JP4136775B2 - Correction apparatus, imaging apparatus, correction method, computer-readable storage medium, and program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to a correction apparatus, an imaging apparatus, a correction method, a computer-readable storage medium, and a program, and in particular, an imaging surface is divided into a plurality of parts and an amplifier that amplifies an imaging signal of each area and is connected to this output The present invention is suitable for use in a correction device that corrects a signal from a solid-state imaging device having a plurality of imaging signal output terminals.
[0002]
[Prior art]
In recent years, due to advances in digital signal processing technology and semiconductor technology, consumer digital video standards for digital recording of standard television, for example, NTSC and PAL video signals, have been proposed. Digital video cameras in which a device and an imaging device are integrated have been commercialized. Some digital video cameras have a still image recording function by taking advantage of the feature of digital recording.
[0003]
Some have a digital I / F for connecting to a computer or the like, and have a function of taking a captured image into the computer. Furthermore, an apparatus that includes a plurality of recording media and that can select a recording medium according to the purpose of use of an image has been put into practical use.
[0004]
In such a device, when a recorded image is reproduced by connecting to a television, the image size is determined by the digital video standard, for example, 720 × 480 pixels, but there is no problem, but via a digital I / F When transferring an image to other media, a larger number of pixels may be required due to image quality problems.
[0005]
Accompanying the increase in the number of pixels in an image sensor, it is necessary to drive the image sensor at a higher frequency in order to read out information on all pixels of the image sensor. There was a problem that caused an increase in electric power.
[0006]
Therefore, a method of increasing the data rate of imaging information while keeping the driving frequency of the imaging element low is considered. As one of such methods, there is a method in which the imaging surface is divided into a plurality of regions, and independent charge transfer units, amplifiers, and output terminals are provided in the respective regions, and the imaging signals are read out in parallel.
[0007]
FIG. 14 shows an example of an imaging apparatus using the imaging element as described above. In FIG. 14, the imaging surface of the imaging device 1400 is divided into two regions on the left and right. 1401 and 1402 are photoelectric conversion and vertical transfer units, 1403 and 1404 are horizontal transfer units, 1405 and 1406 are amplifiers, and 1407 and 1408 are output terminals. By using the imaging device having such a structure, there is an advantage that imaging information having a data rate twice as high as the driving frequency of the imaging device can be obtained.
[0008]
On the other hand, as a disadvantage of this method, due to non-uniformity of the characteristics of the amplifier and external peripheral circuit in each area, when an image is generated by combining two areas, a boundary line is generated due to a level difference between the areas. There was a problem that image quality deteriorated.
[0009]
As a method of reducing the image quality degradation due to these non-uniformities, there is a method in which the correction coefficient is obtained by measuring the black level and the standard white level of each area in advance, and the non-uniformity is corrected by this correction coefficient during imaging. It is considered.
[0010]
FIG. 14 shows a configuration example of such a correction circuit. A subject image formed on the image sensor 1400 by an imaging optical system (not shown) is converted into an electrical signal by the image sensor 1400, and an output terminal 1407 and an output terminal 1407 according to a drive pulse supplied from a drive timing generation circuit (not shown). 1408 is output.
[0011]
The two systems of image signals obtained from the image sensor 1400 are subjected to analog signal processing by the analog signal processing units 1409 and 1410 and then AD-converted and supplied to the black level correction circuits 1411 and 1412 and the black level difference detection circuit 1413. Is done. The black level difference detection circuit 1413 detects a black level difference from two systems of image signals and calculates a correction coefficient.
[0012]
This correction coefficient is supplied to black level correction circuits 1411 and 1412, and the black level difference is corrected based on the correction coefficient. For detection of the black level difference, a signal of an optical black pixel of the image sensor 1400 is used. The detection and the correction value calculation are performed only once at a predetermined time, and the obtained correction coefficient is stored in the memory 1420, so that the black level is determined by the correction coefficient stored in the memory 1420 without performing detection at the subsequent photographing. Difference correction is performed.
[0013]
Next, each signal is supplied to white level correction circuits 1414 and 1415 and a white level difference detection circuit 1416. The white level difference detection circuit 1414 detects a difference in white level from two types of image signals and calculates a correction coefficient. This correction coefficient is supplied to black and white level correction circuits 1414 and 1415, and the difference in white level is corrected based on the correction coefficient.
[0014]
To detect the difference in white level, the image sensor 1400 is irradiated with uniform light that provides a standard white level, and the image signal at that time is used. The detection and the correction value calculation are performed only once at a predetermined time, and the obtained correction coefficient is stored in the memory 1421, so that the detection is not performed at the time of subsequent photographing and the correction coefficient stored in the memory 1421 is used. White level difference correction is performed.
[0015]
The white level corrected signal is subjected to γ correction processing, contour correction processing, color correction processing, and the like by the camera signal processing circuit 1418 after the left and right images are combined as a single image by the screen combining circuit 1417. The luminance signal and the color difference signal are output from the output terminal 1419.
[0016]
[Problems to be solved by the invention]
However, in the above conventional example, since the correction coefficient is calculated only under a predetermined condition such as capturing a standard white image, there is a problem that the real-time property is lacking. For this reason, dynamic fluctuations such as temperature fluctuations or fluctuations with time cannot be dealt with, and unevenness between regions may not be sufficiently corrected.
[0017]
The present invention has been made in view of the above-mentioned problems, and eliminates subject-dependent level difference components and determines the correction coefficient to cope with dynamic fluctuations such as temperature fluctuations or temporal fluctuations. An object is to enable real-time correction of non-uniformity among a plurality of imaging regions.
