screen, and this is approximated as representing white data of the screen under the imaging light source. Because it can be guessed.

Patent Document 1 describes that a high color temperature side pull-in limit frame and a low color temperature side pull-in limit frame are provided.
The pull-in limit frame is a frame that determines the pull-in limit that the pull-in is performed when the color difference data is in the inner range and is not performed in the outer range. As for the reason for providing two pull-in limit frames in this way, if one frame is provided along the black body radiation curve, the color reproducibility is reduced by changing the color of the image portion with the original color such as green. Therefore, two frames each having a limited frame range are provided so as not to cause this.

  Since the color balance adjustment amount can be extracted by this pull-in operation, the gain (amplification or attenuation factor) of the variable gain amplifier corresponding to each of the three primary color signals is set based on the extracted color balance adjustment amount. The primary color signal output from the variable gain amplifier after the gain change is substantially balanced in R, G, and B. Therefore, in the image obtained by synthesizing the primary color signals output from the variable gain amplifier, if there is a display area representing white, the color of that area is corrected to fit under the imaging light source, and almost achromatic white and Become.

By the way, the white balance adjustment method includes a feedback method that sets a gain value based on the output of a variable gain amplifier and a feedforward method that sets a gain value based on a primary color signal before being input to the variable gain amplifier. Yes (see Patent Document 1).
Of these, the feedback method makes it easy to predict the gain value, but there is a possibility of oscillation because the control is not an open loop. In order to prevent this, a technique is known in which a dead zone indicating a prohibited range of color temperature adjustment is provided around the target value of the pull-in operation (see, for example, Patent Documents 2 and 3).
The dead zone for this purpose is set as a relatively small frame around the target value, so that white balance is automatically adjusted by disturbances such as minute transformation of the coordinate system and noise, preventing the balance point from shifting and improving stability. It is possible to keep.

By the way, while the subject is changed and projected in the shooting environment of the same light source, the white balance is hardly shifted greatly. However, in the case of moving from the room (generally a low color temperature) to the room outside (generally a high color temperature) or vice versa, the color temperature of the light source is different and the white balance is greatly shifted. If the convergence speed is increased unconditionally at this time, the gain of the white balance amplifier changes greatly from screen to screen, which is unfavorable in appearance.
In this regard, Patent Document 3 describes that a stop area for determining a convergence condition for white balance control and a restart area for restarting are provided.

In white balance adjustment, an error occurs in control due to various factors.
For example, when the white balance is taken under a certain color temperature, the pull-in restriction frame moves in conjunction with this, and there is a disadvantage that the white balance cannot be taken (see Patent Document 2). In particular, when a transition from the low color temperature side to the high color temperature side is performed, a point inside the original drawing restriction frame may be out of the frame in the shifted drawing restriction frame, and white balance adjustment may not be performed.
In Patent Document 2, a reference point is provided in the pull-in limit frame, and the fluctuation of the pull-in limit frame itself is corrected by moving the reference point along the black body radiation curve.
JP 2004-320671 A JP-A-11-331854, paragraphs [0005] to [0007], etc. Japanese Patent Laid-Open No. 2003-324754

  In the techniques described in Patent Documents 1 to 3, when shooting under a low color temperature and shooting under a high color temperature are switched, control in consideration of such a change in shooting conditions (color temperature transition) is not. That is, a transition form such as a transition from the high color temperature side to the low color temperature side, conversely, a transition from the low color temperature side to the high color temperature side, or a transition within the high color temperature side or the low color temperature side. However, this is not considered in any of the above-mentioned patent documents.

In addition, the feed-forward white balance adjustment circuit controls the convergence speed when switching between shooting at low color temperature and shooting at high color temperature, so that the white balance amplifier gain changes by image. No technology has been proposed to prevent the feeling of discomfort from increasing.
In addition, if the technique described in Patent Document 3 is used, the convergence speed can be controlled by providing a dead zone, a stop region including the dead zone, and a restart region including the dead zone. It is a thing.

Furthermore, in the technique of Patent Document 2 that corrects the fluctuation of the pull-in limit frame along the natural light blackbody radiation curve, correct correction is performed on the high color temperature side (natural light side), but artificial light such as a light bulb (indoors) ) It is considered that correct correction may not always be made at the time of the lower shooting (low color temperature side).
As a result of insufficient correction on the low color temperature side, for example, when a transition is made from the low color temperature side to the high color temperature side, the point that should originally enter the pull-in restriction frame may be outside the frame. In this case, in the control using the pull-in limit frame, the pull-in operation is not always performed from the outside of the frame, and it cannot be said that the white balance adjustment functions sufficiently.
The fluctuation of the pull-in limit frame is considered to be mainly caused by a shift in the coordinate system during pre-white balance adjustment.

  The problem to be solved by the present invention is to appropriately adjust the white balance at the time of switching between the high color temperature side and the low color temperature side in the feedforward method.

  A white balance adjustment circuit according to a first aspect of the present invention includes a primary color separation circuit that sequentially separates image signals input in time series into primary color signals, an integration unit that integrates the primary color signals, and an obtained color The coordinate conversion means for converting the integral value for each color into a coordinate value on a predetermined color coordinate, and information on the brightness or color temperature of the image are input, and on the high color temperature side and low color temperature side based on the information A color temperature transition detecting means for detecting that the color temperature of the image has transitioned, and the coordinate value or the color temperature value of the light source estimated from the coordinate value is drawn into a predetermined target value, thereby white balance the image. Gain setting means that can set a gain value for adjustment for each primary color signal, and a variable gain amplifier that amplifies the primary color signal for each color with the set gain value, the gain setting means, Color in the frame The position, range, and effectiveness of the dead zone that is the prohibited range of degree adjustment can be set independently on the high color temperature side and low color temperature side on the color coordinates, and the transition is based on the detection result of the color temperature transition. Whether or not the dead zone is valid and invalid on the color temperature side, and if the dead zone on the color temperature side that has been transitioned is valid as a result of the decision, whether the coordinate value is drawn into the target value using the valid dead zone judge.

  A camera device according to a first aspect of the present invention includes a signal processing block including an image sensor, an operation unit, a memory that can be rewritten by an external instruction via the operation unit, and a white balance adjustment circuit. The white balance adjustment circuit is obtained by a primary color separation circuit that sequentially separates image signals generated by the image sensor and inputted in time series into primary color signals, and an integration unit that integrates the primary color signals. The coordinate conversion means for converting the integral value for each color into a coordinate value on a predetermined color coordinate, and information on the brightness or color temperature of the image are input, and the high color temperature side and the low color temperature side based on the information The color temperature transition detecting means for detecting the transition of the color temperature of the image and the coordinate value or the color temperature value of the light source estimated from the coordinate value to the predetermined target value. Gain setting means capable of setting a gain value for white balance adjustment for each primary color signal, and a variable gain amplifier for amplifying the primary color signal for each color with the set gain value, The setting means uses the frame parameter in the memory to determine the position, range, and effectiveness of the dead zone where the color temperature adjustment is prohibited within the frame on the high color temperature side and the low color temperature side on the color coordinate. It is possible to set independently, and on the side of the color temperature that has been transitioned based on the detection result of the color temperature transition, the dead zone valid and invalid are determined with reference to the dead zone enable flag in the memory, and as a result of the determination, When the transitional dead zone on the color temperature side is valid, it is determined whether the coordinate value is drawn into the target value using the valid dead zone.

  A white balance adjustment method according to a first aspect of the present invention includes a primary color separation step for sequentially separating image signals input in time series into primary color signals, integration of the primary color signals, and integration for each obtained color. A coordinate conversion step for converting the value into a coordinate value on a predetermined color coordinate and information on the brightness or color temperature of the image are input, and the color of the image on the high color temperature side and the low color temperature side based on the information In order to adjust the white balance of the image by detecting a color temperature transition detecting step for detecting that the temperature has changed, and drawing the color temperature value of the light source estimated from the coordinate value or the coordinate value into a predetermined target value. A gain setting step for setting the gain value for each primary color signal, and a signal amplification step for amplifying the primary color signal for each color with the set gain value. In the gain setting step, The position, range, and effectiveness of the dead zone where the color temperature adjustment is prohibited within the frame are set independently on the high color temperature side and low color temperature side on the color coordinates, and based on the detection result of the color temperature transition The dead zone is determined to be valid or invalid on the transitioned color temperature side, and if the dead band on the transitioned color temperature side is valid as a result of the determination, the coordinate value is drawn into the target value using the valid dead zone. After the determination, the gain value is set for each color by the pull-in.

  A white balance adjustment circuit according to a second aspect of the present invention includes a primary color separation circuit that sequentially separates image signals input in time series into primary color signals, an integration unit that integrates the primary color signals, and an obtained color. The coordinate conversion means for converting the integral value for each color into a coordinate value on a predetermined color coordinate, and information on the brightness or color temperature of the image are input, and on the high color temperature side and low color temperature side based on the information A color temperature transition detecting means for detecting that the color temperature of the image has transitioned, and the coordinate value or the color temperature value of the light source estimated from the coordinate value is drawn into a predetermined target value, thereby white balance the image. Gain setting means that can set a gain value for adjustment for each primary color signal, and a variable gain amplifier that amplifies the primary color signal for each color with the set gain value, the gain setting means, Color in the frame The pull-in limit frame as the degree adjustment range can be set on the high color temperature side and the low color temperature side on the color coordinates, and the detection result of the color temperature transition indicates that the pull-in operation is permitted outside the pull-in limit frame. If the transition is determined based on the color temperature side and the drawing operation outside the frame is permitted as a result of the determination, the pull-in operation is performed even if the coordinate value is located outside the transition limit frame on the color temperature side that has been transitioned. Execute.

