WO2000019710A1

WO2000019710A1 - Ccd readout method

Info

Publication number: WO2000019710A1
Application number: PCT/AU1999/000824
Authority: WO
Inventors: Michael C. B. Ashley
Original assignee: Unisearch Limited
Priority date: 1998-09-25
Filing date: 1999-09-24
Publication date: 2000-04-06
Also published as: AU6319999A; AU759445B2

Abstract

A readout method is described that allows multiple readouts of a sub-frame of a Charge Coupled Device (CCD), while continuing to accumulate charge in another part of the CCD. The method uses reverse-clocking of the CCD in the direction away from the readout register after the sub-frame has been read out, in order to restore the remaining charge to its initial position. If the process is performed quickly, there is no need to use a shutter to reduce smearing of the image during the readout of the sub-frame. Depending on the size and position of the sub-frame it is possible to use the CCD readout register to store the entire sub-frame prior to digitisation, thereby minimising smearing. Fast readout of a sub-frame is possible since the readout algorithm clears the sub-frame region without requiring all rows in the CCD to be clocked; if saturation of the remaining region threatens to cause the over-flow of charge into the sub-frame, the number of reverse-clocking cycles can be increased in order to dump the unwanted charge on the side of the CCD away from the readout register.

Description

CCD Readout Method

Technical Field

This invention relates generally to the field of imaging using charge coupled devices (CCDs), and in particular the invention provides an improved tracking system for use with a CCD imaging system.

Background Art

CCDs are electronic devices that record the spatial distribution of incident radiation. The radiation may take the form of photons or particles. These incident photons or particles liberate electrons in the CCD, that then fall into the nearest potential well. The potential wells are called "pixels", and are formed by a combination of implanted energy barriers and voltages on electrodes. The pixels are arranged in a rectangular pattern on the CCD surface. The time period over which the energy is sampled is known as the "integration time" or "exposure time". During this time, the electrons gradually accumulate in the pixels.

The charges in each potential well may be moved to an adjacent pixel. The rectangular grid of pixels in a CCD is conventionally thought of as being aligned in vertical "columns" and horizontal "rows". The CCD has multiple (usually two or three, sometimes four) electrical connections ("electrodes") per pixel arranged horizontally across the device. By varying the voltages on these electrodes (the "vertical phases") it is possible to move the entire pattern of electrons stored in the CCD pixels vertically. The electrons stored in one row are moved into a neighbouring row during this process. This is referred to as "clocking". The CCD also has one or more "readout registers", which are special rows at the top or bottom of the CCD that have "horizontal phase" electrodes, allowing charge to be clocked out horizontally into a readout amplifier. Electrons are clocked into the readout register from the adjacent row on the CCD using the vertical phases. During horizontal or vertical clocking, there is always one column or row (at the end of the register or the top or bottom of the CCD) that does not have charge transferred into it; this column or row is left with no charge, and is therefore "cleared". The readout register can be cleared in its entirety by horizontally clocking it by at least the number of columns that it contains. The imaging surface of the CCD can be cleared by vertically clocking it by at least the number of rows that it contains. To readout the CCD, the usual technique is to vertically clock one row at a time into the readout register, and, between rows, to horizontally clock the readout register into the readout amplifier. This is repeated until all the rows of the CCD have been clocked into the readout register, and the pixels of each row have all been clocked into the readout amplifier(s).

In this way, the electrons from each pixel are presented serially to one or more readout amplifiers. The output from the amplifier(s) is an analog signal that can be subsequently processed using analog electronics and digitised using an analog-to-digital converter (ADC) to provide a computer- readable representation ("image") of the original energy deposition.

A problem that can occur in the use of CCDs is that, if the pattern of incident radiation moves in relation to the CCD during the exposure time, the resulting image will be blurred, or smeared. This is particularly problematic when the exposure period is of the order of seconds or even minutes, over which period the image and the CCD must be kept in alignment. It is therefore desirable to provide autoguiding during the exposure period in order to align the image and the CCD accurately throughout.

