JPH03173288A - Solid-state image pickup device - Google Patents

Abstract

PURPOSE:To correct distortion in the subscanning direction at readout of a solid-state image pickup element by reading a picture for each scanning line in the subscanning direction orthogonal to a main scanning direction and varying a drive pulse frequency of the solid-state image pickup element is this case. CONSTITUTION:A microprocessor 4 calculates the correction quantity of each longitudinal line based on a correction data to apply correction control to the timing of a drive pulse of a CCD drive pulse generating circuit 5. A picture signal whose longitudinal picture distortion is corrected is read out of a CCD3G with horizontal and vertical CCD drive pulses generated from the circuit 5. The time interval of the signal read by the pulse train for vertical CCD drive is not constant. Then the signal is converted into a consecutive signal by a sample-and-hold circuit 6 and an LPF 7. The picture signal read in the longitudinal direction should be separated into the number of scanning lines, a-sets in the stage of the conversion into a television signal finally, then the signal is converted into a discrete signal being a pulse signal whose time interval is equal by the sample-and-hold circuit 8. The signal is converted into a digital signal at an A/D converter 9 and stored sequentially in a frame memory 10.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a solid-state imaging device such as a color television camera that uses a plurality of solid-state imaging devices such as CCD as an imaging device.
In particular, it relates to electrical correction of overlay errors between images of each color in a plurality of solid-state image sensors and geometric distortion of images of each color caused by imaging optical systems and the like.

(Prior Art) In general, a color television camera captures an image of a subject by separating it into the three primary colors of red, green, and blue, and reproduces a color image by superimposing the images of each color. If chromatic aberration of magnification remains in the image pickup lens, the sizes of images of each color will not match, resulting in color shift in the reproduced color image.Furthermore, if distortion aberration remains in the imaging lens,
Geometric distortions such as a barrel shape or a pincushion shape occur in images of each color.

Conventionally, in an image pickup tube type color television camera, correction of the overlay error between images of each color and geometric distortion has been performed by controlling an electron beam deflection system for reading signal charges. However, in the case of solid-state imaging device cameras that have recently become widely used, correction can only be made by controlling the deflection system.

Therefore, as a method for electrically performing correction, a method for performing image correction by controlling the frequency of the drive pulse of the solid-state image sensor is disclosed in Japanese Patent Laid-Open No. 58-59690, which performs interpolation processing when reading signals from the solid-state image sensor. JP-A No. 61-89791 discloses an image correction method that performs image correction using a digital signal processing technique, and JP-A No. 62-230267 discloses an image correction method that manually inputs an image signal into a frame memory and performs image correction using digital signal processing technology. has been done.

(Problems to be Solved by the Invention) However, the above-mentioned conventional examples compatible with solid-state image sensor cameras each have the following problems.

First, in the method of controlling the frequency of the driving pulse of the solid-state image sensor, correction in the main scanning direction (generally horizontal direction) of the image is achieved by changing the driving frequency.
The superposition drive frequency cannot be changed from the value determined by the number of frames per second and the number of running lines in one frame of the TV system being used (15.72 to 11Z for the NTSC system). Correction of an image in the sub-scanning direction (generally vertical direction) is impossible in principle. Also,
Solid-state image sensor? '-The L method, which performs interpolation processing when reading out such a signal, and the method, which manually inputs the image signal into a frame memory and performs No. 1 processing on the digital IE, both require complex interpolation in order to perform high-precision interpolation iE. IK detection requires interpolation, and is not suitable for devices such as television cameras that are required to process images almost in real time.

The present invention has been made in view of the problems of the prior art as described above, and it is possible to eliminate alignment errors between images of each color and geometric distortion of images of each color, which occur in a multi-plate solid-state image sensor color imaging device, with high precision. An object of the present invention is to provide a solid-state imaging device that can perform corrections.

(Means for Solving the Problems) In order to achieve the above object, the present invention configures a solid-state imaging device as shown in (1) and (2) below.