  A camera device according to a second aspect of the present invention includes a signal processing block including an image sensor, an operation unit, a memory that can be rewritten by an external instruction via the operation unit, and a white balance adjustment circuit. The white balance adjustment circuit is obtained by a primary color separation circuit that sequentially separates image signals generated by the image sensor and inputted in time series into primary color signals, and an integration unit that integrates the primary color signals. The coordinate conversion means for converting the integral value for each color into a coordinate value on a predetermined color coordinate, and information on the brightness or color temperature of the image are input, and the high color temperature side and the low color temperature side based on the information The color temperature transition detecting means for detecting the transition of the color temperature of the image and the coordinate value or the color temperature value of the light source estimated from the coordinate value to the predetermined target value. Gain setting means capable of setting a gain value for white balance adjustment for each primary color signal, and a variable gain amplifier for amplifying the primary color signal for each color with the set gain value, and the gain setting The means can use the frame parameter in the memory to set a pull-in limit frame having a color temperature adjustment range in the frame on the high color temperature side and the low color temperature side on the color coordinates, and the pull-in limit The permission of the pull-in operation outside the frame is determined by referring to the out-of-frame pull-in permission flag in the memory on the color temperature side where the transition has been made based on the detection result of the color temperature transition, and as a result of the determination, the pull-in operation outside the frame is performed. When the operation is permitted, the pull-in is executed even when the coordinate value is located outside the pull-in limit frame on the color temperature side after the transition.

  A white balance adjustment method according to a second aspect of the present invention includes a primary color separation step for sequentially separating image signals input in time series into primary color signals, integration of the primary color signals, and integration for each obtained color. A coordinate conversion step for converting the value into a coordinate value on a predetermined color coordinate and information on the brightness or color temperature of the image are input, and the color of the image on the high color temperature side and the low color temperature side based on the information In order to adjust the white balance of the image by detecting a color temperature transition detecting step for detecting that the temperature has changed, and drawing the color temperature value of the light source estimated from the coordinate value or the coordinate value into a predetermined target value. A gain setting step for setting the gain value for each primary color signal, and a signal amplification step for amplifying the primary color signal for each color with the set gain value. In the gain setting step, A pull-in limit frame whose color temperature adjustment range is within the frame is set on the high color temperature side and low color temperature side on the color coordinates, and permission for pull-in operation outside the pull-in limit frame is permitted for the color temperature transition. Based on the detection result, it is determined on the side of the color temperature that has transitioned, and if the result of the determination is that the drawing operation outside the frame is permitted, even if the coordinate value is located outside the drawing-in restriction frame on the side of the color temperature that has transitioned By executing the pull-in, the gain value for each color is set.

  According to the present invention, in the feedforward method, it is possible to appropriately perform white balance adjustment according to the color temperature transition mode at the time of switching between the high color temperature side and the low color temperature side.

  Embodiments of a camera device including a white balance adjustment circuit and a white balance adjustment method according to the present invention will be described below with reference to the drawings.

<Configuration>
FIG. 1 shows a block configuration of the camera apparatus of the embodiment.
A camera device 10 illustrated in FIG. 1 includes a lens 11, an image sensor 12, a preamplifier 13, an A / D (analog / digital) converter 14, a primary color separator 15, a white balance amplifier 16, an optical detector (OPD) 17, and a controller. 18. Other γ (gamma) corrector, color difference matrix device, D / A (digital / analog) converter, etc. (not shown). The primary color separator 15, white balance amplifier 16, OPD 17, controller 18, other γ (gamma) corrector (not shown), color difference matrix unit, and the like are realized as a digital signal processing IC, for example. One or both of the A / D converter 14 and the D / A converter can be provided in the digital signal processing IC.

  The lens 11 projects an image of a subject (not shown) on the imaging surface of the imaging element 12. The image pickup device 12 includes, for example, a CCD (Charge Coupled Device), a CMOS (Complementary Metal Oxide Semiconductor) sensor, and the like, converts an image transmitted through the lens 11 into an electric signal, and supplies the electric signal to the preamplifier 13.

  The preamplifier 13 samples and holds the image signal from the image sensor 12 to extract necessary data, and performs gain control in order to adjust to an appropriate level. The output signal of the preamplifier 13 is output to the A / D converter 14.

  The A / D converter 14 converts the output signal from the preamplifier 13 from an analog signal to a digital signal, and outputs the digital signal to a digital signal processing IC connected to a subsequent stage.

The primary color separator 15 separates the digital signal from the A / D converter 14 into red, green, and blue primary color signals R, G, and B, respectively, and these primary color signals R, G, and B are respectively variable gain amplifiers 16r. , 16g, 16b.
The primary color signals R, G and B are used for white balance adjustment. For this purpose, these primary color signals R, G, B are also supplied to the OPD 17.

  The OPD 17 has a circuit that integrates the primary color signals R, G, and B for each field (or frame). For this reason, the OPD 17 obtains integrated value data Ir, Ig, Ib of, for example, field units of the primary color signals R, G, B, and outputs these integrated value data Ir, Ig, Ib to the controller 18.

  The controller 18 is hardware capable of computer-based operation under the control of a built-in or external central processing unit (CPU) such as a microcomputer or a DSP (Digital Signal Processor). The CPU executes processing such as coordinate conversion, setting and control of a color temperature pull-in limit frame, color temperature estimation, gain setting for each of the primary color signals R, G, and B in accordance with a program read from the built-in or external memory. Details of each process will be described later. Hereinafter, a case where these series of processes are executed according to software will be mainly described, and a modification example by hardware processes will be described as appropriate.

The white balance amplifier 16 has three variable gain amplifiers 16r, 16g, and 16b corresponding to the primary color signals R, G, and B.
The gains of the variable gain amplifiers 16r, 16g, and 16b can be changed as necessary based on the gain values Gr, Gg, and Gb from the controller 18. The variable gain amplifier 16r amplifies (or attenuates) the primary color signal R from the primary color separator 15 using the gain value Gr from the controller 18, and outputs it to the γ corrector as the primary color signal R ′. Similarly, the variable gain amplifier 16g amplifies (or attenuates) the primary color signal G using the gain value Gg from the controller 18, and the variable gain amplifier 16b amplifies the primary color signal B using the gain value Gb from the controller 18. (Or attenuate) and output the primary color signals G ′ and B ′ to the γ corrector.

A γ corrector (not shown) connected to the subsequent stage of the white balance amplifier 16 performs γ correction based on the primary color signals R ′, G ′, and B ′, and generates a video signal by being combined with a luminance (Y) signal (not shown). To do.
The signal output from the γ corrector passes through a color difference matrix unit (not shown), undergoes compression processing, etc., and is then recorded as a video signal on a recording medium such as a magnetic tape or a magneto-optical disk, or wired or wirelessly. Is transmitted.
Although not shown in FIG. 1, the camera device 10 can set flag information and the like used for control of the controller 18 by the user, and has an operation unit for that purpose. The flag information and the like are stored in the CPU and in a data rewritable memory (for example, a non-volatile memory) connected to the CPU. Specific contents of the flag information will be described later.

The white balance adjustment circuit shown in FIG. 1 is a feedforward circuit that sets the gain of the white balance amplifier 16 using the primary color signal from the primary color separator 15.
Compared with the feedback circuit that determines the amount of gain adjustment using the output of the white balance amplifier 16, the feedforward circuit has better processing responsiveness (convergence when obtaining the optimum gain value), and an optimum gain value can be obtained. The control system never oscillates.

Hereinafter, the various processes described above performed by the controller 18 will be described.
FIG. 2 is a diagram illustrating functions of a program executed by the controller 18. Note that these means conceptually include hardware such as a controller 18 and a memory for executing processing.
As functions of the execution program, there are coordinate conversion means 18A, color temperature transition detection means 18B, pull-in limit frame determination means 18C, dead zone frame determination means 18D, color temperature estimation means 18E, and pull-in means 18F. Of these, means other than the color temperature transition detection means 18B can be included in the “gain setting means” of the present invention. The operation of each of these means will be sequentially described below.

<Coordinate transformation>
This process is executed by the coordinate conversion means 18A in FIG.
As shown in FIG. 1, the controller 18 receives, for example, field-unit integrated value data Ir, Ig, Ib from the OPD 17 and converts them into coordinate values on predetermined color coordinates. The coordinate value on the color coordinate of the integral value data is an evaluation value representing the color tone when the color temperature of the field screen, that is, the achromatic white portion is in the screen.
This coordinate conversion is performed by software processing of the CPU, but since it is simple addition / subtraction, it may be performed by a logic circuit although not particularly shown. In the case of a logic circuit, the burden on the CPU can be reduced, and in the case of software processing, there is an advantage of suppressing an increase in circuit scale, and a preferable processing method can be selected from any viewpoint.

As the color coordinates, “R + B-2G” is taken on one axis and “R−B” is taken on the other axis, or this color coordinate is normalized by the green signal data value G, “(R + B -2G) / G "and" (RB) / G "color coordinates can be used. The case where the latter is used will be described below.
However, as will be described later, when calculating (R−B) / G value, (R + B−2G) / G value, (R−B) / G = R / GB−G, (R + B−2G) Since / G = R / G + B / G-2, the gain setting of the present embodiment using (R−B) / G value and (R + B−2G) / G value is R / G, B / This can be realized even in the G coordinate system, and in this case, the same effect is obtained.

Next, a specific method of coordinate transformation will be described. The method described here is merely an example, and other methods can be applied.
In this coordinate conversion, gains for the three primary colors at the time of pre-white balance adjustment, that is, R gain Gr (p), G gain Gg, and B gain Gb (p) are used. “P” in this reference symbol indicates a gain obtained by pre-white balance adjustment (hereinafter referred to as pre-white balance gain). Here, the reason why the G gain Gg is not obtained by the pre-white balance adjustment is that it is only necessary to adjust the other two colors with reference to green (G).
The value of the G gain Gg is stored, for example, in a storage area of the CPU, a ROM connected to the CPU, or the like.