In particular, exposure times for astronomical applications are often of the order of minutes, during which time the image may shift on the CCD due to imperfections in the telescope tracking, vibration of the telescope, atmospheric turbulence, and so on. The standard technique for compensating for this effect is to place an additional sensor (another CCD, or photomultiplier tube, or avalanche photo-diode) close to the main sensor (or possibly in an auxiliary telescope) and to read out the additional sensor at a higher rate than the main exposure time. By positioning a "guide star" (i.e., a bright star close to the region being imaged in the sky) on the additional sensor, it is possible to sense the error in the telescope position and therefore correct the error.

The main disadvantages of this technique are (1) the difficulty in ensuring that there is no relative movement between the main CCD and the additional sensor, and (2) the cost and complexity of the additional sensory system.

For example, US Patent No. 5,525,793 discloses an optical head including a primary imaging CCD and a second tracking CCD placed adjacent to the primary CCD. The image incident upon the second CCD preferably includes a bright reference object, and is clocked out many times during the exposure of the primary CCD. The relative shift of the reference object may be used to provide tracking information to the optical head.

The disclosure of US 5,525,793 has the problem that the essential second CCD adds significantly to the cost of the optical head. Additional electronics must be provided, along with associated added complexity of the control software. A highly precise mount for the second CCD must also be provided. Additionally, as the two CCDs are mounted at different positions within the optical head, it is possible that there will be some relative movement between the two, or some change or difference in alignment, creating errors between the drift measured on the secondary CCD and the actual drift occurring on the primary CCD.

Another problem with CCDs is that, in cases where too much incident radiation falls onto a pixel of the CCD, the potential well associated with that pixel may fill up and electrons may "overflow" from that potential well into adjacent potential wells. This causes image degradation.

Disclosure of Invention

According to a first aspect, the present invention provides a method of reading image information from a CCD of a CCD imaging system, the method including the steps of: clocking the CCD in a first direction to move charges representing pixels of image information into a readout register from an original image position in the CCD; clocking the readout register to read out the image information; and clocking the CCD in a second direction opposite to the first direction to clear a region of the CCD adjacent to the readout register.

Embodiments of the present method may digitise only selected pixels that have been clocked into the readout register.

Embodiments of the method may delay all digitisation until after the CCD has been clocked away from the readout register.

According to a second aspect, the present invention provides a method of reading image information from a CCD of a CCD imaging system, the CCD being divided into at least two frames including an image frame and a secondary frame, the method including reading out at least part of the secondary frame of the CCD, including the steps of: clocking imaging information in the CCD from an original position towards a readout register, to move at least a portion of the image information in the secondary frame into the readout register; clocking the readout register to read out the image information obtained from the secondary frame; and clocking the CCD away from the readout register at least until the image information of the image frame is returned to the original position.

The secondary frame preferably has fewer rows than the image frame. The secondary frame readovit time is preferably minimised. Embodiments of the second aspect of the present invention may minimise the secondary frame readout time by digitising only selected pixels of the secondary frame, minimising the delay associated with digitising unwanted pixels.

Embodiments of the second aspect of the present invention may read out a nominal sub-frame of the secondary frame, the readout register having at least as many pixels as the sub-frame, and the sub-frame being read into the readout register such that no two pixels of the sub-frame are read into the same pixel of the readout register. Such embodiments allow the CCD to be vertically clocked away from the readout register prior to serial digitisation of the contents of the readout register. As digitisation is usually a relatively slow step compared to clocking, the secondary frame readout is completed much more rapidly than if some digitisation was required prior to reverse clocking.

Other embodiments of the method of the second aspect of the invention may minimise the secondary frame readout time and the effect of

"pocket pumping" by minimising the number of rows in the secondary frame. "Pocket pumping" results from electron traps in the CCD and is exacerbated by a large number of shifts back and forth.