[0018]
[Means for Solving the Problems]
  The correction device according to the present invention is a correction device that corrects a plurality of imaging signals from a plurality of output units of an imaging element, and includes an output level detection unit that detects levels of the plurality of imaging signals, and the plurality of imaging signals. Level adjusting means for adjusting the level of the image pickup signal, and correction coefficient determining means for determining a correction coefficient for reducing a level difference between the plurality of imaging signals, wherein the correction coefficient determining means is the output level detecting means. A gain error calculation means for calculating a gain error between the plurality of imaging signals from the level detection results of the plurality of imaging signals output from, and when the gain error calculated by the gain error calculation means is input, When a gain error between the imaging signals exceeds a threshold for allowing a level difference between the plurality of imaging signals, a predetermined reference value is output, and the gain error is the threshold A non-linear processing means for outputting the input gain error as it is when not exceeding, determining the correction coefficient based on the signal output from the non-linear processing means, and the correction determined by the correction coefficient determining means A coefficient is given to the level adjusting means to perform adjustment so that a level difference between the plurality of image pickup signals becomes small.
  The correction method of the present invention is a correction method for correcting a plurality of imaging signals from a plurality of output units of an image sensor, and includes an output level detection step for detecting levels of the plurality of imaging signals, and the plurality of imaging signals. A level adjustment step for adjusting the level of the image pickup signal, and a correction coefficient determination step for determining a correction coefficient for reducing a level difference between the plurality of imaging signals. The correction coefficient determination step includes the output level detection step. A gain error calculation step of calculating a gain error between the plurality of imaging signals from a level detection result of the plurality of imaging signals output in the step, and a gain error calculated in the gain error calculation step When a gain error between the imaging signals exceeds a threshold for allowing a level difference between the plurality of imaging signals, a predetermined reference value is output, and the gain error is A non-linear processing step that outputs the input gain error as it is when the threshold value is not exceeded, determining the correction coefficient based on the signal output in the non-linear processing step, and the determined correction coefficient The level adjustment step is performed so that the level difference between the plurality of imaging signals is reduced.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Next, an embodiment in which the correction of the present invention is applied to an imaging apparatus will be described with reference to the accompanying drawings.
(First embodiment)
FIG. 1 is a diagram schematically showing an embodiment in which the correction apparatus of the present invention is applied to a single-panel video camera.
In FIG. 1, 100 is a CCD area sensor having an imaging region divided into two and each having an output terminal, 101 is a photoelectric conversion unit and vertical transfer unit, and 103 and 104 are horizontal transfer units, with the center of the screen as a boundary. It is divided into two in the left-right direction.
[0020]
  Reference numerals 105 and 106 denote output amplifiers that amplify the signal charge, and 107 and 108 denote output terminals for the imaging signal.(Output part)It is. Reference numerals 109 and 110 denote analog front ends that perform AD conversion with the correlated double samples. Reference numerals 111 and 112 denote black level detection and correction means; 113 and 114, gain adjustment means for adjusting gain; and 115, screen composition means for synthesizing two image signals to generate one image.The black level detection and correction means 111 and 112 and the gain adjustment means 113 and 114 function as level adjustment means for adjusting the level of the imaging signal.
[0021]
116 is a step evaluation value generating means for detecting non-uniformity between the two systems, 117 is a microcomputer for controlling the system, 118 is a camera signal processing means, 119 is an output terminal, and 120 is a rewritable nonvolatile memory. It is memory. In the present embodiment and later-described embodiments, a gain adjusting unit 113, 114, a step evaluation value generating unit 116, and a microcomputer 117 constitute a correction device for detecting and correcting nonuniformity between two systems. is doing.
[0022]
  next,in frontThe operation of the video camera of the present embodiment in the above configuration will be described.
  An object image formed on the CCD 100 by an imaging optical system (not shown) is converted into an electric signal by the photoelectric conversion unit 101 and then transferred horizontally.PartDivided into two systems by 103 and 104 and supplied to output amplifiers 105 and 106.
[0023]
The signal charge is amplified to a predetermined level by the output amplifiers 105 and 106 and output from the first output terminal 107 and the second output terminal 108. Hereinafter, an imaging signal obtained from the first output terminal 107 is referred to as a left channel signal, and an imaging signal obtained from the second output terminal 108 is referred to as a right channel signal.
[0024]
The left and right two-line imaging signals are subjected to correlated double sample processing and AD conversion by the analog front ends 109 and 110, and then supplied to the black level detection and correction means 111 and 112. The black level detection and correction means 111 and 112 use the dummy signal portion or the optical black signal portion of the image pickup signal so that the black level of the two systems of the image pickup signal coincides with “0” of the digital code. Correction is performed. As a result, the error of the offset component between the two systems is removed.
[0025]
The signal whose black level is corrected is subjected to gain adjustment by the gain adjusting means 113 and 114. The gain applied at the time of gain adjustment is supplied from the microcomputer 117. In the conventional imaging apparatus, the gain of the signal amount in a low illumination environment is increased by an analog circuit. However, in the imaging apparatus that handles two systems of imaging signals as in the present embodiment, the gain adjustment by the analog circuit is performed. Can cause non-uniformity between the two systems. Therefore, in the present embodiment, gain adjustment is performed by digital calculation using the gain adjusting means 113 and 114, thereby eliminating the influence of circuit variations, temporal variations, and temperature variations.
[0026]
Further, not only gain adjustment for image brightness but also correction of gain error between the two systems is performed here. In general, the gain difference between the two systems depends on the output level of the CCD area sensor 100.
[0027]
FIG. 3 is a characteristic diagram showing an example of an output level between two systems and a gain difference between channels. In FIG. 3, the horizontal axis is the left channel output level of the CCD 100, and the vertical axis is the ratio of the input signal (left channel) of the gain adjusting means 114 to the input signal (right channel) of the gain adjusting means 113, that is, 2 The signal level gain difference between systems is shown.
[0028]
For example, assuming that the left channel output level of the CCD 100 when an object of a certain brightness is imaged is L0 and the right channel output level is L0right, the gain difference E0 at this time is given by the following equation (1).
E0 = L0right / L0 (1) formula
[0029]
As shown in this figure, since the relationship between the signal level and the gain difference is not constant, the gain correction amount is not a fixed value, and it is necessary to vary the correction amount according to the gain increase amount.