(1) In a solid-state imaging device for forming an image consisting of a plurality of scanning lines in the main scanning direction, the image is read out for each scanning line in the sub-scanning direction orthogonal to the main scanning direction, and
A solid-state imaging device that corrects distortion in the sub-scanning direction when reading out the solid-state imaging device by making the drive pulse frequency of the solid-state imaging device variable when performing this reading.

(2) A solid-state image sensor capable of reading out signals for each line in the sub-scanning direction, and a readout frequency in the sub-scanning direction from the solid-state image sensor f that is varied according to predetermined correction data. A correction circuit for correcting distortion, a frame memory for storing the output of this correction circuit, and a distortion in the main scanning direction by varying the reading frequency in the -F scanning direction from this frame memory according to predetermined correction data. and a correction circuit that corrects 11 in the solid-state image sensor.
A solid-state imaging device that corrects image signals independently in the main scanning direction and sub-scanning direction.

(Action) + Wf (1), (2) ノM seini, k'),
Alignment error between images of each color in the body image sensor f and geometric distortion of images of each color can be corrected with high precision.

(Example) Hereinafter, the present invention will be explained in detail by way of example.

FIG. 1 shows the present invention in a three-plate CCD (chargeco
1 is a block diagram of a first embodiment of the present invention applied to a color television camera (uploaded device); FIG. The light emitted from the object passes through the imaging lens 1 and is separated by the color separation prism 2 into three primary colors: red light R1, green light G, and blue light B. The optical image formed by these three primary color lights R, G, and B is , respectively C
Images are formed on CDs 3R, 3G, and 3B (solid-state image sensors-F).

The imaging lens 1 generally has residual chromatic aberration of magnification and distortion, and the alignment errors between the RG and B color images caused by these aberrations, as well as the nucleus of the geometric distortion of each image, are shown in Figure 2. I will explain using The figure shows images formed on the CCDs 3R, 3G, and 3B when an achromatic rectangular object 50 is imaged. 51, 52, 53 are 3R and 3G respectively
.

3BF images, and as shown in Figure (b), they are distorted in size due to lens distortion and do not match in size due to lateral chromatic aberration.

The correction method for the signals from the CCD 3R13G and 3B is the same, although the correction method is different, so the correction method will be explained here by taking the signal from the CCD 3G as an example.

Signal readout methods from CCDs include interline transfer methods, frame transfer methods, etc. Regarding interline transfer methods, the differences between a general TV camera CCD and the CCD of this example are shown in Figure 3. explain. Third
Figure (a) shows the signal readout method of a general CCD for a television camera, and Figure 3 (b) shows the signal readout method of the CCD of this embodiment. In FIG. 3(a), the charge created by the photodiode 60 during the vertical blanking period! [Transferred directly to CCD 61, and then every time one scanning line advances! [Direct CCD transfers one clock at a time, water, F,
A signal charge for one scanning line is sent to the CCD 62. Then, the signal 'I' sent to the horizontal CCD is sequentially transferred to the signal detection unit 63 in accordance with the speed of the horizontal scanning. Since it is not possible to read out only the main signal charge, it is impossible to perform any vertical correction of the image.Therefore, in order to make it possible to perform vertical image correction by controlling the IC direct CCD drive pulse frequency. , in this example, Fig. 3 (b
) is used. CC in Figure 3(b)
In D, the roles of the city direct CCD and the horizontal CCD are changed compared to the CCD shown in FIG. 3(a). That is, the charge created by the photodiode 64 is first transferred to the horizontal CCD 6.
The signals of one line in the vertical direction are transferred at a time and read out by the signal detecting section 67 through the vertical CCD 66.

Next, a method of correcting the image in the vertical direction, which is performed at the stage of reading out the vertical signals from the CCD as described above, will be described. Consider the example shown in FIG. 4 as an image to be corrected. In other words, an image that is distorted as shown by the dotted line 70 due to the lens distortion aberration is eventually corrected as shown by the dashed-dotted line 71, but first vertical correction is performed and the image is distorted as shown by the solid line 72.
Consider doing something like this.