In some embodiments of the second aspect of the invention, the secondary frame may comprise one row. In such embodiments, the method may include the preliminary step of storing an image template for cross- correlation against the image accumulated in the secondary frame.

In other preferred embodiments of the second aspect of the invention, the secondary frame may comprise two or three rows. Preferably, the secondary frame is situated adjacent to the readout register. In some embodiments of the invention, the CCD may be divided into more than two nominal regions, including a plurality of secondary frames. One or more of the secondary frames may be read out in accordance with the method of the present invention. Preferably, the method of the first and second aspects of the invention is implemented using computer software. The software used can preferably centroid, cross-correlate images, generate template subframes and calculate correction signals.

The CCD used in the first and second aspects of the present invention preferably has three electrodes per pixel. The CCD used in the first and second aspects of the present invention may have four electrodes per pixel.

According to a third aspect, the present invention provides a method of obtaining imaging information during an exposure period of a CCD, the CCD being divided into at least two frames including an image frame and a secondary frame, the method including the steps of: exposing the CCD to incident radiation; reading out at least part of the secondary frame at least once during the exposure period, the reading out of at least part of the secondary frame including the steps of: - clocking the CCD towards a readout register to move image information from the secondary frame into the readout register, without moving any rows of the image frame into the readout register; clocking the readout register to read out the image information of the secondary frame; and - clocking the CCD away from the readout register to restore the image information in the image frame to the original position; and reading out the image frame. During the exposure period of the imaging frame, the secondary frame is preferably read out a plurality of times. Each time the secondary frame is read out imaging information may be obtained.

Embodiments of the method preferably minimise smearing of an image captured by the image frame. Smearing may be minimised by closing a shutter while some or all of the secondary frame is read out. Smearing is preferably minimised by minimising the time for which the image frame is dislocated, that is, minimising the secondary frame readout time. The secondary frame readout time may be minimised by using a readout amplifier that gives the option of two digitisation modes, a fast lower accuracy mode, and a slower high accuracy mode, and using the fast mode of digitisation when reading out the secondary frame. Embodiments of the present invention may minimise the secondary frame readout time by digitising only selected pixels of the secondary frame, minimising the delay associated with digitising unwanted pixels.

Some embodiments of the present invention may only read out a nominal sub-frame of the secondary frame, the readout register having at least as many pixels as the sub-frame, and the sub-frame being read into the readout register such that no two pixels of the sub-frame are read into the same pixel of the readout register. Such embodiments allow the CCD to be clocked away from the readout register prior to serial digitisation of the contents of the readout register. As digitisation is usually a relatively slow step compared to clocking, the secondary frame readout is completed much more rapidly using this arrangement than if some digitisation was required prior to reverse clocking.

The present invention may be used to autoguide a CCD imaging device during an exposure period. The imaging information obtained by each secondary frame readout may be used to determine whether the image is drifting relative to the CCD.

Embodiments of the invention may implement drift correction. Drift correction may be implemented by calculating correction signals based on the imaging information and sending them to one or more motor drives, or to a piezo translator attached to an optical element, or to a positioning system for a device carrying the CCD camera.

In other embodiments, drift correction may be implemented by moving the image information carrying charge on the CCD in one or two dimensions by an amount equal to the detected drift. This keeps the charge distribution correctly aligned with the incident radiation. This form of drift correction is called charge shuffling.

Embodiments of the invention may implement charge shuffling drift correction in one dimension only, by clocking the image towards or away from the readout register with correction in the other dimension being achieved by traditional physical translation techniques. Alternative embodiments may use advanced CCD architectures to allow charge shuffling in two dimensions.

In embodiments where charge shuffling is used to compensate for image drift, the CCD used in accordance with the present invention preferably possesses individually controllable phase electrodes. This enables the use of sub-pixel compensation, wherein the image captured on the CCD may effectively be moved by less than a pixel, by leaving an appropriate vertical phase high.