[0030]
In this embodiment, the reference level Lref is set for the signal after gain adjustment, and the level difference between the two systems is always 0 at the reference level Lref regardless of the gain increase amount, that is, the signal of each channel is set to the reference level Lref. Gain correction is performed so that they match. As the reference level Lref, a gray level of about 75% is selected after the γ correction with respect to the reference white.
[0031]
For example, when the left channel output level of the CCD 100 is L0 and the gain is increased so that the output level of the gain adjusting unit 114 becomes the reference level Lref, the gain A0 given to the left channel gain adjusting unit 114 can be expressed by the following equation. .
A0 = reference level Lref / L0 (2)
[0032]
Further, the gain A0right given to the gain adjusting unit 113 of the right channel at this time can be expressed by the following equation, with the gain correction amount being C0.
A0right = A0 x C0 (3)
And C0 is calculated | required by following Formula.
C0 = 1.0 / E0 (4) formula
[0033]
Similarly, when the left channel output level of the CCD 100 is L1, the gain correction amount C1 when the gain increase amount is such that the output level of the gain adjusting means 114 becomes the reference level Lref is obtained by the following equation.
C1 = 1.0 / E1 (5)
[0034]
FIG. 4 shows a characteristic example of the gain correction amount with respect to the gain increase amount. This correction characteristic is different for each component of the CCD 100 or the analog front ends 109 and 110.
[0035]
  Next, measurement of gain correction characteristics will be described.
  The step evaluation value generation means 116Functions as output level detection means, as will be described later in FIG.Within the specified rectangular area near the boundary of the divided areaFramePixel valueDetect and its detected resultThe evaluation value of the screen level difference is calculated based on the above and output to the microcomputer 117.
[0036]
  FIG. 2 shows an example of a rectangular area in the screen. As shown in FIG. 2, rectangular areas 203 and 204 are arranged in the vicinity of the boundary between the divided areas 201 and 202.I.e. number of framesIs setThe This setting is set by a frame number setting means (not shown). AndPixel values in this area are used for evaluation of the screen level difference.
[0037]
  In the CCD 100, an on-chip color filter is attached to the pixel portion in order to capture a color image with a single plate.in frontThe on-chip color filter has an arrangement as shown by 205 in FIG. The step evaluation value generation means 116 selects a pixel value of one of these colors,Generate an evaluation value,In the areaAverage the evaluation valuesThe average value is calculated, and this is used as the evaluation value of the screen level difference.
[0038]
When measuring the gain correction characteristic, a subject with uniform brightness is imaged, and the microcomputer 117 sets the same gain multiplier to the gain adjusting means 113 and 114. The average level of the pixels in one rectangular area 203 is set to the left channel level, and the average level of the pixels in the other rectangular area 204 is set to the right channel level and output to the microcomputer 117.
[0039]
  The microcomputer 117 calculates the gain correction amount of the right channel as described above with reference to the level of the left channel. Such a measurement is performed at a predetermined interval at the output level of the CCD 100.(timing)In this way, a gain correction characteristic is generated.
[0040]
  The microcomputer 117 isAs correction coefficient storage meansThe generated gain correction characteristic is stored in a rewritable nonvolatile memory 120 such as an EEPROM (Electrically Erasable Programmable Read Only Memory). The generation of the gain correction characteristic is executed at the time of factory adjustment, for example. Therefore, the gain difference remains as an error without being able to cope with dynamic fluctuations such as aging fluctuations and temperature fluctuations.
[0041]
Next, correction of the residual gain error during general imaging will be described.
FIG. 5 shows a correction coefficient determined by correcting the subject-dependent level difference component, and the level of each imaging signal output from a different output terminal of the CCD area sensor by applying the determined correction coefficient to the gain adjusting means. It shows the configuration of a block for correcting a residual gain error, which is executed by a microcomputer 117 which is a correction coefficient determining means for performing adjustment so as to reduce the difference. Signals A, B, C, and D in FIG. 5 correspond to signals A, B, C, and D in FIG. 1, where symbol A is a left channel step evaluation value, symbol B is a right channel step evaluation value, Symbol C is a gain adjustment value for the left channel, and symbol D is a gain adjustment value for the right channel.
[0042]
  The left channel step evaluation value A and the right channel step evaluation value B input to the microcomputer 117 are gain errors.CalculationThe gain error amount E is obtained by inputting to the means 501. The gain error amount E is obtained by the following equation.
E = B / A (6)
[0043]
  Gain errorCalculationThe gain error amount E obtained by the means 501 is simply a ratio of pixel levels, and is affected not only by the non-uniformity between channels but also by the level difference of the subject itself. Therefore, in order to correct the gain error correctly, it is necessary to eliminate the subject-dependent level difference component. In this embodiment, the subject-dependent level difference component is used as a limiter means.(Threshold setting means and nonlinear processing means)502 and the integration means 503 are excluded.
[0044]
  Limiter means(Nonlinear processing means)The input / output characteristics of 502 are shown in FIG. The origin in FIG. 6 represents a point where limiter input = limiter output = 1.0. Since the ratio is the level between channels, the value when there is no gain error is 1.0.
[0045]
  As shown in FIG. 6, the ratio of the level differences is the threshold TH.SuperThe limiter output is 1.0. The threshold TH is associated with the residual gain error amount.By the threshold setting means described aboveIt is determined. By this processing, those having a large level difference are regarded as subject-dependent level differences and are eliminated.
[0046]
FIG. 7 shows the internal configuration of the integrating means 503. The input signal X (0) is compared with the signal Y (−1) delayed by a predetermined time in the subtracting means 701, and then multiplied by a coefficient k in a coefficient unit 702. The output of the coefficient unit 702 is added to the delay signal by the adding means 703 and becomes an output while being supplied to the delay means 704. The output signal is expressed as an equation as Y (0) as follows.
Y (0) = kX (0) + (1-k) Y (-1) (0 <k <1) (7)
[0047]
The delay time is equal to the CCD vertical scanning period. By this process, an average value of error amounts for the past 1 / k frames is obtained. Normally, the subject is not fixed for a long time in the angle of view, and by taking an average of a plurality of frames, the subject-dependent level difference component is canceled and eliminated.