An example of how to obtain the pre-white balance / gain is as follows.
For example, when the camera apparatus 10 is shipped and adjusted, a white subject is photographed under a light source having a predetermined reference color temperature, for example, a color temperature of 3200 [K]. At this time, integrated value data Ir, Ig, and Ib are obtained from the OPD 17 of FIG.
The integral value data Ir, Ig, and Ib used for the pre-white balance store the data ratio of the primary color signals R, G, and B of the image from which the integral value data is obtained. Therefore, the integral value data Ir, Ig, and Ib can be used to represent the color temperatures (color balance) of the primary color signals R, G, and B. Conversely, the integral value data Ir, Ig, and Ib can be replaced with primary color signal data values R, G, and B.
From the above, gains Gr (p), Gg (p), Gb (p) for each color at the time of pre-white balance using the following relational expressions (1-1), (1-2), (1-3) )

[Equation 1]
Ir × Gr (p) = Ig × Gg = Ib × Gb (p) (1-1), that is,
R × Gr (p) = G × Gg = B × Gb (p) (1-2),
R × Gr (p) −G × Gg = B × Gb (p) −G × Gg = 0 (1-3)

Expressions (1-2) and (1-3) indicate that the ratios of the primary color signals R ′, G ′, and B ′ output from the variable gain amplifiers 16r, 16g, and 16b constituting the white balance amplifier 16 of FIG. This is an expression representing the white balance adjustment operation of 1: 1: 1. From this equation, the ratio of the pre-white balance gain Gr (p) and Gb (p) is obtained. By multiplying this ratio by a predetermined proportional coefficient, each value of the pre-white balance gain Gr (p) and Gb (p) is obtained for each color and stored in a predetermined memory.
At this time, a parameter representing the black body radiation curve measured in advance is also stored in a predetermined memory.

  When pre-white balance adjustment is actually performed while using the camera device, the pre-white balance gain Gr (p), Gb (p), and G gain Gg obtained in advance are read from the memory, and the values and values obtained from the OPD 17 are obtained. Color signal level value after pre-white balance adjustment using the following equation (2) from the integrated value data Ir, Ig, Ib (→ replaced by R, G, B) (hereinafter referred to as amplifier output color level value) R (p), G (p), B (p) are obtained.

[Equation 2]
R (p) = R × Gr (p)
G (p) = G × Gg
B (p) = B × Gb (p) (2)

From this amplifier output color level value R (p), G (p), B (p), (R (p) −B (p)) / G (p) value (hereinafter simply referred to as “(R−B) / G value) and (R (p) + B (p) -2G (p)) / G (p) value (hereinafter simply referred to as “(R + B-2G) / G value”). As shown in FIG. 3 and the like, coordinate values on color coordinates (P0 etc.) are obtained with (R−B) / G on one axis and (R + B−2G) / G on the other axis.
The amplifier output color level value obtained by performing the pre-white balance adjustment as described above is obtained by adding the pre-white balance / level to the input primary color signal level value (actually the integrated value data), as is apparent from the above equation (2). The gain Gr (p), Gb (p) or G gain Gg is multiplied by each value. Therefore, the origin O of the color coordinates after the pre-white balance adjustment is 3200 [K] and represents the point where the three primary color signal levels are 1: 1: 1.

<Detection of color temperature transition>
The color temperature transition is a transition between two color coordinate values on the high color temperature side and the low color temperature side in the color coordinate values obtained one after another while processing a plurality of images. The detection of the transition is executed by the color temperature transition detection means 18B in FIG. At the time of detection, not a small transition but a color temperature transition larger than a predetermined standard, in other words, a transition that involves switching between a high color temperature side pull-in limit frame and a low color temperature side pull-in limit frame, which will be described later, is detected. To do. Alternatively, even if the color temperature change is small, a color temperature transition that involves switching between outdoor and outdoor may be detected in the vicinity of the indoor and outdoor thresholds.

FIG. 3 shows an example of transition from the low color temperature side to the high color temperature side on the (R−B) / G, (R + B−2G) / G color coordinates.
The image (frame or field) is switched from the state in which the color difference integrated value P0 exists in the low color temperature side pull-in limit frame LW, and as a result, the transition to the color difference integrated value P1 in the high color temperature side pull-in limit frame HW occurs. Suppose it occurred. For example, such a case is called color temperature transition.

In the example of FIG. 3, the change from the low color temperature side to the high color temperature side is represented by using an integral value (P0 → P1). As an example of detecting the change, luminance information Y is supplied from the OPD 17 of FIG. Then, it may be determined whether to use the low color temperature side pull-in limit frame LW or the high color temperature side pull-in limit frame HW accordingly. In this case, the OPD 17 has a function of inputting a luminance signal Y or extracting a luminance component of the screen from the primary color signals R, G, and B and performing brightness determination. However, the route for obtaining luminance (brightness) information is not limited. For example, when the camera apparatus has an illuminance sensor or the like, brightness determination may be performed from the sensor output.
In addition to the integral value data and luminance information for each color, color temperature information that can be regarded as having undergone a user setting or color temperature transition may be detected. For example, when the user performs a manual operation to perform a color temperature transition setting or an operation that considers a color temperature transition, this may be detected. The detection result of the color temperature transition is used for dead band control described later, permission for pulling out of the limit frame, and the like in addition to switching of the pull-in limit frame.

<Withdrawal limit frame judgment>
As a restriction frame at the time of color adjustment, there are a pull-in restriction frame whose inside is a color adjustment range, and a dead zone frame whose inside is a color adjustment prohibited range. As specific examples of the pull-in limit frame, FIG. 3 shows a low color temperature-side pull limit frame LW and a high color temperature-side pull limit frame HW.
This process is a process for determining whether or not the gain change is necessary by performing permission determination of the pull-in operation using the pull-in restriction frame. This process is executed by the pull-in limit frame determining unit 18C in FIG.
With regard to the pull-in limit frame determination, the relationship between the position of the integrated value on the color coordinate and the pull-in limit frame is checked, and the high color temperature side pull limit frame HW and the low color temperature side pull limit are determined according to the detection result of the color temperature transition. A process of setting or selecting (validating) one of the frames LW and a determination process of performing permission determination of the pull-in operation using the pull-in restriction frame set or selected are executed. Also in this determination process, different determinations can be made according to the detection result of the color temperature transition and the flag information value. Details thereof will be described later.

Specifically, when integral value data Ir, Ig, and Ib representing the color temperature of the current image are converted into color difference integral values on (R−B) / G, (R + B−2G) / G color coordinates. Whether the color coordinate position of the color difference integration value is inside or outside the drawing limit frame on the high color temperature side or the low color temperature side set or selected is checked.
Each of the high color temperature side pull-in limit frame HW on the high color temperature side and the low color temperature side pull-in limit frame LW on the low color temperature side is given by a position parameter on the color coordinate, and is stored in, for example, a ROM or a non-volatile memory. Stored in advance. When the parameters of these pull-in restriction frames can be changed, the parameters are stored in a nonvolatile memory or the like so that they can be rewritten.

<Dead zone determination>
This process is executed by the dead zone determination means 18D shown in FIG.

FIG. 4 shows an example of setting the dead band.
The dead zone frame is a restriction frame in which the inside of the frame indicates a prohibited range of color temperature adjustment. Therefore, the color difference integral value P outside the dead zone frame shown in FIG. 4 is a target of color temperature adjustment, that is, gain change, but not within the frame. For this reason, when the dead zone is entered during several times of control, the control is stopped, and when the dead zone is exited, the control is resumed. For this reason, the dead zone frame can also be referred to as a frame indicating a range in which white balance deviation can be allowed from a value obtained by adjusting the white balance once. If the dead zone is wide, it can be said that the area in the dead zone is the target value, so that the convergence of the control becomes high. Conversely, if it is narrow, the convergence will be low.
The dead zone is originally provided for preventing oscillation in the feedback control method, but in the present invention, this is used for convergence control. Therefore, it is not necessary to provide a relatively small four-way equidistant around the target value (in this case, the origin) as in the case of oscillation prevention. The size necessary for this convergence control and the relative position to the origin are It is decided.

  The dead zone frame determination means 18D determines the control of this process with reference to a dead zone frame validation flag F2 (details will be described later) held in advance in the CPU or in an external memory. As a control method, it may be determined whether or not to provide a dead zone frame by looking at the value of the dead zone frame enable flag F2, or the provided dead zone frame may not be targeted (not activated) at the time of determination.

The dead zone frame determination unit 18D has a function of setting the dead zone frame on the color coordinates. Specifically, as shown in FIG. 4, dead zone frames having predetermined widths in the (R−B) / G axis and (R + B−2G) / G axis directions are provided. The dead zone frame is provided as a parameter for four directions from the center point of the (R−B) / G axis and (R + B−2G) / G axis, and is stored in the CPU or in an external memory in a changeable manner. The dead zone frame can be switched between setting and non-setting, valid and invalid independently on the low color temperature side and the high color temperature side.
The dead band frame can be set by the dead band frame determination unit 18D or the user, and can be changed by the user. The user can set and change the dead zone frame by, for example, an operation via a menu screen.
Details of the dead zone determination will be described later.

<Color temperature estimation>
This process is executed by the color temperature estimating means 18E shown in FIG.
Here, the color temperature estimation is to estimate the current color temperature of the image (field or frame) indicated by the primary color signal input from the primary color separator 15 of FIG. More specifically, when the color temperature is adjusted to the target value (pull-in operation), if the color changes, the starting point of the pull-in operation is determined from the current coordinate point to the black body in order to prevent or suppress the color change. An operation of changing to a predetermined position such as a radiation curve. The color temperature on the black body radiation curve obtained from the current coordinate point is hereinafter referred to as an estimated color temperature.

Next, a color temperature estimation method will be described with reference to FIG.
FIG. 5 is an explanatory diagram of a color temperature estimation method on the high color temperature side.
In this example, the black body radiation curve is approximated by a straight line (hereinafter referred to as a color temperature change straight line LB), and the color temperature is estimated on the high color temperature side.
In FIG. 5, when the X-axis is (R + B-2G) / G and the Y-axis is (RB) / G, the area “A” on the coordinates is the area within the pull-in restriction frame W1 in the third quadrant. And a region on the left side of the color temperature change straight line LB in the frame pull-in restriction frame W1 in the fourth quadrant.
The data on the coordinates obtained by calculating (Ir + Ib-2Ig) / Ig and (Ir-Ib) / Ig based on the integrated value data Ir, Ig, Ib are the circles in the area “A” described above. , The point that intersects the color temperature change line LB by translating these data along the X axis representing (R + B−2G) / G as indicated by the arrow, Estimated temperature.
For example, when a pull-in operation is performed from the estimated color temperature to the coordinate origin O, the color change of the data P1 moves in parallel with the pull-in operation from the data P1 to the origin O of the estimated color temperature on the coordinate. This is equivalent to a color change up to a point P1 ′ where (R + B−2G) / G value is smaller than O. Therefore, the color temperature change in which the G component is stored is shown.