Embodiments of the invention may use more than one of the above methods of drift correction in order to provide a fast, accurate response to detected drift.

The imaging information obtained by each secondary frame readout may also be used to determine whether pixels of the CCD may be close to overflowing, thereby allowing the option of an early termination of the exposure to prevent image degradation associated with such overflowing.

The method of the present invention may be used in astronomical imaging, medical imaging, or even in recreational imaging devices, such as digital video cameras.

Brief Description of Drawings

Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which: Fig. 1 is a pictorial representation of a CCD;

Figs 2a - 2h illustrate a method of reading out a 3x3 sub-frame within the secondary region;

Figs. 3a and 3b illustrate another method of reading out a 3x3 sub- frame;

Figs. 3c and 3d illustrate yet another method of reading out a 3x3 sub- frame; and Fig. 4 is a pictorial representation of an astronomical auto-guider system.

Modes for Carrying Out the Invention

Fig. 1 shows a CCD 10 with x columns and y rows, including a secondary frame 11 of n rows, where n is less than y, and an image frame 12 having y-n rows. In the present embodiment, the CCD 10 has a single readout register 13, at the "bottom" of the imaging region. The secondary frame 11 is adjacent to the readout register 13.

In accordance with the present invention, the n rows of the secondary frame 11 are clocked one at a time into the readout register 13, and each row is horizontally clocked into a readout amplifier 14. The CCD is then reverse clocked by n rows to restore the image frame 12 to its original location. In this manner, the contents of the secondary frame 11 can be repeatedly sampled during the exposure time, while simultaneously allowing the remaining part of the image to continue integrating. This provides the opportunity for auto-guiding, or for estimating when the exposure should terminate to avoid overflow (or saturation of the signal chain, which may happen at a lower number of electrons, depending on the gain of the readout amplifier).

Note that during the reverse clocking operation, the original contents of the secondary frame 11 are removed, and replaced by zero charge, that is, the secondary frame 11 is cleared. Thus, a fresh image is generated in the secondary frame 11 between each readout.

During the time that the secondary frame 11 is being read out, the image frame 12 is shifted away from its original location, and any radiation that falls on the CCD 10 will produce electrons in the wrong pixels, i.e., the image in the image frame 12 will be smeared by up to n rows. This effect can be minimised in a number of ways, for example: closing a shutter during the time that the secondary frame 11 is read out, in order to block the radiation impinging on the CCD 10, or - minimising the time required to read out the secondary frame 11.

Provided the total time spent reading out the secondary frame 11 and reverse- clocking is negligible compared with the exposure time for the image frame 12, the resultant degradation of the image in the image frame 12 will be insignificant for most purposes. Minimising the readout time of the secondary frame 11 can be implemented by minimising the number of rows in the secondary frame 11. One row may be sufficient for purposes such as exposure time estimation and auto-guiding in one dimension, and the correct techniques can enable two dimensional auto guiding. Two rows allows simpler auto-guiding in two dimensions, while three or more rows improves the accuracy of the auto- guiding and gives increased stability in the event of rapid unpredictable image motion.

The secondary frame readout time may also be minimised by minimising the number of pixels in the secondary frame 11 that are digitised. The digitisation process is often the most time-consuming part of reading out a CCD, requiring typically 20 microseconds per pixel, compared to horizontal and vertical clocking periods of the order of 2 microseconds. For example, as shown in Figure 2, it may be the case that only a 3 x 3 pixel sub-frame 15 is required to determine the auto-guiding correction, in which case only 9 ADC conversions are necessary, as opposed to x x 3 conversions if the secondary frame was digitised in its entirety.

The secondary frame readout time may also be minimised by using the readout register 13 as a summing register and postponing all digitisation until the image frame 12 is back in its original position. A general clocking sequence required to read out an m x n pixel sub-frame, adjacent to the readout register 13, from the secondary frame 11 could be:

1) Clear the readout register 13.

2) Clock one row into the readout register 13.