[0048]
Through the processing as described above, a level difference due to a subject is eliminated, and a gain error caused by non-uniformity between channels is extracted. The gain error amount is then multiplied by a coefficient by the correction amount control means 504. This coefficient corresponds to the feedback gain of the gain error correction loop. When the gain is large, the correction capability is high, but becomes unstable against disturbance such as false detection, and when the gain is small, it is stable against disturbance, but the correction capability is low.
[0049]
The output of the correction amount control means 504 is supplied to the gain correction amount calculation means 506.
The output of the gain correction characteristic table 505 is also supplied to the gain correction amount calculation means 506. The gain correction characteristic table 505 is a table of the gain correction characteristics already described, and as illustrated in FIG. 4, a gain correction amount can be obtained corresponding to the gain increase amount.
[0050]
In the gain correction amount calculation means 506, the gain adjustment value for the right channel is actually calculated by multiplying these two input signals and the gain increase amount. Then, the gain adjustment value calculated in this way is supplied to the gain adjustment means 113 shown in FIG. The gain adjustment unit 114 is supplied with the gain increase amount itself.
[0051]
The signal after gain adjustment is supplied to the screen composition unit 115 and the step evaluation value generation unit 116. The screen synthesizing unit 115 synthesizes two signals and outputs them to the camera signal processing circuit 118 as one screen image. The camera signal processing circuit 118 performs signal processing such as γ correction, color correction, and contour correction, and outputs the image signal from the terminal 119.
[0052]
(Second Embodiment)
FIG. 8 is a signal processing block diagram for explaining the second embodiment of the correction apparatus of the present invention. The specific configuration of the entire imaging apparatus is the same as that of the first embodiment. The signal processing shown in FIG. 8 is executed in the microcomputer 117 in FIG. Also, the rectangular area for evaluation value measurement is the same as in the first embodiment.
[0053]
The gain error between the two systems is not constant with respect to the output level of the CCD 100 as shown in FIG. On the other hand, there are subjects with various brightness in the natural image at the time of general photographing. Therefore, when a gain error is measured from a general natural image, a value obtained by multiplying the gain error characteristic curve shown in FIG. 3 by the brightness distribution frequency in the rectangular area is obtained, and a correct gain error amount cannot be calculated. There is a case. Therefore, an embodiment in which the brightness of the image is taken into account in calculating the gain error amount will be described.
[0054]
  The left channel step evaluation value A and the right channel step evaluation value B input to the microcomputer 117 are gain errors.CalculationInput to the means 801 to obtain the gain error amount. Gain errorCalculationSince the configurations and operations of the means 801 and the limiter means 802 are the same as those in the first embodiment, the description thereof is omitted.
[0055]
The output of the limiter unit 802 is input to the integration unit 803. The right channel step evaluation value B is also input to the integrating means 803 at the same time. The internal configuration of the integrating means 803 is shown in FIG. The configuration and operation other than the coefficient control unit 901 are the same as those in the first embodiment. 9, 902 corresponds to the subtracting means 701 in FIG. 7, 903 corresponds to the coefficient unit 702 in FIG. 7, 904 corresponds to the adding means 703 in FIG. 7, and 905 corresponds to the delay means in FIG. Corresponding to 704.
[0056]
  FIG. 10 shows coefficient control means.(Frame number control means)A coefficient control characteristic 901 is shown. In FIG. 10, the horizontal axis represents the right channel step evaluation value B, and the vertical axis represents the coefficient supplied to the coefficient unit 903. The reference level Lref described in the graph indicates the reference level described in the measurement of the gain correction characteristic of the first embodiment.
[0057]
As shown in FIG. 10, when the right channel step evaluation value is within the range of the threshold value d centering on the reference level, the output coefficient becomes a predetermined value k, but when it deviates from that level, 0 is output. Is output. In the vicinity of the threshold value, an intermediate value between 0 and k is output in order to prevent unstable operation due to a rapid change in coefficient.
[0058]
By performing such coefficient control, the evaluation value is integrated only when the brightness of the image is close to the reference level, and those outside the reference level are excluded from the integration. As a result, the residual error can be corrected with high accuracy with respect to dynamic fluctuations at the reference level at the time of initial adjustment.
[0059]
The output of the integration means 803 is then multiplied by a coefficient in the correction amount control means 804. The operations of the correction amount control unit 804, the gain correction characteristic table 805, and the gain correction amount calculation unit 806 are the same as those in the first embodiment.
[0060]
Returning to FIG. 1, the obtained left channel gain adjustment value C and right channel gain adjustment value D are supplied to gain adjustment means 114 and 113, respectively.
[0061]
(Third embodiment)
FIG. 11 is a block diagram illustrating a configuration example of an apparatus that performs signal processing of the imaging apparatus according to the third embodiment of this invention. The specific configuration of the entire imaging apparatus is the same as that of the first embodiment. FIG. 11 shows a configuration example of the step evaluation value generation means 116 in FIG. The rectangular area for evaluation value measurement is the same as that in the first embodiment.
[0062]
As described above, since the gain error between the two systems is not constant with respect to the output level of the CCD 100, if a level difference is measured from a general natural image, a correct gain error amount may not be calculated. Therefore, an embodiment in consideration of the brightness of the image will be described.
[0063]
The outputs of the gain adjusting means 113 and 114 correspond to the left channel input and the right channel input in FIG. 11, respectively. Since the left channel and the right channel have the same configuration and operation, the left channel will be described here.
[0064]
The left channel input is input to integrating means 1102 (1107) and level detecting means 1104 (1109), respectively. The integrator 1102 (1107) integrates the input signal only when both the gate signal generated by the gate generator 1101 (1106) and the level detection signal generated by the level detector 1104 (1109) are “true”. When either one of the gate signal and the level detection signal is “false”, the input signal is not taken into the integrator.
[0065]
The gate generating means 1101 (1106) operates in synchronization with the horizontal scanning and vertical scanning of the image signal, and the output is “true” during the period of the rectangular region 203 shown in FIG. 2 and the color filter is in a desired color state. Otherwise, it operates so that the output becomes “false”. With this signal, only one color input signal in the rectangular area is selected as an integration target.
