CN1882031A - Method and equipment for forming multi-focusing images - Google Patents

Abstract

A method of producing a multifocus stack image of an object is provided. The stack of images has multiple images of the target, each image having a corresponding focus range or position. The method utilizes relative scanning motion between a target and an array of photodetectors that is used to repeatedly receive image information from the target as scan lines during scanning. The relative focus between the object and the array is varied during scanning between the focal ranges or positions of the individual images to obtain each image in the stack. Each image is thus formed from image information obtained from various focus ranges or positions during scanning. Apparatus for carrying out the method is also provided.

Description

Method and apparatus for forming multi-focus stack images

technical field

The present invention relates to methods and apparatus for forming multi-focus stack images.

Background technique

In some market areas such as medical applications, samples are examined using a microscope and a digital image is generated using a two-dimensional digital camera attached to the microscope. The observed sample captured by the digital camera is limited in area. For a 40x objective, the typical area is only 0.7mm wide. Given an active area of 64 x 24 mm on a microscope slide, the possible sample area is only a very small area. One approach to this is to step and repeat over the entire sample area, also known as macro dithering. A more preferred process is to use a line scanning unit similar to that disclosed in US 6,711,283, where 0.7 x 64 mm strips of data can be collected. Adjacent strips can then be scanned so that the images can be stitched or stitched together, as explained in patent GB 2206011.

One problem with strip scanning, as mentioned in US 6,711,283, is that focus must be maintained throughout the scan length. As an example, the depth of focus for a 40x lens with a numerical aperture of 0.65 is approximately 1 micron. Typical microscope slides are not manufactured to maintain this type of tolerance and may warp by more than 1 micron when mounted due to the mounting method or under gravity. And the samples to be imaged themselves are not flat down to 1 micron. This problem is solved in US 6,711,283 by building a focus map at the scan length alone and dynamically adjusting the focus during the scan to work with this focus map. Unfortunately, building a focus map for each sample is time consuming. A typical focusing method is to scan the same area at different focus levels and then use an evaluation algorithm to determine the best focus. There are several evaluation algorithms that have been used, one example is taking the sum of squares of the differences between adjacent pixels. This scoring algorithm yields a function such as that shown in Figure 1, where peaks (indicated by arrows) are considered focal points.

Another way to solve this problem is to get multiple scans with different levels of focus. This is called the focus stack or Z-stack and is shown in Figure 2. The idea is that at any time at least one of the scanned images is in focus and that the Z-stack 400 of images can be combined later to give a single in-focus image. Software to combine images is available from several suppliers. The problem with this approach is that for small depths of focus over the sample focus variation, many image layers need to cover the complete focus range and this can be time consuming. The line of best focus is shown on 401 . The images in the stack at different locations 402 provide little useful information.

There is thus a need to address these disadvantages.

Contents of the invention

According to a first aspect of the present invention we provide a method of generating a multi-focus stack image of an object, the stack image comprising a plurality of images of the object, each image having a corresponding focal range or position, the method comprising:

causing relative scanning motion between the target and the photodetector array for repeatedly receiving image information from the target during scanning in scan lines; and

causing changes in the relative focus between the target and the array between said focus ranges or positions of the individual images during the scan, resulting in images in the stack, wherein each image is obtained from each focus range or position during the scan The image information is formed.

The present invention thus differs considerably from prior art methods. In the existing method, stack images are obtained in successive image frames, but in the present invention images are obtained by scanning lines. Importantly, the scanlines are obtained by repeatedly swapping between different focus positions of the image (fixed or within a range) during the scan, and are then used to form the stack at the end of the scan. Images with different focus. These images can be thought of as focus levels or focus layers. When discussing scanning, it is understood that under arrays with more than one row of detectors the term encompasses the output of several detector rows. However, the number of detectors on a scan line is orders of magnitude larger than the small number (approximately less than 16) of such an array.

The present invention thus has significant advantages over the prior art in that relatively inexpensive equipment can be used. It also avoids the need to scan the target multiple times, which can cause problems in recording the resulting information. In addition, no auxiliary equipment is required for generating the focus map beforehand. It also allows differentially focused image information to be obtained for very localized areas in a short period of time and without the need for subsequent scanning. This helps to ensure that no changes occur in the optics or sample at any stage of the intervention and can provide an on-line change in focus level for objects with significant configuration in the scan.

The method is typically repeated for several regions of the target and a multifocus stack image is obtained in a single scan of the target. Preferably, the relative focus is modified after a scanline of image information has been obtained from the target for a particular image in the stack, such that the scanline is obtained for at least one other image before another scanline (image information) is obtained again for said particular image . Focusing can thus be cycled between images during scanning to construct these images from the individual scan lines. Image information of a common area or portion of the target can be obtained for each image, and this can be achieved by suspending relative motion during the scan. Alternatively, the scanning motion may be substantially continuous, and the image information obtained for each image may then be interpolated to form corresponding images within the stack with different focus positions or ranges.