3) Have n rows been clocked? If so, jump to step 6. 4) Clock the readout register 13 horizontally by at least m columns, without digitisation.

5) Jump to step 2.

6) Reverse clock the vertical phases by n rows.

7) Clock the readout register 13 horizontally (without digitisation) until the first pixel from the m x n pixel sub-frame is just about to enter the readout amplifier 14.

8) Clock the readout register horizontally by m x n pixels, with digitisation.

Note that this technique leaves image frame 12 away from its original position for only 2n vertical clock periods plus m x (n-1) horizontal clock periods. For a 3 x 3 pixel sub-frame 15, and 2 microsecond clocks, this requires only 24 microseconds, which will result in negligible smearing for many applications.

This technique is shown in Figs. 2a - 2g for a 3 x 3 pixel sub-frame. Fig. 2a shows the readout register 13 after it has been cleared. One row is then clocked into the readout register 13, as shown in Fig. 2b. Pixels "A", "B", and "C" of interest are then clocked along the readout register by three pixels, as shown in Fig. 2c. Another row of the sub-frame 15 (containing pixels "D", "E", "F") is then clocked into the readout register 13, as shown in Fig. 2d. Once again, the readout register 13 is clocked sideways, as shown in Fig. 2e. The final row of the sub-frame 15 (containing pixels "G", "H", "I") is then clocked into the readout register 13, as shown in Fig. 2f. The nine pixels of the sub-frame 15 may then be read out. Note that part of the image frame represented by "x" has been clocked into the sub-frame 15. This is restored to the image frame when the CCD is clocked away from the readout register 13 by three rows, as shown in Fig 2g. This step also clears the sub-frame, allowing a fresh image to be recorded in the sub-frame 15.

Note that this readout process results in all the pixels in Fig. 2h that are labelled with the same letter being summed together. This technique is therefore ideally suited for images where there is an isolated bright object (e.g., a "guide star") in the m x n pixel sub-frame, so that the summing of the surrounding pixels that occurs when the rows are clocked into the readout register 13 does not greatly affect the resultant image. However, even when the surrounding image does contribute to the value in the horizontal register 13, later analysis of the digitised data to compensate for this effect, possibly with reference to previously stored images, may still allow use of this method.

The summing of signal from unwanted pixels can be partially avoided if the sub-frame is sufficiently close to one or other end of the readout register. Figure 3 shows an alternative method of reading out a sub-frame 15 which may be used where the sub-frame 15 is situated close to one or other end of the readout register. The readout register is clocked by sufficient columns, either towards or away from the readout amplifier as appropriate, so that the sub-frame rows are always added into cleared locations within the readout register. Fig. 3a shows the sub-frame before readout, with all pixels that will be summed, labelled with the same letter. Fig. 3b shows the situation after all the pixels in the sub-frame have been shifted into the readout register. Note that the readout register has been shifted towards the readout amplifier by 5 pixels between each row of the sub-frame, thereby providing cleared pixels into which to add the next row. The net result is that fewer pixels have been summed together, making the correction for this effect less uncertain. Furthermore, it is possible to have some choice over the pixels that are summed by appropriate shifting of the readout register. This may be advantageous should there be a bright object adjacent to the sub- frame. Once all pixels of the sub-frame 15 have been clocked into the readout register 13, the CCD is reverse clocked by three rows to restore the image frame to its original position, the readout register 13 is clocked into the readout amplifier 14, and the pixels of the sub-frame 15 are digitised.