[0066]
  Further, the level detection means 1104 (1109) operates so that the output becomes “true” only when the input signal is larger than the lower limit level Llim and smaller than the upper limit level Ulim, and otherwise the output becomes “false”. is there. This signal is selected as an integration target only when the input signal is within a certain reference level range. The lower limit level Llim and the upper limit level Ulim are set with an appropriate margin (for example, ± 5%) with respect to the reference level Lref described in the measurement of the gain correction characteristic of the first embodiment.This setting is set by upper / lower limit level setting means (not shown).
[0067]
With these signals, only a signal having a desired color in the rectangular area of the input signal and having a level within the reference range is selected as an integration target. On the other hand, the gate signal and the level detection signal are input to the pixel counting unit 1105 (1110. In the pixel counting unit, the count is reset at the beginning of the rectangular area, and both the gate signal and the level detection signal are “true”. The counter is incremented, so the counter output indicates the number of integrated input signals.
[0068]
The outputs of the integrating means 1102 and the pixel counting means 1105 (1110) are then supplied to the normalizing means 1103 (1108). In the normalizing means 1103 (1108), the average value is calculated by dividing the integral signal by the counter output. The calculated average value and counter output are output to the microcomputer 117 as a step evaluation value for the left channel.
[0069]
  FIG. 12 shows a block diagram of signal processing executed by the microcomputer 117. The left channel step evaluation value A and the right channel step evaluation value B input to the microcomputer 117 are gain errors.CalculationInput to the means 1201 to obtain the gain error amount. Gain errorCalculationSince the configuration and operation of the means 1201 and the limiter means 1202 are the same as those in the first embodiment, a detailed description is omitted.
[0070]
The output of the limiter unit 1202 is input to the integration unit 1203. The integration unit 1203 also receives the left channel count value and the right channel count value obtained from the step evaluation value generation unit 116 at the same time. The internal configuration of the integrating means 1203 is shown in FIG. The configuration and operation other than the coefficient control unit 1301 are the same as those in the first embodiment. 13 corresponds to the subtracting means 701 in FIG. 7, 1303 corresponds to the coefficient unit 702 in FIG. 7, 1304 corresponds to the adding means 703 in FIG. 7, and 1305 denotes the delay means in FIG. Corresponding to 704.
[0071]
The coefficient control means 1301 outputs 0 when either the left channel count value or the right channel count value is 0, and otherwise outputs a predetermined coefficient k.
[0072]
By performing such coefficient control, the evaluation value is excluded from the integration when there is no pixel having a level near the reference level in the rectangular area. As described above, by selecting only pixels having a level close to the reference level in the rectangular area and generating the evaluation value, it is possible to correct the residual error with respect to the dynamic fluctuation with high accuracy.
[0073]
The output of the integration unit 1203 is then multiplied by a coefficient by the correction amount control unit 1204. The operations of the correction amount control unit 1204, the gain correction characteristic table 1205, and the gain correction amount calculation unit 1206 are the same as those in the first embodiment. Returning to FIG. 1, the obtained left channel gain adjustment value C and right channel gain adjustment value D are supplied to gain adjustment means 114 and 113, respectively.
[0074]
(Another embodiment of the present invention)
The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device.
[0075]
Further, in order to operate various devices so as to realize the functions of the above-described embodiments, a transmission medium such as the Internet is transmitted from a storage medium to an apparatus connected to the various devices or a computer in the system. The program implemented by operating the various devices according to the program stored in the computer (CPU or MPU) of the system or apparatus is supplied via the software program code for realizing the functions of the above-described embodiments Are included in the scope of the present invention.
[0076]
In this case, the program code of the software itself realizes the functions of the above-described embodiments, and the program code itself and means for supplying the program code to the computer, for example, the program code are stored. This storage medium constitutes the present invention. As a storage medium for storing the program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.
[0077]
Further, by executing the program code supplied by the computer, not only the functions described in the above-described embodiments are realized, but also the OS (operating system) or other operating system in which the program code is running on the computer. It goes without saying that the program code is also included in the embodiment of the present invention even when the functions described in the above-described embodiment are realized in cooperation with application software or the like.
[0078]
Further, after the supplied program code is stored in the memory provided in the function expansion board of the computer or the function expansion unit connected to the computer, the CPU provided in the function expansion board or function expansion unit based on the instruction of the program code Etc. perform part or all of the actual processing, and the functions of the above-described embodiment are realized by the processing.
[0079]
Examples of embodiments of the present invention are listed below.
[Embodiment 1] A correction device for correcting a plurality of imaging signals from a plurality of output units of an imaging device,
A level adjusting means for adjusting the levels of the plurality of imaging signals; and a correction coefficient determining means for determining a correction coefficient for reducing a level difference between the plurality of imaging signals. The correction coefficient is determined by correcting a subject-dependent level difference component, and the determined correction coefficient is applied to the level adjustment means to perform adjustment so that the level difference of each image pickup signal is reduced. Correction device.
[Embodiment 2] It has an output level detection means for detecting an output level of the plurality of imaging signals,
2. The correction apparatus according to claim 1, wherein the correction coefficient determination unit determines a correction coefficient so that a level difference between the plurality of imaging signals is reduced based on the level detection result.
[0080]
[Embodiment 3] The output level detection means includes an area selection means for selecting a predetermined area in the plurality of imaging signals, and an average level calculation for calculating an average level in the area selected by the area selection means. And a correction device according to the second embodiment.
[0081]
[Embodiment 4] The output level detection means includes color signal selection means for selecting a predetermined color signal in the plurality of imaging signals,
4. The correction apparatus according to claim 1, wherein the output level detection result is generated based on a color signal selected by the color signal selection unit.
