Solid-State Color TV Camera Image Size Control

This invention relates to control of registration of a color television camera including more than one imager, particularly when a zoom lens is used.

Color television cameras have traditionally used several electron-beam camera tubes optically coupled to a color-splitting optical system to produce from each tube a signal representative of a color component of the light from a scene being imaged. The color-splitting optical system most often was a prism assembly with dichroic surfaces by which the light separation was accomplished. Because of the need for periodic replacement of the camera tubes, the optical coupling of the tubes to the prism is made by a complex mechanical assembly which can be disassembled for tube replacement. However, the camera tubes in order to produce color signals from which a correct image can be formed are required to scan the image simultaneously; i.e. the electron beam of each tube is required to scan that portion of its target upon which the portion of the image is focussed which is at that moment being scanned by the other tubes. Since the focussed image is generally less than 3 cm on a side and often less than 1 cm on a side, the slightest movement of the tube relative to the optical system can cause the scannings to become offset. Since each tube responds to a different color, this causes the image of one color to be offset with respect to the other colors upon subsequent reproduction of the scene. Such an offset is termed misregistration, and is avoided in electron-beam tube television cameras by a combination of mechanical integrity and stable electron-beam deflection or scanning. The requisite mechanical integrity makes it difficult to design a camera mount which allows the camera tube to be moved relative to the optics, so these are ordinarily fixed. The need for adjustment to cause the beam to scan over the proper portion of the image is met by controllable electron-beam deflection circuits. These circuits produce analog deflection signals such as sawtooth current waveforms. Because of their analog nature, the waveform may be adjusted in almost infinitesimally small amounts so as to obtain proper color registration.

The power of optical systems varies with the color of the light being processed. Wide-range zoom lenses are particularly susceptible to differential changes in power as the focal length is changed. The change in power causes the sizes of the different-color images focussed onto the targets of the camera tubes to differ slightly. This causes a misregistration of the camera, which is corrected by mechanically coupling a potentiometer or variable resistor to a movable portion of the optical system of the zoom lens to produce a control signal which varies in response to the focal length and therefore in response to the difference in image size. The control signal is coupled to the deflection circuits of the tubes for varying the deflection so that the scanning of each tube over its image occurs simultaneously with the scans of the other tube or tubes of the assemblage, so that misregistration errors are reduced or eliminated.

Camera tubes are bulky, heavy, fragile and consume substantial power during operation. Attention has therefore been directed to solid-state imager arrays, such as MOS or CCD imagers. These imagers include a plurality of individual light-responsive sites each of which generates signals representative of the total amount of light falling onto the site (also called an element or pixel). The discrete sensing sites cause resolution of the CCD imager to be limited by comparison with that of a modern camera tube, but the advantages of long life, light weight, ruggedness, signal-to-noise and low power consumption make the solid-state imager very attractive for color camera use.

The discrete sites of the solid-state imagers allow the scanning to be accomplished by addressing each site individually. In CCD imager arrays, a tripartite structure is used to accomplish the addressing. The first or "A" portion ("A" register) includes a rectangular array of sensor sites upon which light from the image is focussed. The array is formed of vertical columns separated from each other by stop channels, and within each column the discrete sensor sites are established by gate voltages. The gate voltages may be varied to move the signal generated at a site along the vertical column. The imager array also includes as a second part a "B" or buffer register generally corresponding in size with the "A" register, but isolated from light. Each column of the B register is coupled to an extension of a corresponding column of the A register. In operation, light is allowed to fall onto the A register for a predetermined period of time to form a signal at each site. A clock is then operated during a "pull-down" interval to move the signal of each sensor site vertically along the channel of a column to a corresponding site in the "B" register for temporary storage, thereby allowing the A register to again begin generating signal in response to light for a second or integrating interval, generally corresponding to the active portion of a television vertical interval, i.e. that portion of the vertical interval which is not vertical blanking inteval. During the active portion of the vertical interval, the signal stored in the B register is clocked into a parallel-to-serial "C" register and read out in serial form by a high-speed clock. The time-quantized analog-amplitude (sampled analog) signal derived from the C register is filtered to produce line-sequential analog tellevision signals which may be processed in a manner similar to those of a camera tube.