On the other hand, the area “B” on the coordinates is an area on the right side of the color temperature change line LB in the frame pull-in restriction frame W1 in the fourth quadrant.
When the integration value conversion data converted into the coordinate data by converting the integration value data Ir, Ig, Ib is at the position indicated by the triangle in the region “B”, as indicated by the arrow (R + B− 2G) / G representing the X axis and (R−B) / G representing the Y axis are moved at the same ratio in the negative direction, and the estimated color temperature is obtained from the intersection with the color temperature change line LB.
Specifically, the estimated color temperature is a point where the extension line and the color temperature change straight line LB intersect with each other by moving 45 ° to the lower left on the coordinates from the position of Δ in the region “B”.
In this case, the color change of the data P2 moves on the coordinate axis from the data P2 in parallel with the drawing operation from the estimated color temperature to the origin O, and the (R + B-2G) / G value and (R− B) Both are equivalent to a color change up to a point P2 ′ where both G values are large. For this reason, the R component is preserved (or the B component is reduced), and as a result, the color change of the skin color is particularly prevented.

When the color temperature is estimated in this way, a predetermined color component is stored according to the method, and the color temperature can be corrected without affecting the color reproducibility of the screen.
On the low color temperature side, it is possible to estimate the color temperature on the black body radiation curve or at a predetermined position on the basis of the black body radiation curve by a method suitable for storing a predetermined color component (for example, yellow). It is. The color temperature estimation method is a method of estimation using a black body radiation characteristic approximated by a straight line, but should not be limited to this.

<Pull-in (gain setting)>
This process is executed by the pull-in means 18F in FIG. 2 in order to achieve white balance.
Here, the white balance is taken (adjusted), as described in the explanation of the pre-white balance, so that the ratios of the R, G, B color components of the estimated color temperature are equal. This means adjusting the gain value of the white balance amplifier.

  In calculating the gain value of the white balance adjustment in the controller 18 shown in FIG. 1, the following equation (3-1) generalizing equations (1-2) and (1-3) used for the pre-white balance adjustment: And (3-2) are used.

[Equation 3]
R × Gr (i) = G × Gg (i) = B × G (i) (3-1),
R × Gr (i) −G × Gg (i) = B × G (i) −G × Gg (i) = 0 (3-2)

Here, the symbol “i” is an arbitrary natural number indicating the number of white balance adjustments performed after the pre-white balance adjustment. When the color temperature is estimated, reference symbols R, G, and B represent color signal data values representing the color components of the estimated temperature.
The general formula (4) when the white balance is adjusted i times with respect to the primary color signal of the same image (field or frame) can be expressed as follows. it can. As such a case, there is a step pull-in operation described later.

[Equation 4]
R (i) = R × Gr (i)
G (i) = G × Gg (i)
B (i) = B × Gb (i) (4)

In step pull-in operation, multiple white balance adjustments are made for the same image (field or frame), so there is one integral value data representing the primary color signal data for that screen for each color. In (2), the same color signal data values R, G, B are used. The color signal data corresponds to the color signal data of the estimated color temperature in a 1: 1 ratio, and the color signal data R, G, and B of Equation (4) represent the color signal data after the color temperature estimation. You can consider it.
The pull-in operation is to set the ratio of the left sides of the three formulas (4) to 1: 1: 1. Therefore, if the color signal data R, G, and B of the estimated color temperature are known, the formula ( Using 4), the gain values Gr (i), Gg (i), and Gb (i) for each color can be obtained.
When the gain values Gr (i), Gg (i), and Gb (i) of the respective colors are set in the variable gain amplifiers 16r, 16g, and 16b in FIG. 1, the primary color signals R, G, and B that are input thereafter are white. Balance adjustment is performed. In the white balance adjustment accompanied by the color temperature estimation, the color temperature is estimated along the black body radiation curve on the coordinates (for example, 3200 [K] as a reference) for estimating the color temperature, and the current color is estimated with respect to the estimated color temperature. The origin on the coordinates multiplied by the gain value is achromatic white. This is clear from the above formula (4).

The above is the basic function of the controller 18.
In addition, the controller 18 includes, as auxiliary functions, an operation outside the pull-in limit frame, a function for switching the dead band frame between valid and invalid, a function for giving a dead time frame determination process a time delay (time constant), and a delay (process) for the pull-in operation. A step pull-in operation function for providing a delay), a target value type and value switching function, and a user preset setting function thereof. The user preset can be arbitrarily set according to the user's preference by an operation using a menu screen, an external dedicated button, or the like. When not set, a predetermined default value is used. Or, when there is a color temperature transition, it is automatically switched.
Each content will be described together in the following operation description.

As an explanation of the operation, first, a more detailed explanation will be given of the pull-in limit frame determination and the dead-zone determination while assuming the basic operations of the color temperature transition detection, color temperature estimation, and pull-in (gain setting) described above.
First, as an outline, when there is a color temperature transition between the high color temperature side and the low color temperature side, whether to change the gain according to the values of the out-of-frame pull-in permission flag F1 and the dead zone frame enable flag F2 This determination will be described for each case. At this time, the (R−B) / G value and the (R + B−2G) / G value are calculated from the amplifier output color level values R (p), G (p), and B (p) in the equation (2). Then, it is determined whether or not the coordinate value of the color difference integrated value after the color temperature transition is out of the pull-in limit frame, and if it is out of the frame, the gain is changed according to the value of the out-of-frame pull-in permission flag F1 that determines whether to perform the pull-in operation. Judge whether to perform.

FIG. 6 is a chart showing a mode of determination according to the flag value.
“Case A1 and A2” is the case where the integrated color difference P1 after the color temperature transition is within the pull-in restriction frame of the high color temperature or the low color temperature of the transition destination, and “Case B1 to B4” is outside the frame that did not enter. This is the case (see FIG. 7).
In “cases A1 and A2”, the out-of-frame pull-in permission flag F1 is not referred to, so the value is arbitrary. In “cases B1 to B4”, the out-of-frame pull-in permission flag F1 may be on or off.
“Case A1” and “Case A2” are when the dead zone frame enable flag F2 is on, and “Case A1” is control within the dead zone frame (in the case of the color difference integral value P2 shown in FIG. 8), “Case”. “A2” indicates control outside the dead zone (in the case of the color difference integral value P3 shown in FIG. 8).
“Case B1” and “Case B2” are cases where both the out-of-frame pull-in permission flag F1 and the dead zone frame enabling flag F2 are on, and “Case B1” is the control within the dead zone frame (the color difference integral value shown in FIG. 10). “Case B2” indicates control outside the dead zone frame (in the case of the color difference integral value P6 shown in FIG. 10).
“Case B3” is when the out-of-frame drawing permission flag F1 is on, but the dead-zone frame enabling flag F2 is off (in the case of the color difference integral value P7 shown in FIG. 11). In addition, in “Case B4”, since the gain is not changed when the out-of-frame pull-in permission flag F1 is off, the value of the dead zone frame enabling flag F2 is arbitrary.

<In case of restriction: Cases A1 and A2>
In the case of being in the dead zone frame as in the color difference integrated value P2 shown in FIG. 8, it is not necessary to change the gain without pulling in, or in the case of being out of the frame as in the case of the color difference integrated value P3 to change the gain by pulling in. Conforms to the purpose of
The feature of this embodiment is that the dead zone frame can be turned on and off independently on the high color temperature side and on the low color temperature side.

<When control is not performed outside the limit frame: Cases B1 and B4>
In case B1, the out-of-frame pull-in permission flag F1 is turned on, and for example, it is possible to pull in to the target value provided on the pull-in limit frame (in the example of FIG. 10, the high color temperature side convergence target value HT). The color difference integral value P5 of 10 exists in the dead zone frame provided around the target value HT on the high color temperature side pull-in limit frame HW. For this reason, priority is given to the prohibition of pull-in by the dead band, and in this case, the gain is not changed.
In case B4, since the out-of-frame pull-in permission flag F1 is off, the gain is not changed regardless of the dead zone frame.
Accordingly, in these cases, the gain values of the variable gain amplifiers 16r, 16g, and 16b are maintained as they are, and the white balance is not allowed to follow the color temperature transition, as in the case of outside the normal white balance control pull-in limit frame.

<When control is performed outside the limit frame: Cases B2 and B3>
In cases B2 and B3, the dead zone frame enable flag F2 is turned off (case B3: see FIG. 11), or the coordinate point (color difference integral value P7) representing the current color temperature outside the dead zone frame even if it is turned on. ) Is positioned (Case B2: see color difference integral value P6 in FIG. 10).
However, the case B2 has a delay in control, and the case B3 does not. In this case, when the dead zone enabling flag F2 is on, the control has a delay, and when it is off, there is no delay. Details of the method of providing a delay will be described later. As a result, when the dead zone is turned on, control is stopped when it enters the dead zone and the control stops, so the number of control steps until the white balance adjustment converges is reduced. The non-uniformity is corrected and the white balance adjustment time is always stable and easy to view. However, in cases B2 and B3, the convergence point (high color temperature side convergence target value HT in FIGS. 10 and 11) for white balance when outside the pull-in limit frame is the coordinates (white balance amplifier 16 (FIG. 1) in FIG. 10 and FIG. 11). It is not taken as the center point of the previous coordinate system), but, for example, on the frame in the coordinate system after the white balance amplifier 16 (FIG. 1). Details of this point will be described later.

  Next, the individual operations for the above cases will be described in more detail with reference to FIGS.