Figures 3c and 3d show another method of reading out a sub-frame. Fig. 3c shows the sub-frame 15 before readout, with all pixels that will be summed, labelled with the same letter. Fig. 3d shows the situation after all the pixels in the sub-frame 15 have been shifted into the readout register. In this case, a bright unwanted object in the image, represented by 'o' in Figures 3c and 3d, is adjacent the sub-frame 15, and the sub-frame 15 is close to the edge of the CCD. By clocking the readout register 13 away from the readout amplifier 14 between rows, all pixels of the sub-frame 15 may be read into the readout register 13 without the need for digitisation. By clocking the readout register by 7 columns between rows, it can be seen that the content of pixels 'o' does not corrupt the pixels A' to T. Again, once all pixels of the sub-frame 15 have been clocked into the readout register 13, the CCD is reverse clocked by three rows to restore the image frame to its original position, the readout register 13 is clocked into the readout amplifier 14, and the pixels of the sub-frame 15 are digitised. Pixels 'o' between pixels A' to T in the readout register do not necessarily need to be digitised.

The preferred embodiment of this invention uses a CCD camera with controlling electronics that allow control of the readout process using a computer. The CCD chip and electronics must be capable of clocking the vertical phases in two directions. It would also be advantageous to have a high-speed digitisation mode specifically for digitising the sub-frame, since the ultimate accuracy that comes with slow digitisation is often not necessary.

It would also be desirable to be able to leave an arbitrary vertical phase high, in order to implement the sub-pixel charge-shuffling technique. In the preferred embodiment the CCD control electronics are controlled by a computer that is programmed to implement the present invention. For example, for an astronomical auto-guiding application, the software on the computer should be able to find the center of a guide star image by using a technique such as centroiding. An alternative to centroiding is to cross-correlate sub-frames with a previously generated template sub- frame. The software should also be able to calculate correction signals, and send them to, for example, a piezo translator or a motor drive. The computer might be implemented as a microcontroller that is built into the CCD control electronics.

The present invention allows compensation of mid exposure image drift, while ensuring no relative movement between the drift sensor and the imaging area, and possibly without requiring additional cost and complexity. As shown in Fig. 4, by using the CCD 20 as the primary imaging device and as the auto-guiding sensor, there can be no relative displacement, and the cost of the added functionality is very small, particularly if the CCD controller 21 is designed with flexibility in mind (as most astronomical systems are).

To use the present invention for astronomical auto-guiding, the secondary frame is read out repetitively at a rate which is typically between 100 Hz and 0.01 Hz depending on, among other things, the size of the region, the speed of the CCD electronics, the speed of the computer 22, the brightness of the guide star, and the frequency with which corrections to the image position are necessary. Each secondary frame image is then processed by the computer 22 to measure its offset from the desired position. The offset can be derived from a centroid of one or more bright stars in the secondary frame, or from a cross-correlation of the secondary frame with a previously acquired template image (e.g., the first secondary frame, or an image which can be obtained at leisure prior to the exposure commencing, or an accumulated average of previous secondary frame images).

If the image scene is sufficiently complex (e.g., if there are multiple stars in a single row), it may be possible to use a single row to derive corrections in both axes, by cross-correlating the row with a previously stored template image of two or more adjacent rows. This has the advantage of minimising the number of vertical clock cycles, and hence minimising image smearing and charge transfer efficiency degradation. Once the offset has been determined from the secondary frame readout, the image needs to be shifted relative to the CCD. This can be done using a number of methods including movement of the CCD itself with piezo translators, tilting a mirror in the optical path, or driving the telescope 23 with motors 24. Another possibility is to employ charge-shuffling to move the stored image on the CCD. Typical CCDs can only apply vertical charge shuffling, although more complicated CCD architectures may use 2- dimensional charge shuffling to correct image drift in two dimensions.

Sub-pixel offsetting is possible using charge shuffling by choosing the closest vertical phase to leave positively charged.

The vertical charge-shuffling technique can be vised with telescopes that have no declination motor drive (in which case lines of constant declination should be aligned along columns of the CCD) or which have no provision to vary the right ascension tracking rate (in which case the axis that drifts the most should be aligned along a column).