Figure 5 shows one line in the vertical direction of the CCD.
For simplicity, it is assumed that the number of pixels in the vertical direction is equal to the number of scanning lines a. Consider expanding an image compressed to a number of pixels (b < a) indicated by diagonal lines in the figure to a number of pixels. Su
, SR indicate the topmost pixel and the topmost pixel of the image to be corrected, respectively.

From here, we will begin to explain the JL type correction means.

The signal for each line in the vertical direction as shown in FIG. 5 is transferred to the vertical CCD 66 (see FIG. 3(b)), from which it is read out as a signal for each pixel by a vertical CCD drive pulse. In Figure 6(a) ai! ! A pulse train is shown when the signal of ji components is read out without being corrected. The number of pulses is a%, and the time required to generate a number of pulses is τ. Next, as described above, a pulse train for expanding an image compressed into b pixels into alIhI elements will be explained. In Figure 5, are b pulses required to read out the signals for b pixels from Su to Sl?If the method is to read out the signals on the vertical CCD from the talker, then the first pulse will be at the Sl position. , if the method is to read from above, it must correspond to the position of Su. Furthermore, since the time required to read out the signal of one line in the vertical direction must be the same in all correction states, the time required to generate b pulses must be equal to the time τ in the non-correction state. Furthermore, when the image is distorted due to inter-form distortion aberration as in this example, the distortion is not uniform, and the periphery of both images is strongly compressed, so the periphery of the image is correspondingly greatly compressed. It must be stretched, that is, the pulse interval must be lengthened. From the above, the pulse train necessary for corrected readout is as shown in FIG. 6(b).

Further, here, since no horizontal correction of the image is performed, the horizontal CCD 65 (see FIG. 3(b)) is driven with normal equally spaced pulses.

In order to perform the above-mentioned correction on the entire screen, it is necessary to keep the pulse train width τ constant for each line in the vertical direction and change the number of pulses and the arrangement of the pulses in the pulse train.

The microprocessor 4 in FIG. 1 calculates the amount of correction for each line in the vertical direction based on the correction data, and controls the timing of the drive pulse of the CCD drive pulse generation circuit 5 to correct it.

COD 3 is generated by the horizontal CCD drive pulse and vertical CCD drive pulse generated from the CCD drive pulse generation circuit 5.
As described above, an image signal whose vertical image distortion has been corrected is read from G. Here, the correction data is prepared in advance.
It may be stored in ROM, or it may be stored in RAM from outside. You may also use one with F.

As shown in FIG. 7(a), the signals read out by the pulse train for driving the vertical CCD described above are not constant at intervals of 1:11. Therefore, first, this signal is sampled and held by the sample and hold circuit 6. A continuous signal as shown in FIG. 7(b) is generated by a low-pass filter. The image signal read out in the vertical direction must be separated into a pieces corresponding to the number of scanning lines at the stage of final conversion into a television signal, so the sampling circuit 8 divides the image signal into a number of parts corresponding to the number of scanning lines, as shown in FIG. 7(e). An m-dispersed signal, which is a pulse signal at equal intervals, is generated. The signal is converted into a digital signal by an A/D conversion circuit 9 and sequentially stored in a frame memory 10. When the above operations have been completed for all vertical lines on the screen, the frame memory 10-
)- has an image corresponding to the solid line in FIG. 4 72 stored therein.

Next, a method of performing horizontal correction on an image that has been vertically corrected or completed using the above-described method will be described. The explanation will be given using FIG. A solid line 80 is an image to be corrected in the horizontal direction, a dashed-dot line 81 is an image to be moved forward after the correction is completed, and a diagonal line 82 is one line in the horizontal direction (corresponding to one scanning line) for explaining the correction method.