<When using a dead band (dead band activation flag F2: ON)>
The dead zone determination means 18D in FIG. 2 calculates (R−B) / G value, (R + B−2G) / G value (coordinates of the color difference integral values P5 and P6 in FIG. 10) calculated based on the equation (4). It is discriminated whether or not is outside the dead zone frame or inside.
Regarding the color difference integral value P6 in which the (R−B) / G value and (R + B−2G) / G value calculated based on the equation (4) are outside the dead zone frame, the white balance adjustment value is shifted. Determine and perform the pull-in operation.
At this time, it is preferable to provide a time constant until the start of pull-in. For example, as shown in FIG. 10, the time constant is "a period of several fields, a time to wait for being outside the dead zone", and this delay field number is provided as a parameter. Alternatively, it can be set and changed by the user in the external memory. If the (R−B) / G value and (R + B−2G) / G value calculated based on the equation (4) move within the dead zone within the time constant, the pull-in operation is not performed.
On the other hand, for the color difference integral value P5 in which the (R−B) / G value and (R + B−2G) / G value calculated based on the equation (4) are inside the dead zone frame, the deviation of the white balance adjustment value is different. Judge that it is within the allowable range and do not perform the pull-in operation.

<When no dead band is used (dead band activation flag F2: OFF)>
In the case of FIG. 11, it is determined that the white balance adjustment value is shifted without providing a dead zone or time constant, and a pull-in operation is performed.
As described above, the target value of the pull-in operation is set to the coordinate origin when preset white balance adjustment is performed. On the other hand, (R−B) / G value, (R + B−2G) / G value (coordinate point of color difference integral value P7 in FIG. 11) after the color temperature transition is outside the pull-in limit frame (in the example of FIG. 11, high color temperature). In the case of the side pull-in limit frame HW), the coordinate origin O in FIGS. 8 and 9 is not set as a target value, but is specified by a parameter stored in the CPU or the like or a user-specified target value (FIG. 10, FIG. 11 high color temperature side convergence target value HT, etc.). Then, the control value is drawn while stepping toward the set target value.

FIG. 12 shows a setting example of the high color temperature side convergence target value HT and the low color temperature side convergence target value LT. The coordinate systems in FIG. 12 and FIG. 11 are those before adjustment by the white balance amplifier 16 (FIG. 1), and are hereinafter referred to as “coordinate system before the WB amplifier”. However, in the coordinate system before the WB amplifier, the origin is the color temperature after the pre-white balance adjustment, and the coordinate value represented here is obtained by multiplying the integral value by the gain value of the pre-white balance adjustment. Is. In FIG. 12, the direction of change of each color in natural light is indicated by arrows.
The high color temperature side convergence target value HT is set, for example, at the corner where blue B of the high color temperature side pull-in limit frame HW is the strongest. The low color temperature side convergence target value LT is desirably set to a point where the low color temperature side pull-in limit frame LW intersects with the black body radiation curve, for example.
However, the setting of the convergence target values HT and LT is not limited to the illustrated example, and is arbitrary. However, on the high color temperature side, the black body radiation curve is desirable as described above, and on the low color temperature side, it is desirable to determine the position of the convergence target value according to the indoor light source.

Here, the control value at the time of engraving the step includes a control parameter representing the starting point for each step of the pulling operation, the end point (target) of the pulling operation, the number of repetitions, and the like. The current color difference integral value is set to the control parameter representing the first starting point, and the provisional target value provided between the control parameter representing the starting point and the control parameter representing the convergence target value (final target value) is set. Let them pull in. Then, the temporary target value becomes the next control value, and the new temporary target value is drawn. In this way, in the step pull-in operation, the pull-in operation is repeated while changing the control parameter so that the temporary target value is brought closer to the convergence target value sequentially without pulling in the final convergence target value at once. As a result, the pull-in operation can be provided with a processing delay, and a sudden color change can be prevented. If it is not necessary to prevent abrupt color change, a normal pull-in operation without a step pull-in operation can be performed. For example, as shown in FIG. 6 described above, it can be determined whether or not the step pull-in operation is performed in accordance with whether or not the dead zone frame is provided.
Control parameters such as convergence target values may be given as gain values of the variable gain amplifiers 16r, 16g, and 16b, or may be given as (R−B) / G values and (R + B−2G) / G values. Hereinafter, these two control methods will be described.

First, the case where the convergence target value is given by the gain values of the variable gain amplifiers 16r, 16g, and 16b will be described.
As shown in FIG. 13, the gain value of the R amplifier (variable gain amplifier 16r) and the gain value of the B amplifier (variable gain amplifier 16b) are provided as control parameters. These control parameters can be set by the user independently on the low color temperature side and the high color temperature side.
The current gain value of the white balance amplifier 16 (FIG. 1), which is a control value, is drawn in step by step toward the convergence target gain value GT.

On the other hand, as shown in FIG. 12, the high color temperature side convergence target value HT and the low color temperature side convergence target value LT given by (R−B) / G value and (R + B−2G) / G value are represented. Control parameters can also be provided. This control parameter can also be set by the user independently on the low color temperature side and the high color temperature side.
This parameter is a value obtained by multiplying the pre-white balance adjustment value calculated by the equation (2), and has a meaning as a color temperature, not as a coordinate value.

The drawing operation to the convergence target value HT in FIG. 12 will be described.
When calculation is performed using Expression (2) and Expression (4), (R−B) / G value (R + B−2G) calculated based on the color signal level values R, G, and B of Expression (2). ) / G value indicates the value given by the control parameter. Specifically, when the color signal level values R, G, and B obtained from the equation (2) are substituted into the equation (4), the following equation (5) is established.

[Equation 5]
R (i) = (R (p) / Gr (p)) × Gr (i)
G (i) = (G (p) / Gg (p)) × Gg (i)
B (i) = (B (p) / Gb (p)) × Gb (i) (5)

  Equation (3) is transformed into an equation applicable when performing step pull-in operation on the same image. Specifically, when the color signal level values R, G, and B obtained from the equation (2) are substituted into the equation (3), the following equation (6-1) is obtained. Further, the above formula (5) to formula (6-1) can be transformed into the following formula (6-2).

[Equation 6]
(R (p) / Gr (p)) × Gr (i) = (G (p) / Gg (p)) × Gg (i)
= (B (p) / Gb (p)) × G (i) (6-1)
R (i) -G (i) = B (i) -G (i) = 0 (6-2)

  In the step pull-in operation, it is possible to operate the white balance amplifier 16 (FIG. 1) so that the above equation (6) is satisfied, thereby pulling in the color temperature of the designated control parameter.

  In this way, when the control parameter is given by (R−B) / G value and (R + B−2G) / G value, it is possible to give meaning as a color temperature instead of a coordinate value, and thus has the following advantages.

Normally, the value of the user custom preset (white balance mode such as fluorescent lamp mode, indoor mode, outdoor mode) designates the amplifier value for white balance adjustment. In this case, there is a disadvantage that the pre-white balance adjustment value changes in each camera device. The reason is that the white balance adjustment is a white adjustment for absorbing the solid variation of the image sensor 12 (FIG. 1), and therefore, the pre-white balance adjustment value (gain value) is not the same in all camera devices. Because there is almost no.
For the same reason, even if the gain value of the same white balance amplifier is specified, the color temperature to be drawn in each camera device has changed.
In order to adjust the color temperature to be drawn in, adjustment may be performed in a factory, which causes an increase in adjustment cost.

FIG. 14 and FIG. 15 show the difference between the case where the user custom preset value is designated by the amplifier value of the variable gain amplifier (FIG. 14) and the case where the user custom preset value is designated by the coordinate value (FIG. 15).
As described above, the control target value in the case where the control outside the pull-in limit frame is performed when the photographing under the low color temperature and the photographing under the high color temperature are switched is (RB) / G value, (R + B-2G). The method of calculating the value (gain value) of the white balance amplifier 16 (FIG. 1) with the coordinate value of / G value is a user custom preset (white balance mode such as fluorescent lamp mode, indoor mode, outdoor mode) Application to is possible.
That is, in the user custom preset mode, the gain value corresponding to each mode is held in the memory so that it can be changed. In this embodiment, the pre-white balance gain Gr (p), Gb (p), and G gain Gg ( 3200 [K]) is adjusted, the gain value corresponding to each mode is no longer the correct color temperature designation. That is, when each mode is used, the color temperature to be drawn changes as shown in FIG.
On the other hand, as shown in FIG. 15, instead of gain values corresponding to each mode of the user preset mode, if the coordinate value is stored in the memory, each mode is used in conjunction with the pre-white balance / gain adjustment. Sometimes the temperature drawn in also changes correctly.
From the above, if the user custom preset value is specified as a coordinate value, the adjustment work performed in the factory is not required to match the color temperature to be pulled in, which greatly contributes to shortening the work time at the time of factory shipment. Is obtained.

Up to this point, when switching between the low color temperature side and the high color temperature side and outside the pull-in limit frame, the convergence target value of the pull-in is determined at a predetermined position on the pull-in limit frame. This is the case, for example, when an image that is in the low color temperature side pull-in limit frame LW and has been pulled in and white balance adjusted on the low color temperature side suddenly shifts to the high color temperature side and goes out of the frame This is because it is difficult to determine whether the image should originally be within the frame. That is, in the past, the shift of the pull-in restriction frame was corrected based on the reference point, but in the present embodiment, this is not done and the origin (color balance is 1) when the frame goes out of the frame after transition. The full-scale pull-in operation to the point of 1: 1 is not performed, but the pull-in operation is started toward the convergence target value on the pull-in limit frame so that the full-scale pull-in operation is easily performed for the time being. In the meantime, while the subject changes and the white balance is taken several times, the integral value moves into the limit frame if it should be within the frame. In this case, full-scale pull-in operation is immediately started by normal auto white balance control. On the other hand, if it is originally out of the frame and the drawing operation is unnecessary, the convergence does not proceed even if the drawing operation is performed on the target value on the frame. Therefore, full-scale pull-in operation is not performed.
Such a method has an advantage that appropriate white balance adjustment can be easily performed without performing complicated calculations. Also, if there is a color temperature transition on the low color temperature side and the high color temperature side, it is determined whether the white balance is to be followed according to whether it is outside the frame of the pull-in restriction frame or within the frame. It can be appropriately provided by setting flag information. As a result, the white balance control function is improved, and the white balance performance can be appropriately selected according to the user to be used and the system being constructed.