[0082]
[Embodiment 5] The correction coefficient determining means includes a gain error calculating means for calculating a gain error between the respective imaging signals from a plurality of detection results, and a threshold setting for setting a threshold value for allowing a level difference between the respective imaging signals. And a gain error signal calculated by the gain error calculation means is input, and a reference value is output when the gain error signal exceeds a threshold value, and when the gain error signal does not exceed the threshold value, an input gain error is output. A non-linear processing means for outputting
The correction apparatus according to any one of Embodiments 1 to 4, wherein a correction coefficient is determined based on a signal output from the nonlinear processing means.
[0083]
[Embodiment 6] The correction coefficient determining means includes an evaluation value generating means for generating an evaluation value from a detection result of the output level detecting means, and an average of the evaluation values generated by the evaluation value generating means between a plurality of frames. And averaging means for setting, and a frame number setting means for setting the number of frames to be averaged in the averaging means,
The correction apparatus according to any one of embodiments 1 to 5, wherein a correction coefficient is determined based on an evaluation value obtained by averaging the number of frames set by the frame number setting unit.
[0084]
[Embodiment 7] The correction according to Embodiment 6, wherein the averaging means includes frame number control means for controlling the number of frames to be averaged according to the level of the detection result by the output level detecting means. apparatus.
[0085]
[Embodiment 8] The output level detection means has upper and lower limit level setting means for setting an upper limit level and a lower limit level,
The output level detection result is generated by selecting a signal whose level is lower than the upper limit level and higher than the lower limit level from among the plurality of imaging signals. The correction apparatus as described.
[0086]
[Embodiment 9] A correction device that outputs a plurality of imaging signals,
Based on a plurality of level adjusting means for independently adjusting the levels of the plurality of imaging signals, an output level detecting means for detecting an output level of the plurality of level adjusting means, and a detection result of the output level detecting means Correction coefficient determining means for determining a correction coefficient so as to reduce the level difference of each imaging signal, and correction coefficient storage means for storing the correction coefficient in a recording medium,
An adjustment value determined based on both the correction coefficient stored in the recording medium and the correction coefficient obtained in real time during the imaging operation is given to the level adjustment means so that the level difference between the respective imaging signals is reduced. Correction apparatus characterized by performing various adjustments.
[0087]
[Embodiment 10] The correction apparatus according to Embodiment 9, wherein the correction coefficient storage means stores the correction coefficient determined by the correction coefficient determination means at a predetermined timing in the recording medium.
[0088]
[Embodiment 11] The correction coefficient determination unit determines an adjustment value to be supplied to the level adjustment unit by multiplying a correction coefficient stored in the recording medium by a correction coefficient obtained in real time during an imaging operation. The correction device according to Embodiment 9 or 10, characterized in that.
[0089]
[Embodiment 12] The output level detection means includes an area selection means for selecting a predetermined area in the plurality of imaging signals, and an average level calculation for calculating an average level in the area selected by the area selection means. The correction apparatus according to any one of Embodiments 9 to 11, wherein the correction apparatus comprises: means.
[0090]
[Embodiment 13] The output level detection means includes color signal selection means for selecting a predetermined color signal in the plurality of imaging signals,
13. The correction apparatus according to any one of embodiments 9 to 12, wherein an output level detection result is generated using the color signal selected by the color signal selection unit.
[0091]
[Embodiment 14] The correction coefficient determination means includes a gain error calculation means for calculating a gain error between the respective imaging signals from a plurality of detection results, and a threshold setting for setting a threshold value for allowing a level difference between the respective imaging signals. And a non-linear processing means for outputting a reference value when the signal of the gain error detecting means is input and the gain error exceeds a threshold value, and outputting the input gain error as it is when the gain error does not exceed the threshold value. Prepared,
The correction apparatus according to any one of embodiments 9 to 13, wherein a correction coefficient is determined based on an output signal of the nonlinear processing means.
[0092]
[Embodiment 15] The correction coefficient determining means includes: averaging means for averaging output level detection results between a plurality of frames; and frame number setting means for setting the number of frames to be averaged by the averaging means. Prepared,
The correction device according to any one of embodiments 9 to 14, wherein a correction coefficient is determined based on a level obtained by averaging the number of frames set by the frame number setting means.
[0093]
[Embodiment 16] The correction according to embodiment 15, wherein the averaging means comprises frame number control means for controlling the number of frames to be averaged according to the level of the detection result by the output level detecting means. apparatus.
[0094]
[Embodiment 17] The output level detection means includes upper and lower limit level setting means for setting an upper limit level and a lower limit level, and an upper limit level and a lower limit level set by the upper and lower limit level setting means. 17. The correction apparatus according to any one of Embodiments 9 to 16, wherein a detection result is generated by selecting only a signal whose level is lower than the upper limit level and higher than the lower limit level.
[0095]
[Embodiment 18] A correction method for correcting a plurality of imaging signals from a plurality of output units of an imaging device,
A level adjustment process for adjusting the levels of the plurality of image pickup signals; and a correction coefficient determination process for determining a correction coefficient for reducing a level difference between the plurality of image pickup signals. And correcting the subject-dependent level difference component to determine the correction coefficient, and applying the determined correction coefficient to the level adjustment process so as to reduce the level difference of each image pickup signal. Correction method to do.
[Embodiment 19] It has an output level detection process for detecting an output level of the plurality of imaging signals,
The correction method according to embodiment 18, wherein the correction coefficient determination processing determines a correction coefficient so that a level difference between the plurality of imaging signals becomes small based on the level detection result.
[0096]
[Embodiment 20] The output level detection process includes an area selection process for selecting a predetermined area in the plurality of imaging signals, and an average level calculation for calculating an average level in the area selected by the area selection process. The correction method according to claim 19, further comprising: processing.
[0097]
[Embodiment 21] The output level detection process includes a color signal selection process for selecting a predetermined color signal in the plurality of imaging signals,
21. The correction method according to any one of embodiments 18 to 20, wherein the output level detection result is generated based on a color signal selected by the color signal selection process.