Because of the very long expected life of solid-state imagers, there is no need to mount them to the optical system in a temporary manner, and they may therefore be permanently bonded to the prism as described in U.S. Pat. No. 4,323,918 issued Apr. 6, 1982 in the name of S. L. Bendell. As there described, the registration is determined at the time the imagers are bonded into position. While this is effective, spacers may be necessary to position the various imagers at the proper focal point for the color in question, and the sizes of the images may not be the same at that focal point. Since the size of the imager is established by a photographic process which cannot easily be varied, all solid-state imagers of a given type are of about the same size. There being no convenient way to adjust the size of either the imager or of the image, there may be some residual misregistration. This problem is exacerbated when the optical system has a variable power as when a zoom lens having a variable focal length is varied. With this variation, the sizes of the color images vary with respect to each other. The largest size of the optical image is normally made to be greater than the physical size of the active portion of the imager onto which it is focussed, so that the size variation of the images does not cause edge portions of the image to disappear, but misregistration may occur or change by as much as several pixels. It is desirable to be able to adjust for misregistration in both a fixed and variable focal-length situation when solid-state imagers are used.

A color misregistration correction arrangement for a television camera includes an optical system having differential power causing images of different size for different colors. A first imager coupled to the optical system responds to a first color and includes an array of light-sensing pixels. A second imager is coupled to the optical system and responds to a different color. A clock is coupled to the imagers for sequentially reading signals from the first array within a reference time interval, and is controlled for reading signals from the corresponding array of the second imager in a time selected for reducing misregistration. In a particular embodiment, the time for reading the array of the second imager is varied in response to variations in the focal length of the optical system.

FIGS. 1a and 1b illustrate, partially in block-diagram form, a camera including a prism, an optical system with a four-port color-splitting prism and four discrete-sensor solid-state imagers;

FIGS. 2a, 2b, and 2c illustrate raster patterns aiding in understanding how changes in the power of the optical system can cause misregistration;

FIG. 3 illustrates rasters of a camera and a monitor with timing information aiding in understanding how changing the clock timing can change the apparent size of the image falling on an imager as seen or a monitor;

FIG. 4 illustrates, partially in simplified block-diagram form, a camera arrangement according to an aspect of the invention having a fixed optical system;

FIGS. 5a and 5b illustrate, partially in simplified block-diagram form, a camera arrangement with a variable optical system according to another aspect of the invention;

FIG. 6 illustrates details of the arrangement of FIG. 5; and

FIG. 7 is a timing diagram aiding in understanding the operation of the arrangements of FIGS. 5 and 6.

In FIG. 1, a television camera includes a zoom lens 10 with a manual control ring 12 which lens focusses a scene (not shown) onto the surfaces of a blue-responsive solid-state imager 14, a red-responsive imager 16 and two green-responsive imagers 18 and 20 by way of a prism 22 including dichroic surfaces and a 50% neutral reflecting surface 24 such as a half-silvered surface.

Each solid-state imager may be of the CCD type illustrated in FIG. 1b having an "A" register 90 onto which light is focussed and in the surfaces of which photoelectrons are generated. The photoelectrons are constrained within vertical channels by channel stops 94. Vertical motion of the accumulated image-representative charge is controlled by polyphase clock voltages applied to control electrodes illustrated as 96. After an integrating period, the clock signals are activated to produce a peristaltic motion of the electrons into corresponding portions of a "B" register 98 which is free from light influence. The charge in each packet is clocked in parallel from B register 98 to C-register 99, from which they are clocked in serial fashion by clock signals applied to clock electrodes 97. The signals are clocked to the right or to the left, depending on the phase relation of the polyphase C clock drive, to left or right output couplers or amplifiers 93 or 95, respectively.

The output signals from imagers 14, 16, 18 and 20 are clocked to preamplifiers 26, 28, 30 and 32, respectively, under control of clock signals applied from a clock signal generator 34. The signal from clock generator 34 applied to a sync signal generator 36 for generating the various sync signals, and is also applied to the imagers. The clock signals applied to the number-2 green imager 20 are phase-inverted with respect to those applied to number-one green imager 18 by a phase inverter 38 so that the green imagers are clocked alternately to produce an effective improvement in resolution when the imagers are respectively physically displaced by one-half pixel.