Next, a white balance control method at the time of switching between shooting at a low color temperature and shooting at a high color temperature will be described with reference to flowcharts shown in FIGS.
FIG. 16 shows the entire processing flow, and FIG. 17 shows the processing flow of normal automatic white balance control (AWB).

The process of FIG. 16 is started automatically during shooting or by manual setting according to the key operation of the operation unit by the user.
Step ST1 is an input step of the integral value data output from OPD 17 (FIG. 1).
The OPD 17 receives the R signal, G signal, and B signal output from the primary color separator 15 and integrates them for each field (or one frame) to obtain integrated value data Ir, Ig, and Ib. The obtained integral value data is supplied to the controller 18.

Step ST2 is executed by the coordinate conversion means 18A (FIG. 2) in the controller 18.
The coordinate conversion means 18A converts the integral value data Ir, Ig, Ib input to the controller 18 into coordinate values on (RB) / G, (R + B-2G) / G color coordinates. In this conversion, the pre-white balance / gain stored in the memory in advance is reflected in the coordinate conversion result. Also, the pre-white balance / gain is stored in the memory for each custom preset mode, and the coordinate conversion suitable for the shooting environment can be executed by reading the pre-white balance / gain suitable for the mode according to the user operation. .
As specific processing, as shown in the equation (2), the color signal level values R, G, B (integrated value data Ir, Ig, Ib) when pre-white balance is not performed are adjusted during pre-white balance adjustment. Are multiplied by the color gains Gr (p), Gg, and Gb (p), and a color signal level value R (based on a predetermined color temperature (for example, 3200 [K]) is obtained by a coordinate transformation calculation based on the multiplication result. The values (integrated values) on (R−B) / G, (R + B−2G) / G color coordinates corresponding to p), G (p), and B (p) are obtained.

Step ST3 is executed by the color temperature transition detecting means 18B (FIG. 2) in the controller 18.
The color temperature transition detection means 18B monitors and detects whether the color temperature transition on the high color temperature side and the low color temperature side has occurred since the previous white balance adjustment. For this detection, as described in the item “Detection of color temperature transition”, for example, brightness information such as a luminance signal Y may be used.
If there is a color temperature transition, the process flow proceeds to step ST4, and if not, the process proceeds to step ST11.

First, a case where there is no color temperature transition will be described.
In step ST11, the pull-in limit frame determination unit 18C in FIG. 2 checks whether or not the previous pull-out limit frame control has been executed based on the determination history. Goes to step ST20, and normal AWB control is executed. If the out-of-frame control has been executed last time, in the next step ST12, the attraction-in limit frame determining means 18C checks whether it is currently within the attraction-in limit frame. Pull in. Here, the full-scale pull-in operation is a pull-in operation with the coordinate origin as the convergence target.
On the other hand, if it is outside the pull-in limit frame, the processing flow moves to before step ST5.
The reason for performing such processing is that even if there is no color temperature transition this time, the color temperature transition is performed before that, and the pull-in operation may be performed to the convergence target value on the frame described later. It is a thing. That is, in such a case, when the integral value exists within the frame, it is necessary to start a full-scale pull-in operation immediately without performing coordinate conversion again. On the other hand, if steps ST11 to ST13 are not provided, coordinate conversion is performed in the next step ST20 (see FIG. 17), which is not preferable. The reason for providing steps ST11 to ST13 is to avoid this undesirable situation.

Next, normal AWB control in FIG. 17 will be described.
As in steps ST1 and ST2 in FIG. 16, after the integral value input (ST21) and coordinate transformation (ST22) are performed, it is checked in step ST23 whether the obtained integral value is within the pull-in limit frame.
If it is outside the pull-in limit frame, the process proceeds to return (ST0) without performing processing, and enters a state of waiting for the next integral value input. When the next integration value is input, the processing flow of FIG. 16 is started again.
On the other hand, if it is within the pull-in limit frame, color temperature estimation is executed by the color temperature estimation means 18E in FIG. 2 in the next step ST24. At this time, the color temperature is estimated by a method determined depending on whether the current integrated value is on the high color temperature side or the low color temperature side (see the above-mentioned “Color Temperature Estimation” section).

  In step ST25, (R−B) / G value and (R + B−2G) / G value are calculated for the coordinate point of the estimated color temperature. This calculated coordinate value corresponds to an amplifier input value (color signal data values R, G, B) at the time of white balance calculation described later (see the above equation (4), etc.).

In step ST26, the dead zone determination means 18D in FIG. 2 has the estimated color temperature coordinate values ((R−B) / G, (R + B−2G) / G) in step ST25 outside the dead zone. Or whether it exists inside. As described above, the dead zone frame is provided as a parameter with respect to four directions from the center point of the (R−B) / G axis and (R + B−2G) / G axis, and is independent on the low color temperature side or the high color temperature side. The default value or set by the user can be changed (see FIG. 4).
When coordinate points represented by (R−B) / G value and (R + B−2G) / G value exist inside the dead zone, it is determined that there is no need to adjust the white balance, and the return (RET) : Step ST0) is performed. By the return, in step ST1 of FIG. 16, the next integrated value data is waited for input.
On the other hand, when the coordinate point represented by (R−B) / G value or (R + B−2G) / G value exists outside the dead zone frame, it is determined that the white balance adjustment value is shifted, and the pull-in is performed. In order to perform the operation, the processing flow moves to step ST27.

In step ST27, white balance gain adjustment, that is, calculation (setting) of the gain value obtained by the pull-in operation, and signal amplification by the variable gain amplifiers 16r, 16g, and 16b using the gain value are executed.
Specifically, the gain value when the drawing means 18F of FIG. 2 draws the estimated color temperature to the coordinate origin, that is, the amplifier output color level values R (i), G (i), B (i) is 1: 1. The gain values Gr (i), Gg (i), and Gb (i) of the variable gain amplifiers 16r, 16g, and 16b are set to 1 by calculation (white balance calculation). The relationship between the amplifier output color level value and the gain value is given by the above equations (3-1) to (6-2). Here, i = 1 because it is the first white balance calculation.
The calculated gain values Gr (i), Gg (i), Gb (i) are actually supplied to the variable gain amplifiers 16r, 16g, 16b in FIG. Is amplified or attenuated by the gain values Gr (i), Gg (i), and Gb (i).
Thereafter, the processing flow shifts to return (RET: step ST0) and waits for input of the next integral value data.

The pull-in operation may be performed once on a predetermined convergence target value, for example, the origin O of the color coordinates, but preferably the following method is adopted in order to avoid a rapid change in the color temperature.
In the first method, for example, the number of times the OPD 17 sequentially outputs the integral value data, a time delay defined by other methods (for example, detection of the vertical synchronization signal), etc. are provided, and the pull-in operation is performed only once every several times. This is a method that is not performed. This is the case with the so-called time constant described above.
The second method is a case where the above-described step pull-in operation is performed. In this case, in step ST27, the coordinate origin is not set as the convergence target value, and the gain value when the halfway temporary target value is pulled in is obtained. . Although not particularly shown, a step of incrementing the white balance calculation number i and a step of determining whether this has reached a predetermined number are provided after step ST27. If the predetermined number of times has not been reached, the process flow is returned to step ST23, and the determination result is “from the step ST23 to the predetermined number of times determination step while the temporary target value is brought closer to the convergence target value (coordinate origin) sequentially. Repeat until “Yes”. If the determination result is “Yes”, or if it is determined in step ST26 that it is in the dead zone, the processing flow moves to return (RET: step ST0) and is waiting for input of the next integral value data. It becomes.

Returning to the processing flow of FIG.
If it is determined in step ST3 that there is a color temperature transition between the high color temperature side and the low color temperature side, in the next step ST4, the pull-in limit frame determination unit 18C in FIG. Executed.
At this time, the pull-in limit frame on the side after the color temperature transition is used as a determination criterion. The selection of the pull-in restriction frame uses the detection result when the detection result of the color temperature transition includes information on the direction of the transition. Alternatively, for example, the luminance signal Y may be input and it may be determined whether to use the high color temperature side pull-in limit frame HW or the low color temperature side pull-in limit frame LW accordingly.
When either the high color temperature side pull-in limit frame HW or the low color temperature side pull-in limit frame LW is selected, the coordinate values ((R−B) / G value, (R + B−) obtained by the coordinate conversion in step ST2 are selected. 2G) / G value) is determined to be present inside or outside the low color temperature pull-in limit frame LW or the high color temperature pull-in limit frame HW shown in the coordinates of FIG. 3, for example.
If the coordinate value is within the selected pull-in limit frame, the determination in step ST4 is “Yes”, and the processing flow shifts to normal AWB control in step ST20.
If the coordinate value is outside the selected pull-in limit frame, the process flow proceeds to step ST5, and a series of operations for white balance adjustment is continued.

In step ST5, it is checked by the pull-in limit frame determination means 18C in FIG. Thereby, the user setting of whether to perform out-of-frame control or not is confirmed.
When the out-of-frame pull-in permission flag F1 is off, the processing flow forcibly shifts to return (RET: step ST0), and enters a state of waiting for input of the next integral value data. On the other hand, if the out-of-frame pull-in permission flag F1 is on, the processing flow moves to the next step ST6.

  In step ST6, the pull-in means 18F in FIG. 2 sets a convergence target value in the case of out-of-frame control. For example, in the example shown in FIG. 12, the high color temperature side convergence target value HT or the low color temperature side convergence target value LT corresponds to this convergence target value. The high color temperature side convergence target value HT or the low color temperature side convergence target value LT is given as a parameter as a user setting or default value, and is stored in a predetermined memory or the like.

  In step ST7, the coordinate values ((R−B) / G value, (R + B−2G) / G value) are calculated as in step ST25 of FIG. However, in this case, since the color temperature is not estimated, the data that is the basis of the calculation is obtained in step ST2, and is given by (R−B) / G value and (R + B−2G) / G value. Integrated value data. That is, the coordinate value obtained in step ST2 may be used as it is.