[0098]
[Embodiment 22] The correction coefficient determination process includes a gain error calculation process for calculating a gain error between imaging signals from a plurality of detection results, and a threshold setting for setting a threshold value for allowing a level difference between the imaging signals. Processing and the gain error signal calculated in the gain error calculation processing are input, and when the gain error signal exceeds a threshold, a reference value is output, and when the signal does not exceed the threshold, the input gain error is output. With non-linear processing
The correction method according to any one of embodiments 18 to 21, wherein a correction coefficient is determined based on a signal output from the nonlinear processing.
[0099]
[Embodiment 23] The correction coefficient determination process includes an evaluation value generation process for generating an evaluation value from a detection result of the output level detection process, and an average of the evaluation values generated by the evaluation value generation process between a plurality of frames. An averaging process, and a frame number setting process for setting the number of frames to be averaged in the averaging process,
23. The correction method according to any one of embodiments 18 to 22, wherein a correction coefficient is determined based on an evaluation value obtained by averaging the number of frames set by the frame number setting process.
[0100]
[Aspect 24] The correction according to Aspect 23, wherein the averaging process includes a frame number control process for controlling the number of frames to be averaged according to the level of the detection result obtained by the output level detection process. Method.
[0101]
[Embodiment 25] The output level detection process includes an upper and lower limit level setting process for setting an upper limit level and a lower limit level.
In any one of Embodiments 18 to 23, the output level detection result is generated by selecting a signal having a level lower than the upper limit level and higher than the lower limit level among the plurality of imaging signals. The correction method described.
[0102]
[Embodiment 26] A correction method for outputting a plurality of imaging signals,
Based on a plurality of level adjustment processes for independently adjusting the levels of the plurality of imaging signals, an output level detection process for detecting an output level of the plurality of level adjustment processes, and a detection result of the output level detection process A correction coefficient determination process for determining a correction coefficient so as to reduce the level difference between the respective imaging signals, and a correction coefficient storage process for storing the correction coefficient in a recording medium,
An adjustment value determined based on both the correction coefficient stored in the recording medium and the correction coefficient obtained in real time during the imaging operation is given to the level adjustment process so that the level difference between the imaging signals is reduced. Correction method characterized by performing various adjustments.
[0103]
[Aspect 27] The correction method according to Aspect 26, wherein the correction coefficient storage process stores the correction coefficient determined in the correction coefficient determination process at a predetermined timing in the recording medium.
[0104]
[Embodiment 28] In the correction coefficient determination process, an adjustment value to be supplied to the level adjustment process is determined by multiplying a correction coefficient stored in the recording medium by a correction coefficient obtained in real time during an imaging operation. 28. The correction method according to embodiment 26 or 27, wherein:
[0105]
[Embodiment 29] The output level detection process includes an area selection process for selecting a predetermined area in the plurality of imaging signals, and an average level calculation for calculating an average level in the area selected by the area selection process. The correction method according to any one of embodiments 26 to 28, further comprising: a process.
[0106]
[Embodiment 30] The output level detection process includes a color signal selection process for selecting a predetermined color signal in the plurality of imaging signals,
30. The correction method according to any one of embodiments 26 to 29, wherein an output level detection result is generated using the color signal selected by the color signal selection process.
[0107]
[Embodiment 31] The correction coefficient determination process includes a gain error calculation process for calculating a gain error between imaging signals from a plurality of detection results, and a threshold setting for setting a threshold value that allows a level difference between the imaging signals. And a non-linear process that outputs a reference value when the gain error detection signal is input and the gain error exceeds a threshold value, and outputs the input gain error as it is when the signal does not exceed the threshold value. And
31. The correction method according to any one of embodiments 26 to 30, wherein a correction coefficient is determined based on an output signal of the nonlinear processing.
[0108]
[Embodiment 32] The correction coefficient determination process includes an averaging process for averaging output level detection results between a plurality of frames, and a frame number setting process for setting the number of frames to be averaged in the averaging process. Have
32. The correction method according to any one of embodiments 26 to 31, wherein a correction coefficient is determined based on a level obtained by averaging the number of frames set by the frame number setting process.
[0109]
[Aspect 33] The correction according to Aspect 32, wherein the averaging process includes a frame number control process for controlling the number of frames to be averaged according to the level of the detection result obtained by the output level detection process. Method.
[0110]
[Embodiment 34] The output level detection process is based on the upper and lower limit level setting processes for setting an upper limit level and a lower limit level, and the upper limit level and the lower limit level set by the upper and lower limit level setting process. 34. The correction method according to any one of embodiments 26 to 33, wherein a detection result is generated by selecting only a signal whose level is lower than the upper limit level and higher than the lower limit level.
[0111]
[Embodiment 35] A computer program for causing a computer to execute a correction method for correcting a plurality of imaging signals from a plurality of output units of an imaging device,
A level adjustment process for adjusting the levels of the plurality of image pickup signals; and a correction coefficient determination process for determining a correction coefficient for reducing a level difference between the plurality of image pickup signals. And correcting the subject-dependent level difference component to determine the correction coefficient, and applying the determined correction coefficient to the level adjustment process to adjust the level difference of each imaging signal to be small. A computer program characterized by being executed.
[0112]
[Embodiment 36] A computer program for causing a computer to execute a correction method for outputting a plurality of imaging signals,
Based on a plurality of level adjustment processes for independently adjusting the levels of the plurality of imaging signals, an output level detection process for detecting an output level of the plurality of level adjustment processes, and a detection result of the output level detection process A correction coefficient determination process for determining a correction coefficient so as to reduce the level difference between the respective imaging signals, and a correction coefficient storage process for storing the correction coefficient in a recording medium,
An adjustment value determined based on both the correction coefficient stored in the recording medium and the correction coefficient obtained in real time during the imaging operation is given to the level adjustment process so that the level difference between the imaging signals is reduced. A computer program characterized by causing a computer to perform various adjustments.
[0113]
[Embodiment 37] A computer-readable recording medium in which the computer program according to Embodiment 35 or 36 is recorded.