The blue and red signals amplified by preamps 26 and 28 are applied to signal processors 40 and 42, respectively, of standard types which may include clamps, shading and gamma corrections and other types of signal processing. The alternately-occurring green signals amplified by amplifiers 30 and 32 are applied to a multiplex circuit 44 which because of the alternate clocking of imagers 18 and 20 alternately receives the output signals from the green imagers. The multiplexed green signals are applied to a standard green signal processing circuit illustrated as a block 46. As is often the case, processor 46 is coupled to a green contour generator 48 for generating an edge signal which is added back to the green. The processed signals are applied to an encoder 50 in which the RGB signals are matrixed to YIQ, and the color-representative components are quadrature-modulated onto a subcarrier to form a composite signal such as NTSC or PAL.

The problem to which the invention is directed is that of color misregistration resulting from differences in the power of the optical portions of the camera, especially the lens, as a function of color. This is sometimes known as lateral chromatic aberration. The difference in lens power results in focus of the image at different distances from the lens, and in addition may result in the focussed images having different sizes. FIG. 2 aids in understanding the problem. In FIG. 2a, a focussed image 210 of a series of vertical stripes 212-220 such as a picket fence is illustrated for reference displaced from the surface 222 of the imager with which it is actually congruent. This may represent, for example, the red image focussed onto the red imager. FIG. 2b illustrates the effect of a difference in power of the optical system for some other color, such as blue, for which the power of the lens is greater. While the surface 224 of the blue imager is the same size as that of the red imager as illustrated in FIG. 2b, the focussed blue image 226 (again illustrated displaced from the imager for ease of understanding) is smaller than in the case of the red image. In actual use, the imager will be smaller than the smallest image focussed thereupon so that no "black areas" occur at the sides of any of the images. With the reduced-size image 226 focussed onto imager 224, it is easy to see that the stripes corresponding to 228-236 are physically closer together than the corresponding stripes 212-220 in FIG. 2a. If the stripes are white, so that both the illustrated imagers respond to the light, the stripe 232 will occur in the center of imager 224 and stripe 216 also occurs in the center of imager 222 (assuming that the imagers were initially set up properly). However, the stripes on the sides of the image are relatively displaced. The image produced on a color monitor from composite signals derived from the two imagers illustrated in FIG. 2 is illustrated in FIG. 2c. Since these stripes are of different colors, the center of the raster is registered (216 overlies 232) while the sides are misregistered. As mentioned, the misregistration changes with changes in the focal length or power of lens 10.

FIG. 3 illustrates the principle by which the invention operates. Generally, a monitor which responds to a camera scans horizontally across a raster 312 in a fixed interval, as for example 53.5 .mu.S for NTSC (63.5 .mu.S line interval minus 10 .mu.S blanking), and the camera similarly scans its raster 314 in a like time, to reproduce vertical stripes 316-324 in proportional positions, as illustrated in FIG. 3a for reference. FIG. 3b illustrates the result when the camera scans more slowly but the monitor continues to scan at the original rate. As can be seen, the vertical stripes appear more widely separated on the monitor. The slow scanning of the raster results in loss of some image portions 326-328 at the right and left edges of the image reproduced on the monitor. Naturally, faster-than-normal scanning of the raster by the camera results in apparent compression of the distance between the stripes.

In accordance with the invention, the clocking rate of one or more of the imagers of a color television camera is relatively changed to effectively superpose the images formed from the signals. By this means, the effective relative positioning of the imagers can be adjusted after they are mechanically fixed, and/or misregistration attributable to color aberration may be compensated.

FIG. 4 illustrates in block-diagram form a color television camera in which the optical system is fixed and in which it is desired to compensate for misregistration attributable to relative physical misalignment of the imagers. In FIG. 4, a lens 410 focusses an image (not shown) onto B, R and G imagers 414, 416 and 418 by way of a prism 422. The signal produced by each imager is applied to a corresponding preamplifier 426, 428, 432 for amplification before being applied to further conventional signal processing (not shown). A clock signal generator 434 generates a reference clock signal which is applied to green imager 418. The standard clock frequency is established by assuming that each imager includes 404 active pixels to be scanned during a 53 .mu.S active horizontal interval

the standard clock signal from generator 434 is also applied to sync signal generators and to pull-down clock signal generators illustrated together as a block 450 for control of the frequency thereof.