In step ST8, the same processing as step ST26 of FIG. 17 is performed. That is, the pull-in means 18F in FIG. 2 searches the predetermined memory to check whether the dead zone activation flag F2 is on. Thereby, it is determined whether or not the dead zone frame is used.
When the dead zone enable flag F2 is off, the processing flow is performed to adjust the white balance amplifier using only the low color temperature side pull-in limit frame LW or the high color temperature side pull-in limit frame HW without using the dead zone frame. If the process transitions to ST10 and the dead zone validation flag F2 is on, the process flow proceeds to the next step ST9.
Although the dead zone frame itself is provided as the four-direction parameter as described above, the dead zone frame is provided around the high color temperature side convergence target value HT or the low color temperature side convergence target value LT of FIG. (See FIG. 10). This dead zone can also be set as a default value independently on the low color temperature side or the high color temperature side or can be changed by the user (see FIG. 4).

In step ST9, the dead zone determination means 18D in FIG. 2 determines whether the coordinate values ((R−B) / G value, (R + B−2G) / G value) in step ST7 exist outside the dead zone. To determine if it exists.
When (R−B) / G value and (R + B−2G) / G value exist inside the dead zone, it is determined that it is not necessary to adjust the white balance, and a return (RET: step ST0) is performed. . Returning waits for input of the next integral value data.
On the other hand, when (R−B) / G value and (R + B−2G) / G value exist outside the dead zone frame, the processing flow is step ST10 in order to perform white balance adjustment even outside the pull-in limit frame. Migrate to

In step ST10, white balance gain adjustment, that is, calculation (setting) of the gain value obtained by the pull-in operation, and signal amplification by the variable gain amplifiers 16r, 16g, and 16b using the gain value are executed.
Specifically, the gain value when the pull-in means 18F in FIG. 2 pulls the integral value outside the pull-in limit frame into the high color temperature side convergence target value HT or the low color temperature side convergence target value LT, that is, the amplifier output color level value. The gain values Gr (i), Gg (i), and Gb (i) of the variable gain amplifiers 16r, 16g, and 16b with R (i), G (i), and B (i) being 1: 1: 1 are calculated. Find (white balance calculation). The relationship between the amplifier output color level value and the gain value is given by the above equations (3-1) to (6-2). Here, i = 1 because it is the first white balance calculation.
The calculated gain values Gr (i), Gg (i), Gb (i) are actually supplied to the variable gain amplifiers 16r, 16g, 16b in FIG. Is amplified or attenuated by the gain values Gr (i), Gg (i), and Gb (i).
Thereafter, the processing flow shifts to return (RET: step ST0) and waits for input of the next integral value data.

The pull-in operation in the out-of-frame control may be performed once for a predetermined convergence target value, for example, the high color temperature side convergence target value HT or the low color temperature side convergence target value LT in FIG. In order to avoid an abrupt change in temperature, the first method described above (time delay method having a time constant) or the second method described above (step pulling operation method for delaying a process having a delay) ) Is executed.
Even in the second method, although not particularly illustrated, a step of incrementing the white balance calculation number i and a step of determining whether this has reached a predetermined number are provided after step ST10. If the predetermined number of times has not been reached, the process flow is returned to step ST4, and from step ST4 to the predetermined number of times determination step, the provisional target values are sequentially set to the convergence target value (high color temperature side convergence target value HT or low color temperature). While being close to the side convergence target value LT), it is repeated until the determination result becomes “Yes”. If the determination result is “Yes”, or if it is determined in step ST9 that it is in the dead zone, the processing flow proceeds to return (RET: step ST0) and is waiting for input of the next integral value data. It becomes.

  The primary color signals R ′, G ′, and B ′ after white balance adjustment as described above are output from the white balance amplifier 16 of FIG. 1, and after this, the primary color signals are subjected to predetermined processing, recording, etc. Is done.

According to the above embodiment, the following benefits can be obtained.
There is a flag indicating that the dead zone is set independently or not on the high color temperature side and the low color temperature side, or a flag indicating that the set dead zone is enabled (for example, the dead zone enabling flag F2), which is a default value. Can be changed by user operation. In this embodiment, since the dead band is used in the feedforward method, there is no viewpoint of oscillation prevention as in the feedback method, and the convergence condition is turned on and off by setting the flag on the high color temperature side and the low color temperature side. Uses independent settings and switching. Therefore, each dead zone has a size and range suitable for this purpose. In addition, the size and range of the dead zone are also set by parameters and can be changed by the user.
The combination of such flag setting and switching pattern, and also the dead zone parameter, when there is a transition between high color temperature and low color temperature, when there is no transition, when there is a transition between high color temperature and low color temperature It can be controlled according to the transition mode such as from which direction the transition has occurred. For this reason, the convergence speed of white balance control can be kept constant in various modes of color temperature transition. In addition, the white balance control value does not change greatly from field to field, and a stable weight balance can be adjusted without a sense of incongruity.
When this is combined with a time delay or processing delay, the visual stability during white balance control is further increased.

Conventionally, the control is such that no pull-in operation is performed outside the pull-in limit frame regardless of whether there is a color temperature transition.
On the other hand, in the present embodiment, when the integrated value is out of the frame after the color temperature transition, it is not possible to determine whether or not this should be the pull-in operation. The drawing operation is started toward the second target value (the high color temperature side convergence target value HT or the low color temperature side convergence target value LT) on the frame or the like. Thereafter, when the subject changes and the integrated value falls within the pull-in limit frame, a full-scale pull-in operation toward the first target value (coordinate origin, etc.) is started immediately. In addition, when it is still outside the pull-in limit frame, since the full pull-in operation is not performed, white balance adjustment is not performed.

Further, even when there is no color temperature transition, it is monitored whether or not out-of-frame control has been executed due to the color temperature transition last time (step ST11 in FIG. 16). As a result of the previous out-of-frame control, if it is determined in the next step ST12 that it is within the pull-in limit frame, a full-scale pull-in operation is immediately performed.
On the other hand, if the determination in step ST12 is “No” and the process flow moves to step ST5, there was no color temperature transition this time, but the out-of-frame control of the pull-in restriction frame was executed last time, and the current This is a case where the integrated value also remains outside the pull-in limit frame. In this case, out-of-frame control after step ST5 is executed.
As described above, the control is different between the case where the frame is out of the frame as a result of the color temperature transition, and the case where the frame is out of the frame because the subject whose color is not to be changed is captured. It effectively prevents a pull-in operation that changes the color of the subject by the control, and realizes white balance control with high followability and responsiveness.
The out-of-frame pull-in permission flag F1 used at this time is set to a combination of on and off according to the color temperature transition mode, so that the white balance can be accurately executed while preventing the color change of the subject. ing. Then, as with the out-of-frame pull-in permission flag F1 and the dead zone activation flag F2, the change by the user is possible.

  Furthermore, by giving the control value as a coordinate value, the influence of the coordinate system slightly deviating for each pre-white balance is eliminated, and accurate white balance control is realized. Also, the pre-white balance / gain value read from the memory is given to the pre-white balance as a coordinate value, and an optimum value is provided for each user preset mode. When the user selects a mode, an optimum gain value is set by this, so that the coordinate system shift itself is reduced and more accurate white balance adjustment is realized. This mode switching information can be used as color temperature information for controlling dead zone control and out-of-frame control.

  As described above, according to the present embodiment, when there is a change in shooting at the low color temperature side and the high color temperature side, the white balance control function is improved by providing a determination to follow the white balance. The white balance performance can be selected according to the user to be used and the system being built.

It is a block diagram of a camera device of an embodiment. It is an execution program functional diagram of a controller. It is a figure which shows the example of the drawing restriction | limiting frame of the low color temperature side and the high color temperature side on a color coordinate, and its transition example. It is a figure which shows the example of a setting of a dead zone frame on a color coordinate. It is explanatory drawing of the color temperature estimation method by the side of high color temperature. It is a chart which shows the mode of judgment according to a flag value. It is a figure which shows the example which becomes out of a frame by color temperature transition. It is operation | movement explanatory drawing according to a dead zone determination result. It is operation | movement explanatory drawing when there is no dead zone. It is operation | movement explanatory drawing at the time of out-of-frame control in case a dead zone is provided. It is operation | movement explanatory drawing at the time of out-of-frame control when a dead zone is not provided. It is a figure which shows the example of a 2nd target value (HT, LT). It is explanatory drawing of the step pull-in operation | movement when a control value is made into an amplifier value. It is a figure for demonstrating the inconvenience in having a control value with an amplifier value. It is a figure for demonstrating the advantage in having a control value by a coordinate value by contrasting with FIG. It is the whole flowchart. It is a flowchart of normal automatic white balance control (AWB).

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Camera apparatus, 12 ... Image sensor, 15 ... Primary color separator, 16 ... White balance amplifier, 16r, 16g, 16b ... Variable gain amplifier, 17 ... Optical detector (OPD), 18 ... Controller, 18A ... Coordinate conversion means, 18B: Color temperature transition detection means, 18C: Pull-in limit frame determination means, 18D: Dead band frame determination means, 18E: Color temperature estimation means, 18F: Pull-in means, HW: High color temperature side pull-in limit frame, LW: Low color Temperature side pull-in limit frame, HT ... High color temperature side convergence target value, LT ... Low color temperature side convergence target value, F1 ... Outside frame allowance flag, F2 ... dead zone activation flag

Claims (17)