Now, assuming that the total number of pixels in the horizontal direction is p, and the number of pixels included in the diagonal area 82 is q (q<p), to enlarge an image of q pixels to ρ pixels, use the same method as in FIG. 6(b). The signal can be induced using a pulse train. However, in this case, the number of pulses is nine, and the time corresponding to τ is automatically determined when the television system is determined. Historically, in this case, since the image correction has been completed in the vertical direction, the signal for each line is read out using equally spaced pulses, and the readout frequency is also automatically determined based on the television system. Is it necessary to change the pulse train for signal readout in accordance with the correction data in the horizontal direction correction as well as in the vertical direction?The total correction amount p and generation of the corresponding pulse train are determined by the microprocessor 4. This is performed using the No. 15 read pulse generation circuit 11.

As mentioned above, the signals read out for each horizontal line from the frame memory 10 are converted into analog signals by the D/A conversion circuit 12, and then converted to analog signals by the sample and hold circuits 13. It passes through a low-pass filter 14 and becomes a continuous signal. The signal is exactly equivalent to the signal obtained by scanning an image sensor with a normal television camera, and the image distortion has been corrected.

Up to this point, the image correction method for the signal from CC03G has been explained. For images formed on CCD 3R and 3B, image distortion is corrected in the same way as 3G images,
Also, the size deviation of the images on 3R, 3G, and 3B due to the lateral chromatic aberration of the lens is, for example, based on the image of 3G, 3R, 3G, and 3B.
It is sufficient to control the CCD drive pulse and the signal readout pulse from the frame memory so that the sizes of the 3B images match.

Below 1. CCD 3R obtained as follows.

No. 18 from 3G and 3B is either converted to NTSC TV (+"j ""j) and displayed on a monitor, or the image is a very becomes beautiful.

Although this embodiment adopted a method of performing correction by stretching the image, the present invention is not limited to this, and a method of performing correction by compressing the image may also be adopted. For example, CCD vertical direction C pixels (c>a
) when you want to compress the image into a pixel image,
It is sufficient that the pulse train of the vertical CCD driving pulses has a large time width, as in FIG. 6(b), and is composed of a total of C pulses having intervals corresponding to the way the image is distorted.

Although this embodiment is an example of a 3-panel CCD camera, the present invention is also applicable to color cameras other than 3-panel cameras such as 2-panel and 4-panel cameras, single-panel monochrome television cameras, and solid-state imaging devices other than CCDs such as Mo5y elements as image sensors. Needless to say, this method is applicable to the multi-disc color camera used.

Furthermore, it goes without saying that the present invention can also be used to correct geometric distortion caused by lens distortion in the present invention, a nine-plate color television camera, and a monochrome television camera.

Further, color separation is not limited to a prism, and any suitable means such as a combination of a half mirror and a single color filter can be used.

Next, a second embodiment of the present invention will be described in which the present invention is applied to a color television camera with a zoom lens. In a zoom lens, when the usage conditions (focus position, focal length, aperture value) change due to zooming, etc., the chromatic aberration of magnification and distortion change accordingly. Therefore, the ccDg motion pulse control for correcting the image and the signal readout pulse control from the frame memory must also be changed depending on the usage state of the lens. This embodiment makes it possible to always perform optimal image correction by following the fluctuations of the Δ-rate chromatic aberration and distortion in real time.

The correction method itself is the same as the first embodiment, but in the second embodiment, the optimum correction amount is calculated from the signal representing the lens condition,
The CCD drive pulse and the signal read pulse from the frame memory can be controlled accordingly.

This embodiment will be explained using FIG. 9. Figure 9 shows only the lens portion that has the ability to output correction data.
It is assumed that the correction data output from the lens side is input to the microprocessor 4 shown in FIG. In FIG. 9, 90 is the zoom lens body, 91, 92, and 93 are potentiometers built into the zoom lens 90. Potentiometer 91 is the focus position, potentiometer 92 is the zoom ratio, and potentiometer 93 is the aperture value data. is provided for reading out. 94.9
596 respectively 1 said potentiometer 91, 92°9
3 is an A/D conversion circuit that converts the data from 3 into a digital signal, and 97 is a ROM that stores information on magnification chromatic aberration and distortion aberration in several usage states (focus position, focal length, aperture value) of the zoom lens 90. , 98 is a microprocessor that calculates and outputs appropriate correction data based on data from the potentiometers 91, 92, and 93 and information in the ROM.