The blue and red imagers 414 and 416 receive C register clock signals from variable-frequency clock signal generators 436 and 438, respectively. Generator 436 is similar to generator 438. Generator 438 generates stable signals for driving the register of the imager by means of a crystal-controlled oscillator arranged with a variable voltage-controlled capacitor for variation over a relatively small frequency range near a selected operating point of the crystal, as is known. For example, the range of frequency variation of oscillator 438 may be selectable within 1% of 7.622641 MHz, based upon the assumption that the physical location of the imagers on the prism and differential power of the lens together result in an initial misregistration of not more than four pixels out of 404, or about 1%. Thus, it is necessary to produce a clock signal having a stable frequency between 7.546414 MHz and 7.698867 MHz, or .+-.76 KHz from the 7.622641 MHz reference frequency. Potentiometer 446 is arranged with its wiper to pick off the voltage required for the desired apparent size of the red image as viewed on a monitor. The equivalent BLUE SIZE control for clock generator 436 is used to adjust the apparent size of the blue image, so that both the red and blue images are superimposed on the green image and registration results.

FIG. 5 illustrates partially in schematic and partially in block form a camera having a zoom lens coupled for control of registration. In FIG. 5a, a zoom lens 510 includes a control ring 512 for mechanically controlling the power of the optical portions of the lens. Lens 510 is optically coupled by a prism 522 to imagers 514-518, the signals from which are coupled by preamplifiers 526-532 to suitable filters, signal processing and encoding circuits illustrated together as block 534. Encoded composite NTSC signals are produced, as well as R, G and B baseband signals. A variable-frequency clock signal generator 576 is controlled by a direct voltage from a buffer amplifier 572, which receives as an input voltage the output of a variable voltage divider including a fixed resistor 570 and switch-selected variable resistors 560-568. A switch designated generally as 550 includes a wiper 552 mechanically coupled to zoom control 512 for coupling resistor 570 variously to grounded resistors 560-568 as the wiper sequentially encounters contacts 553-557 as control 512 is varied. Thus, stepwise-continuous voltages may be selected for control of the "C" register clock signal for control of the apparent size of blue imager 522 as zoom lens 510 is varied in power. The green imager receives a fixed clock signal and its image remains constant in apparent size. Switch 550 may be ganged with a further switch 580 coupled to a resistor bank 582 corresponding to resistors 560-568 for controlling a drive circuit and VCO illustrated together as a block 584. The clock signal produced by VCO 584 controls the "C" register of red imager 516. The arrangement of FIG. 5a may not be satisfactory for some critical uses, changes in the power of the zoom lens for the different colors may occur as a function of zoom angles. The arrangement of FIG. 5a changes the clock rate, but these changes have least effect at the left side of the raster and greatest effect at the right end, thus causing a second-order registration error resulting from centering error of the red and blue images.

FIG. 5b is a block diagram of the arrangement of FIG. 5a additionally including centering control blocks 590, 592 for centering adjustment of the red and blue rasters displayed on a monitor relative to the green raster for correction of residual centering error resulting from effective size adjustment. Centering gating controls 590 and 592 are mechanically coupled to zoom control 512, as suggested by the dashed line, for receiving an indication of the zoom position. FIG. 6 illustrates details of block 590, and block 592 is similar. In FIG. 6, the variable clock signal is gated through AND gate 632 by a "C enable" signal generated on conductor 634 by flip-flop (FF) 630. A D FF 626 periodically sets its Q output low at the beginning (T.sub.o) of each horizontal (H) drive pulse, thereby disabling transistor 636 and allowing capacitor 624 to begin charging towards the reset (R) threshold voltage of FF 626 through a resistor 612-620 selected by a switch 610 mechanically coupled to control 512. At a time T.sub.3 established by the time constant of resistors 612-622 and capacitor 624, FF 626 is reset, causing a positive-going transition at its Q output. The positive-going Q signal from FF 626 resets FF 630, and enables clock pulse counter 628, which counts 404 clock pulses (the number of pixels in a line of the imager) to again set FF 630. AND 632 is enabled during the count interval of counter 628.