Also, the above realignment techniques may be combined. Foiexample, if the image needs to be realigned by 2.10 pixels vertically, the vertical charge shuffling technique can rapidly account for 2.00 pixels of the correction, and then the (relatively slow) motor drive can make up the final 0.10 pixel adjustment. This may provide a response to image drift that is both rapid and provides sub pixel accuracy. Although the invention has been described with reference to particular examples of the invention, it should be appreciated that it may be exemplified in other forms. For instance, the method of the present invention may be used in medical imaging, or be applied in digital video cameras. The method of the present invention can be generalised to a wide variety of readout algorithms for special purposes, including multiple sub- frames, non-rectangular sub-frames, and sub-frames that are not adjacent to the readout register.

The present invention allows higher speed readout of a sub-frame than is usually possible, due to the ability to clear the sub-frame of charge without having to vertically clock the entire CCD.

Often, an object of interest in an image being captured by the CCD only occupies a small portion of the CCD imaging surface, and it is desired to repetitively read out the image as often as possible. For example radiation from a star may only fall on 4 pixels. The readout of information relating to a small object may be achieved much more rapidly by only reading out the small portion of the CCD which includes the object of interest. To further increase the speed of such readout, the image may be situated on the CCD such that the object of interest is close to the readout register, and so a minimal number of rows needs to be clocked into the readout register, minimising the amount of time needed to read out the required information.

Additionally, the CCD must be cleared before the next image can be recorded and read out. Rather than going through the time consuming process of clocking all the rows of the CCD (and thereby clearing it), the reverse clocking clears the secondary frame much more rapidly, as significantly fewer rows need to be clocked in order to clear the region of interest. For example, in this application, the image frame may not be used. To avoid the possibility of saturation in the image frame causing an overflow of charge into the secondary frame, it may be desirable to reverse-clock the CCD by more than n rows, so that the image frame is gradually "scrolled off the top of the CCD.

In some applications, it may be desirable to monitor the rate at which radiation is accumulating during the exposure. The present invention allows this to be performed using the secondary frame without significantly disturbing the remaining image. At regular intervals, the secondary frame can be read out, and the image analysed to determine the flux of radiation since the secondary frame was last read. This information may be used for a variety of purposes, for example, the exposure can be terminated early if it is determined that the CCD is close to saturation, or the exposure can be terminated early if conditions have changed (e.g., in an astronomical context, clouds may have started to interfere with the photometric stability).

The present invention may also be used to assist imaging in cases where a faint object of interest is adjacent to a bright object. By positioning the bright object in the secondary frame, the charge build-up associated with the bright object can be regularly read out in accordance with the present invention, thereby clearing the region, and removing the possibility of electron overflow, which otherwise may limit the exposure period or image quality.

It will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.

Claims

CLAIMS:

1. A method of reading image information from a CCD of a CCD imaging system, the method including the steps of: clocking the CCD in a first direction to move charges representing pixels of image information into a readout register from an original image position in the CCD; clocking the readout register to read out the pixels of image information; and clocking the CCD in a second direction opposite to the first direction to clear a region of the CCD adjacent to the readout register.

2. The method of claim 1 wherein only selected pixels of the image information are digitised after being read out of the readout register.

3. The method of claim 1 or claim 2 wherein the digitisation occurs after all of the selected pixels have been moved into the readout register.

4. The method as claimed in any preceding claim wherein the control of the CCD is implemented using computer software.

5. The method according to claim 4 wherein the software includes functions to centroid, cross-correlate images, generate template sub-frames and calculate correction signals.

6. The method as claimed in any preceding claim wherein the CCD has individually controllable phase electrodes and the clocking of the CCD is performed using a multiphase clock whereby the image is moved in sub-pixel increments.

7. A method of reading image information from a CCD of a CCD imaging system, the CCD being divided into at least two frames including an image frame and a secondary frame, the method including reading out at least part of the secondary frame of the CCD, including the steps of: clocking imaging information in the CCD from an original position towards a readout register, to move at least a portion of the image information in the secondary frame into the readout register; clocking the readout register to read out the image information obtained from the secondary frame; and clocking the CCD away from the readout register at least until the image information of the image frame is returned to the original position.