A primary color separation circuit that sequentially separates image signals input in time series into primary color signals;
Integrating means for integrating the primary color signal;
Coordinate conversion means for converting the obtained integral value for each color into a coordinate value on a predetermined color coordinate;
Input information on the brightness or color temperature of the image, and a color temperature transition detection means for detecting that the color temperature of the image has transitioned between the high color temperature side and the low color temperature side based on the information;
A gain setting means capable of setting a gain value for white balance adjustment for each primary color signal by drawing the coordinate value or a color temperature value of a light source estimated from the coordinate value into a predetermined target value;
A variable gain amplifier that amplifies the primary color signal for each color at the set gain value;
The gain setting means includes
The position, range and effectiveness of the dead zone where the color temperature adjustment is prohibited within the frame can be set independently on the high color temperature side and low color temperature side on the color coordinates,
Based on the detection result of the color temperature transition, the validity and invalidity of the dead zone is determined on the side of the transitioned color temperature, and when the dead zone on the side of the transitioned color temperature is valid as a result of the determination, the effective dead zone is used. A white balance adjustment circuit for determining whether to draw the coordinate value into the target value. The gain setting means includes
Of the two dead zone enable flags corresponding to the high color temperature side and the low color temperature side, which are held in a rewritable memory in response to an external instruction input via the operation unit, the color temperature transition flag 2. The white balance adjustment circuit according to claim 1, wherein the dead zone validation flag corresponding to the color temperature side that has transitioned based on the detection result is referenced to determine whether the dead zone is valid or invalid. The gain setting means includes
Prior to the pull-in operation, one coordinate value corresponding to the mode designation input from the outside via the operation unit is read from a plurality of different coordinate values corresponding to the user custom preset mode stored in the memory, The white balance adjustment circuit according to claim 1, wherein a gain value for white balance adjustment of the image is calculated based on the read coordinate value. A primary color separation circuit that sequentially separates image signals input in time series into primary color signals;
Integrating means for integrating the primary color signal;
Coordinate conversion means for converting the obtained integral value for each color into a coordinate value on a predetermined color coordinate;
Input information on the brightness or color temperature of the image, and a color temperature transition detection means for detecting that the color temperature of the image has transitioned between the high color temperature side and the low color temperature side based on the information;
A gain setting means capable of setting a gain value for white balance adjustment for each primary color signal by drawing the coordinate value or a color temperature value of a light source estimated from the coordinate value into a predetermined target value;
A variable gain amplifier that amplifies the primary color signal for each color at the set gain value;
The gain setting means includes
The pull-in limit frame with the color temperature adjustment range in the frame can be set on the high color temperature side and the low color temperature side on the color coordinates,
The permission of the pull-in operation outside the pull-in limit frame is determined on the color temperature side that has been transitioned based on the detection result of the color temperature transition, and when the pull-out operation is permitted outside the frame as a result of the determination, the transition has occurred. A white balance adjustment circuit that performs the pull-in even when the coordinate value is located outside the pull-in limit frame on the color temperature side. The gain setting means includes
The detection result of the color temperature transition among the out-of-frame pull-in permission flags corresponding to the high color temperature side and the low color temperature side, which is held in a rewritable memory in response to an external instruction input via the operation unit 5. The white balance according to claim 4, wherein, based on the reference, an out-of-frame pull-in permission flag corresponding to the transitioned color temperature side is referred to, and permission / non-permission of the pull-in operation outside the transition of the transition limit frame on the color temperature side is determined. Adjustment circuit. The gain setting means includes
When the pull-in operation outside the frame of the pull-in restriction frame is permitted, the coordinate value outside the frame is pulled into the second target value that is different from the first target value at the time of the pull-in operation from the frame and the position on the color coordinate. 5. The white balance adjustment circuit according to 5. The gain setting means includes
The second target value is read out and used when performing a retraction operation outside the retraction restriction frame from a rewritable memory in response to an external instruction input via the operation unit. White balance adjustment circuit. The gain setting means includes
In the transition from the high color temperature side to the low color temperature side and the transition from the low color temperature side to the high color temperature side, the relative position with respect to the pull-in restriction frame is different from the outside of the frame using the second target value. The white balance adjustment circuit according to claim 6 or 7, wherein a pull-in operation is performed. The gain setting means includes
When performing the pull-in operation outside the pull-in limit frame based on the detection result of the color temperature transition, the target value of the pull-in operation is determined by the gain value for each color of the variable gain amplifier or the coordinate value of the color coordinate The white balance adjustment circuit according to claim 4, which can be defined. The gain setting means includes
Prior to the pull-in operation, one coordinate value corresponding to the mode designation input from the outside via the operation unit is read from a plurality of different coordinate values corresponding to the user custom preset mode stored in the memory, The white balance adjustment circuit according to claim 4, wherein a gain value for adjusting the white balance of the image is calculated based on the read coordinate value. The gain setting means includes
The position, range and effectiveness of the dead zone where the color temperature adjustment is prohibited within the frame can be set independently on the high color temperature side and low color temperature side on the color coordinates,
Based on the detection result of the color temperature transition, the validity and invalidity of the dead zone is determined on the side of the transitioned color temperature, and when the dead zone on the side of the transitioned color temperature is valid as a result of the determination, the effective dead zone is used. The white balance adjustment circuit according to claim 4, wherein it is determined whether to draw the coordinate value into the target value. The gain setting means includes
Of the two dead zone enable flags corresponding to the high color temperature side and the low color temperature side, which are held in a rewritable memory in response to an external instruction input via the operation unit, the color temperature transition flag The white balance adjustment circuit according to claim 11, wherein the dead zone validation flag corresponding to the color temperature side that has transitioned based on the detection result is referenced to determine whether the dead zone is valid or invalid. The gain setting means includes
When permitting the pull-in operation from outside the pull-in limit frame based on the detection result of the color temperature transition, the first target value at the time of pull-in operation from within the frame is different from the second target value where the position on the color coordinate is different. Is determined according to the detection result, and when the dead zone is valid, the pull-in operation has a time delay, and the dead zone is invalid. The white balance adjustment circuit according to claim 11, wherein the pull-in operation is controlled so as not to have a time delay. An image sensor, an operation unit, a memory rewritable by an external instruction via the operation unit, and a camera device having a signal processing block including a white balance adjustment circuit,
The white balance adjustment circuit
A primary color separation circuit that sequentially separates image signals generated in the image sensor and input in time series into primary color signals;
Integrating means for integrating the primary color signal;
Coordinate conversion means for converting the obtained integral value for each color into a coordinate value on a predetermined color coordinate;
Input information on the brightness or color temperature of the image, and a color temperature transition detection means for detecting that the color temperature of the image has transitioned between the high color temperature side and the low color temperature side based on the information;
A gain setting means capable of setting a gain value for white balance adjustment for each primary color signal by drawing the coordinate value or a color temperature value of a light source estimated from the coordinate value into a predetermined target value;
A variable gain amplifier that amplifies the primary color signal for each color at the set gain value;
The gain setting means includes
Using the frame parameters in the memory, the position, range, and effectiveness of the dead zone where the color temperature adjustment is prohibited within the frame can be set independently on the high color temperature side and low color temperature side on the color coordinates. And
Based on the detection result of the color temperature transition, validity and invalidity of the dead band are determined on the side of the transitioned color temperature with reference to the dead band activation flag in the memory, and the dead band of the transitioned color temperature side as a result of the determination A camera device that determines whether the coordinate value is drawn into the target value by using the effective dead zone when is effective. An image sensor, an operation unit, a memory rewritable by an external instruction via the operation unit, and a camera device having a signal processing block including a white balance adjustment circuit,
The white balance adjustment circuit
A primary color separation circuit that sequentially separates image signals generated in the image sensor and input in time series into primary color signals;
Integrating means for integrating the primary color signal;
Coordinate conversion means for converting the obtained integral value for each color into a coordinate value on a predetermined color coordinate;
Input information on the brightness or color temperature of the image, and a color temperature transition detection means for detecting that the color temperature of the image has transitioned between the high color temperature side and the low color temperature side based on the information;
A gain setting means capable of setting a gain value for white balance adjustment for each primary color signal by drawing the coordinate value or a color temperature value of a light source estimated from the coordinate value into a predetermined target value;
A variable gain amplifier that amplifies the primary color signal for each color at the set gain value;
The gain setting means includes
Using the frame parameter in the memory, a pull-in limit frame having a color temperature adjustment range in the frame can be set on the high color temperature side and the low color temperature side on the color coordinates,
The permission of the pull-in operation outside the pull-in limit frame is determined by referring to the out-of-frame pull-in permission flag in the memory on the color temperature side based on the detection result of the color temperature transition. When the pull-in operation is permitted, the camera device executes the pull-in even when the coordinate value is located outside the transition limit frame on the color temperature side that has been changed. A primary color separation step for sequentially separating image signals input in time series into primary color signals;
A coordinate conversion step of integrating the primary color signal and converting the obtained integrated value for each color into a coordinate value on a predetermined color coordinate;
Inputting information about the brightness or color temperature of the image, and detecting a color temperature transition that detects that the color temperature of the image has transitioned between the high color temperature side and the low color temperature side based on the information;
A gain setting step for setting, for each primary color signal, a gain value for white balance adjustment of the image by drawing the coordinate value or the color temperature value of the light source estimated from the coordinate value into a predetermined target value;
A signal amplifying step for amplifying the primary color signal for each color with the set gain value;
In the gain setting step,
The position, range and effectiveness of the dead zone where the inside of the frame is the prohibited range of color temperature adjustment is set independently on the high color temperature side and low color temperature side on the color coordinates,
Determine the validity and invalidity of the dead zone on the side of the color temperature transitioned based on the detection result of the color temperature transition,
As a result of the determination, after determining whether or not the coordinate value is drawn into the target value using the effective dead zone when the dead zone on the color temperature side that has been transitioned is valid,
A white balance adjustment method in which a gain value is set for each color by the pull-in. A primary color separation step for sequentially separating image signals input in time series into primary color signals;
A coordinate conversion step of integrating the primary color signal and converting the obtained integrated value for each color into a coordinate value on a predetermined color coordinate;
Inputting information about the brightness or color temperature of the image, and detecting a color temperature transition that detects that the color temperature of the image has transitioned between the high color temperature side and the low color temperature side based on the information;
A gain setting step for setting, for each primary color signal, a gain value for white balance adjustment of the image by drawing the coordinate value or the color temperature value of the light source estimated from the coordinate value into a predetermined target value;
A signal amplifying step for amplifying the primary color signal for each color with the set gain value;
In the gain setting step,
Set the pull-in limit frame with the color temperature adjustment range inside the frame on the high color temperature side and low color temperature side on the color coordinates,
Permission of pull-in operation outside the pull-in limit frame is determined on the color temperature side that has transitioned based on the detection result of the color temperature transition,
As a result of the determination, when the drawing operation outside the frame is permitted, the gain value for each color is obtained by executing the drawing even when the coordinate value is located outside the drawing limit frame on the transitioned color temperature side. Set the white balance adjustment method.

Images