FIG. 7 illustrates the timing sequence which allows apparent centering of the imager for correction of registration. FIGS. 7a-7f are for reference, and FIGS. 7g-7l illustrate the changes that occur when the zoom is varied, thereby varying the clock rate for registration correction and also varying the resistance in the charge path of capacitor 624. In FIG. 7, the horizontal drive pulse is illustrated as having a duration T.sub.o -T.sub.1. The Q output of FF626 goes low at time T.sub.0, as illustrated in FIG. 7c, thereby disabling transistor 636 and allowing capacitor 624 to charge, as illustrated in FIG. 7d. Capacitor 624 reaches the threshold voltage at the R input of FF626 at time T3, causing its Q output to go high, enabling counter 628 and resetting FF 630. The high Q output of FF 630 illustrated in FIG. 7f enables gate 632 beginning at time T.sub.3, and gate 632 passes VCO clock pulses to its imager. Counter 628 counts 404 pulses and produces at time T.sub.6 a high Q output signal as illustrated in FIG. 7e for setting FF630, the Q output of which goes low to disable AND 632 and terminate clocking of the C register for the horizontal line in question.

By comparison, FIGS. 7g-7l illustrate the changes in gate timing which occur when the zoom control is operated to change the lens power in such a manner that the clock rate for one of the imagers must be increased. With the increase in clock rate established by a set of switched resistors as described in conjunction with FIG. 5, the resistor bank 612-620 is set up for a lower charging rate of capacitor 624. Consequently, the Q output of FF 626 goes low at time T.sub.0 as before, but the lower changing rate of capacitor 624 causes the capacitor to reach the R threshold voltage of the FF at a time T.sub.4 which is later than time T.sub.3. Consequently, FF626 is reset later than was the case at the lower clock rate, and counter 626 begins counting at later time T.sub.4. However, the counting ends at a time T.sub.5 earlier than time T.sub.6, because of the higher clock rate. The counting interval of counter 628 establishes the gating interval during which clock pulses are applied to the imager for generating image signals. At the higher clock rate, counter 628 started later and ended earlier than when counting the lower-frequency clock pulses. Consequently, the time of the center of the counting interval relative to H drive remains the same, at time T.sub.C. Thus, notwithstanding the change in clock rate, resistors 612-620 may be adjusted for each clock rate to provide an effective centering control.

Setup of the camera having both clock frequency and centering controls involves manipulation of each pair of controls (CENTERING and SIZE) for each of the red and blue images. If a zoom lens is used, the adjustments must be made at each position of the zoom lens for which a separate set of SIZE and CENTERING resistors is provided.

Other embodiments of the invention will be obvious to those skilled in the art. The entire SIZE and CENTERING control arrangement may be digital in nature rather than analog as shown, in which case the control information for the various positions of the zoom lens may be stored in various addresses of a ROM accessed by addresses generated by a rotation-to-address encoder mechanically coupled to the zoom control. The frequency modification of FIG. 4 may be implemented as a VCO controlled by a phase-lock loop to which a stable controllable low-frequency (0 to 140 KHz) oscillator is coupled for offsetting the frequency of the reference oscillator upward from 7.546 MHz. Alternatively, the frequency modification may be implemented as a high-frequency oscillator controlled in a PLL by use of a programmable frequency divider to produce at the divider output the desired frequencies near 7 MHz. The number of imagers used is a matter of choice, and the image size control may be applied only to one of many. Also, the size control may be used on all of the imagers of a TV camera, if desired, to control the effective magnification of the image, thereby providing an effect similar to that of a zoom lens but which remain in focus. The centering adjustment may be a separate adjustment not coupled to a zoom lens. Line scan arrays may be used instead of rectangular arrays, and the number of elements of the array is a mere matter of design choice. Generally similar techniques may be used for correction in the vertical direction.