In FIG. 9, signals output from potentiometers 91, 92, and 93 are respectively output from A/D conversion circuits 94 and 9.
5.96, it is manually powered by the microprocessor 98. The microprocessor 98 reads out the chromatic aberration of magnification and distortion information from the ROM 97 according to the input data, performs interpolation calculations if necessary, and outputs appropriate correction data according to the state of use of the lens. The supplementary iE data is input manually to the microprocessor 4 in FIG. 1, and image correction for each color of R2O and B is performed in the same manner as in the previous embodiment. As soon as the usage status of the lens changes, correction data appropriate for the new usage status is calculated, and image correction is performed based on the data, so that beautiful television images are always obtained.

In addition, in the embodiment of multiplexing, the -iE scanning direction is
Although one P direction and the sub-scanning direction are vertical, the reverse direction may be used. Alternatively, the distortion in the horizontal direction may be corrected first, and then the distortion in the vertical direction may be corrected, and the negative images may be outputted in one line in the vertical direction and then supplied to a printer or the like.

In addition, when printing out using a printer, it is preferable to print by scanning the print heads in the vertical direction in the horizontal direction, in terms of the dot density of the print head, so image data in the vertical direction is Output one line at a time is preferred. In that case, the solid-state image sensor shown in 3rd IA(a) can be used.

〔Effect of the invention〕

As explained below, according to the present invention, it is possible to increase the overlay error between images of each color in a plurality of solid-state image sensors and the image distortion of each color of each solid-state image sensor, which is caused by lateral chromatic aberration and distortion of the lens. You can obtain beautiful images with accurate correction.

Moreover, since there is no need to perform complicated interpolation calculations between pixels during the correction process, the circuit scale of the correction device can be relatively small.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of the first embodiment of the present invention, Fig. 2 is a diagram illustrating the effects of lateral chromatic aberration and distortion of the lens on images;
Figure 3 is a diagram explaining the signal readout method from the CCD, Figure 4 is a diagram explaining the order of image correction, and Figure 5 is a diagram explaining the signal readout method from the CCD.
6 is a diagram showing a pulse train for reading signals from the CCD, and FIG. 7 is an enlarged view of one line in the vertical direction of , FIG. 8 is a diagram illustrating horizontal image correction, and FIG. 9 is a configuration diagram of a second embodiment of the present invention. 1...Imaging lens 2...3 color separation prism 3R, 3G, 3B-...C0D 4...Microprocessor 5...-CCD D movement Pulse generation circuit 6...
- Sample hold circuit 7...Low pass filter 8...Sampling circuit 9...-A/D conversion circuit 10...Frame memory 11...1. No. 1 read pulse generation circuit 12.
...-D/A conversion circuit 13... Sample hold circuit 14...
・Low pass filter

Claims (2)

[Claims] (1) In a solid-state imaging device for forming an image consisting of a plurality of scanning lines in the main scanning direction, the image is read out for each scanning line in the sub-scanning direction orthogonal to the main scanning direction, and
A solid-state imaging device characterized in that distortion in the sub-scanning direction is corrected when reading out the solid-state imaging device by making variable the drive pulse frequency of the solid-state imaging device when performing this reading. (2) A solid-state image sensor capable of reading signals for each line in the sub-scanning direction, and distortion in the sub-scanning direction by varying the readout frequency in the sub-scanning direction from this solid-state image sensor according to predetermined correction data. a correction circuit that corrects the distortion, a frame memory that stores the output of this correction circuit, and a readout frequency in the main scanning direction from the frame memory that is varied in accordance with predetermined correction data to correct distortion in the main scanning direction. What is claimed is: 1. A solid-state imaging device comprising: a correction circuit; the solid-state imaging device comprising: a correction circuit; the solid-state imaging device corrects an image signal obtained by the solid-state imaging device independently in a main scanning direction and a sub-scanning direction;