8. The method of claim 7 wherein the secondary frame has fewer rows than the image frame.

9. The method of claim 7 or claim 8 wherein only selected pixels of the secondary frame are digitised.

10. The method of claim 9 wherein the secondary frame is read out by reading out a nominal sub-frame of the secondary frame, the readout register being able to simultaneously hold all of the pixels of the sub-frame, and the sub-frame being read into the readout register such that no two pixels of the sub-frame are read into the same pixel of the readout register.

11. The method of any one of claims 7 to 10 wherein the secondary frame has from one to three rows.

12. The method of any one of claims 7 to 11 wherein the method includes the preliminary step of storing an image template for cross-correlation against the image accumulated in the secondary frame.

13. The method of any one of claims 7 to 12 wherein the secondary frame is situated adjacent to the readout register.

14. The method of any one of claims 7 to 13 wherein the CCD is divided into more than two frames.

15. The method of any one of claims 7 to 14 wherein the control of the CCD is implemented using computer software.

16. The method of claim 15 wherein the software includes functions to centroid, cross-correlate images, generate template sub-frames and calculate correction signals.

17. The method of any one of claims 7 to 16 wherein the CCD has individually controllable phase electrodes and the clocking of the CCD is performed using a multiphase clock whereby the image is moved in sub-pixel increments.

18. The method of any one of claims 7 to 17 wherein the CCD has three electrodes per pixel.

19. The method of any one of claims 7 to 17 wherein the CCD has four electrodes per pixel.

20. A method of obtaining imaging information during an exposure period of a CCD, the CCD being divided into at least two frames including an image frame and a secondary frame, the method including the steps of: exposing the CCD to incident radiation; reading out at least part of the secondary frame at least once during the exposure period, the reading out of at least part of the secondary frame including the steps of: clocking the CCD towards a readout register to move image information from the secondary frame into the readout register, without moving any rows of the image frame into the readout register; clocking the readout register to readout the image information of the secondary frame; and clocking the CCD away from the readout register to restore the image information in the image frame to the original position; and reading out the image frame.

21. The method of claim 20 wherein the secondary frame is read out a plurality of times during the exposure period.

22. The method of claim 20 or 21 wherein smearing of the image information in the image frame is minimised by closing a shutter while some or all of the secondary frame is read out.

23. The method of any one of claims 20 to 22 wherein a time taken foieach readout of the secondary frame is minimised by using a readout amplifier that provides a fast lower accuracy digitisation mode, and a slower high accuracy digitisation mode, the fast mode of digitisation being used when reading out the secondary frame.

24. The method of any one of claims 20 to 23 wherein a time taken for each read out of the secondary frame is minimised by digitising only selected pixels of the secondary frame.

25. The method of any one of claims 20 to 24 wherein the secondary frame is read out by reading out a nominal sub-frame comprising the selected pixels of the secondary frame, the readout register being able to simultaneously hold all of the pixels of the sub-frame, and the sub-frame being read into the readout register such that no two pixels of the sub-frame are read into the same pixel of the readout register.

26. The method of any one of claims 20 to 25 wherein imaging information obtained by each secondary frame readout is used to determine drifting of an image relative to the CCD.

27. The method of any one of claims 20 to 26 wherein drift correction is implemented by calculating correction signals based on the image information of the secondary frame and sending the signals to one or more location devices for altering the location of the CCD or a structure on which the CCD is carried.

28. The method of any one of claims 20 to 27 wherein drift correction is implemented by moving the image accumulated on the CCD by an amount related to the detected drift.

29. The method of any one of claims 20 to 28 wherein the CCD possesses individually controllable phase electrodes and clocking of the CCD is performed using a multiphase clock whereby the image is moved in sub-pixel increments.

30. The method of any one of claims 20 to 29 wherein the image information obtained from each secondary frame readout is used to predict an overflow condition in pixels of the CCD.