The array may be a one-dimensional array defining a first direction, and the relative scanning motion is in a direction substantially perpendicular to the first direction. The focus positions or ranges may be evenly or non-uniformly spaced in focus from each other. The focus range includes a focus region between two out-of-focus positions defining the range. When using focus ranges, the focus ranges may or may not overlap for different images in the stack. When a focus range is used, the method may include using image information from scanlines during scanning to modify the focus to obtain successive scanlines for each image as scanning continues. A focus rating curve as a function of focus position may be used for this purpose, and the method may then further comprise controlling the focus on the images in the stack so as to cover an "ideal" focus position for a particular region. The centermost image (group) in the stack can be arranged to correspond to the ideal focus position in these regions.

Each image in the stack can be utilized to generate an output image having a depth of focus according to the focal range or location of the image from which it was constructed.

The array may include a plurality of pixels arranged in sub-arrays generally along the scan direction. They may be spaced substantially along the scan direction and each sub-array adapted to receive light of a corresponding colour.

In some instances image information is obtained from adjacent regions. When "m" integer sub-arrays are provided, it is preferable that the spacing of the sub-arrays is mn-1 seen from the array in units of area width, where n is a non-zero integer, typically to make the image information obtained from different areas at different times consistent. intertwined.

Determines the speed of relative motion based on the dimensionality and orientation of the array.

According to a second aspect of the present invention we provide an apparatus for producing a multi-focus stack image of an object, the stack image comprising a plurality of images of the object with corresponding focus ranges or positions, the apparatus comprising:

an array of photodetectors for receiving image information from the image in scan lines;

a scanning device for providing relative motion between the array and the target;

focusing means for controlling the relative focus between the array and the target, and

a control system for operating the scanning device to cause relative scanning motion between the target and the photodetector array; for adapting control of the array to repeatedly receive image information from the target during scanning; and for operating the focusing means to cause a change in the relative focus between the target and the array during scanning between said focal ranges or positions of the individual images, whereby each image in the stack is obtained by a respective focal range during scanning Or the image information obtained at the position is formed.

The array may comprise a one-dimensional array or a plurality of sub-arrays arranged in a direction substantially perpendicular to the scan direction. The number of pixels can be large, for example about 5000. A number "m" of sub-arrays may be provided, the sub-arrays being spaced such that the corresponding image information derived from each segment is spaced in the image by approximately an integer number of segments in the scan direction. This separation may be actual physical separation or equivalently optical separation, for example provided by the use of beam splitters with physically separated sub-arrays.

For color images, preferably the sub-arrays each include filters to receive light corresponding to some particular colour. The focusing component may achieve focusing by moving the array or target or, when it includes an imaging lens, by moving an imaging lens or components of an imaging lens. When the focusing component includes stacked mirrors, the focusing component achieves focusing by moving the stacked mirrors. It is also possible to use windows with controllable optical thickness for this purpose, for example optoelectronically active quartz windows or rotatable windows with variable optical thickness as a function of the rotation angle.

The apparatus and method can be used in several imaging applications, but it has been found to be particularly beneficial in microscopes where the field of view and depth of field are typically quite efficient.

Description of drawings

Some examples according to the method and apparatus of the present invention are now described with reference to the accompanying drawings, in which:

Figure 1 shows a focus evaluation curve known in the art;

Figure 2 shows a prior art stack image and ideal focus;

Figure 3 is a schematic perspective view of array scanning;

Fig. 4 shows the arrangement of scanning lines in scanning;

Figure 5 shows trifocal stacking with a "stop-start" scan;

Figure 5a shows a specific three-focus stack scheme;

Figure 6 shows a three-focus stack with smooth scan;

Figure 7 shows images of uneven spacing in the stack;

Figure 8 shows a non-planar image in a stack;

Figure 9 shows non-uniform spacing between images during scanning;

Figure 10a shows the intersection of the lower image;

Figure 10b shows the intersection of the upper and lower focus stack images;

Figure 11 shows in-focus evaluation curves during focus tracking;

Figure 12 shows the edges of the focus rating curve during focus tracking;

Figure 13 shows an image stack for tracking focus;

Figure 13a shows the stack for this sample;

Figure 14 shows the variation in focus extremes of the stack of images;

Figure 15 shows the focus evaluation curves at the two positions of Figure 14;

Figure 16 shows the three outer stack image positions in focus;

Figure 17 shows three positions, two of which are on the edge of focus;

Figure 18 shows a multi-row detector array;

Figure 19a shows three adjacent row detectors with three focus stacks and smooth scan;

Figure 19b shows four adjacent row detectors with three focus stacks and smooth scan;

Figure 19c shows two adjacent row detectors with three focus stacks and smooth scan;

Figure 20 shows three row detector arrays spaced apart;

Figure 21a shows a 3-row detector with a 2-row spacing;

Figure 21b shows a 3-row detector with a 5-row spacing;

Figure 21c shows a 3 row detector with 8 row spacing.