[0114]
【The invention's effect】
  As described above, according to the present invention, when the calculated gain error exceeds the threshold, the correction coefficient is determined based on a predetermined reference value, and when the calculated gain error does not exceed the threshold, the calculated gain error is determined. Since the correction coefficient is determined based on this, it is possible to correct non-uniformity between a plurality of imaging regions in real time in response to dynamic fluctuation such as temperature fluctuation or temporal fluctuation. This makes it possible to correct inhomogeneities between a plurality of imaging regions in real time, and to eliminate the steps that appear in the image satisfactorily.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an embodiment of the present invention and showing a configuration of a first embodiment in which a correction apparatus of the present invention is applied to a video camera.
FIG. 2 is a diagram illustrating a rectangular area at a boundary portion of a divided screen.
FIG. 3 is a characteristic diagram showing a CCD output level and a gain difference between channels.
FIG. 4 is a diagram illustrating a gain correction characteristic with respect to a gain-up amount.
FIG. 5 is a block diagram showing a configuration example of means for executing a gain adjustment value calculation procedure in the first embodiment;
FIG. 6 is a diagram showing input / output characteristics of a limiter in the first embodiment.
FIG. 7 is a block diagram showing a configuration example of integration means in the first embodiment.
FIG. 8 is a block diagram illustrating a configuration example of a unit that performs a gain adjustment value calculation procedure according to the second embodiment;
FIG. 9 is a block diagram illustrating a configuration example of integrating means in the second embodiment.
FIG. 10 is a diagram illustrating coefficient control characteristics in the second embodiment.
FIG. 11 is a block diagram illustrating a configuration example of a step evaluation value generation unit according to the third embodiment.
FIG. 12 is a block diagram showing a configuration example of means for executing a gain adjustment value calculation procedure in the third embodiment;
FIG. 13 is a block diagram illustrating a configuration example of integrating means in the third embodiment.
FIG. 14 is a block diagram illustrating an example of a conventional correction apparatus.
[Explanation of symbols]
100 CCD area sensor
101 photoelectric conversion unit and vertical transfer unit
103, 104 Horizontal transfer section
105, 106 Output amplifier
107, 108 Imaging signal output terminal
109, 110 Analog front end
111, 112 Black level detection and correction means
113, 114 Gain adjusting means
115 Screen composition means
116 Step evaluation value generating means
117 Microcomputer that controls the system
118 Camera signal processing means
119 Output terminal
120 Rewritable non-volatile memory

Claims (11)

A correction device that corrects a plurality of imaging signals from a plurality of output units of an imaging element,
Output level detection means for detecting levels of the plurality of imaging signals;
Level adjusting means for adjusting the levels of the plurality of imaging signals;
Correction coefficient determining means for determining a correction coefficient for reducing a level difference between the plurality of imaging signals,
The correction coefficient determining means includes a gain error calculating means for calculating a gain error between the plurality of imaging signals from the level detection result of the plurality of imaging signals output from the output level detecting means,
When the gain error calculated by the gain error calculation means is input,
When a gain error between the imaging signals exceeds a threshold for allowing a level difference between the plurality of imaging signals, a predetermined reference value is output,
When the gain error does not exceed the threshold, non-linear processing means that outputs the input gain error as it is,
While determining the correction coefficient based on the signal output from the nonlinear processing means,
A correction apparatus characterized in that the correction coefficient determined by the correction coefficient determination means is applied to the level adjustment means so that the level difference between the plurality of image pickup signals is reduced. The correction coefficient determining means includes an evaluation value generating means for generating an evaluation value from the detection result of the output level detecting means,
Evaluation value averaging means for averaging evaluation values generated by the evaluation value generating means between a plurality of frames;
Frame number setting means for setting the number of frames to be averaged in the evaluation value averaging means,
2. The correction apparatus according to claim 1, wherein a correction coefficient is determined based on an evaluation value obtained by averaging the number of frames set by the frame number setting means. 3. The correction apparatus according to claim 2, further comprising: a frame number control unit that controls the number of frames to be averaged by the evaluation value averaging unit in accordance with a level of a detection result by the output level detection unit. The output level detection means includes upper and lower limit level setting means for setting an upper limit level and a lower limit level,
2. The correction device according to claim 1, wherein a detection result of an output level is generated by selecting a signal having a level lower than an upper limit level and a level higher than a lower limit level among the plurality of imaging signals. The correction device includes correction coefficient storage means for storing the correction coefficient in a recording medium,
An adjustment value determined on the basis of both the correction coefficient stored in the recording medium and the correction coefficient obtained in real time during the imaging operation is given to the level adjustment means, and the level difference between the plurality of imaging signals is reduced. The correction apparatus according to claim 1, wherein adjustment is performed so that The correction apparatus according to claim 5, wherein the correction coefficient storage unit stores the correction coefficient determined by the correction coefficient determination unit at a predetermined timing in the recording medium. The correction coefficient determining means determines an adjustment value to be supplied to the level adjusting means by multiplying a correction coefficient stored in the recording medium by a correction coefficient obtained in real time during an imaging operation. The correction device according to claim 5.   An imaging apparatus comprising: the correction apparatus according to claim 1; and the imaging element. A correction method for correcting a plurality of imaging signals from a plurality of output units of an imaging element,
An output level detection step for detecting levels of the plurality of imaging signals;
A level adjustment step for adjusting the levels of the plurality of imaging signals;
A correction coefficient determining step for determining a correction coefficient for reducing a level difference between the plurality of imaging signals,
The correction coefficient determining step calculates a gain error between the plurality of imaging signals from a level detection result of the plurality of imaging signals output in the output level detection step;
When the gain error calculated in the gain error calculation step is input,
When a gain error between the imaging signals exceeds a threshold for allowing a level difference between the plurality of imaging signals, a predetermined reference value is output,
When the gain error does not exceed the threshold, a non-linear processing step that outputs the input gain error as it is,
Determining the correction factor based on the signal output in the nonlinear processing step;
A correction method comprising: adjusting the determined correction coefficient to the level adjustment step so as to reduce a level difference between the plurality of image pickup signals.   A computer-readable storage medium storing a program for causing a computer to execute each step of the correction method according to claim 9.   A program causing a computer to execute each step of the correction method according to claim 9.

Images