Figure 22a shows a 3-row detector with 2-row spacing, three focus stacks and smooth scan;

Figure 22b shows a 3-line detector with 5-line spacing, three focus stacks, and smooth scan; and shows 3 focus positions;

Figure 22c shows a 4-row detector with 3-row spacing, three focus stacks and smooth scan;

Figure 23 shows a non-adjacent row detector with narrowed lines;

Figure 24 shows RGB scanned with switched light and shows a 3 row detector with 2 row spacing and 3 focus positions;

Figure 25a shows multiple detectors arranged in color groups;

Figure 25b shows multiple detectors arranged in a color sequence;

Figure 26a shows a device containing two beam splitters and three arrays;

Figure 26b shows two arrays, two mirror structures;

Figure 26c shows three arrays, two mirror structures;

Figure 27a illustrates the movement of the detector head;

Figure 27b illustrates the movement of the imaging lens;

Figure 27c illustrates the use of a movable beam stacker;

Figure 27d illustrates movement of the sample;

Figure 27e illustrates moving parts within the imaging lens;

Figure 27f shows an example of dimmable thickness;

Figure 27g employs a rotating window of variable optical thickness; and

Figure 28 shows the distribution of variable optical thickness.

Detailed ways

Standard line scanning involves a single row x pixel array, each pixel typically corresponding to a detector in the array. For color, a single row is provided for each of the three colors (eg red green blue). The single row is then traversed in a direction perpendicular to the detector array rows. The traverse speed is set to traverse the detector a distance of one pixel in the scan direction after one "line time" of the detector so that the next line time produces a row of pixels adjacent to the previous row. This is illustrated in Figure 3, where the 1D array is scanned in the direction indicated by the arrow.

Figure 4 is a perspective layout of the 1D array scan seen from the tail of the 1D array. The direction of traverse is indicated by an arrow, where the first scan line is marked with a "1", the second line with a "2", and so on. The simplest embodiment is to scan and adjust the focus to a different focus stack position during the move to the next row. Figure 5 shows the situation where three such stacks are obtained. This involves a stop-start traverse scan, but does not require interpolation of scan lines within the same image. It can be seen that the focus direction indicated by the vertical arrows in Figure 5 is roughly in this case perpendicular to the traverse (scan) direction. Figure 5a shows this scheme in more detail, where the linear array 1 has an orientation into the plane of the figure. A lens is used to achieve focus variation. A sample with variable thickness is shown at 15 positioned on a slide acting as a support. The position of the scanning line is indicated by arrow X, while the scanning direction is shown by Y. In this embodiment the traversing has a "stop-start" action in the traversing, which is not always desirable because stopping and starting the traversing mechanism can cause position errors that show up as jerks in the image.

In another embodiment the traversing can be arranged to be smooth (fixed scan speed), but three times slower than in FIG. 4 and produce the image shown in FIG. 6 . Interpolation methods can then be used to clean each Z-stack image with neighboring images if desired. Although these examples show three focus stack images, preferably from 2 to any practical number of multi focus stack images can be produced. It is also not necessary to have the focused stack images equidistantly spaced, so that for example it is possible to have 5 stack images including three central stacks and two peripheral focused stacks. This is shown in Figure 7, where rows 1 and 5 and the more closely spaced rows 2, 3 and 4 are more spaced apart.

There is no need to keep the focus stack in a fixed plane. This is illustrated in FIG. 8 . Such a situation may arise in a stack image acquisition system that follows a predetermined non-planar projection during scanning, or that follows a non-planar surface in a sample by iteratively determining the best focus position during scanning. As illustrated in Figure 9, the spacing between lines in adjacent images within a stack need not be constant. For example in Figure 9, the top and bottom images in the stack exhibit a variable spacing, while the three central image rows have a fixed spacing in the scan.

In some cases, as shown in Figures 10a, 10b, focusing stacks may be arranged to cross each other. In Figure 10a, the lowest image in the stack intersects the next-lowest image during the scan, while in Figure 10b the upper and lower images intersect the adjacent image layers, and the order at the start of the scan is 1, 2, 3, 4, 5 and when the scan ends the order becomes 2, 1, 3, 5, 4.

A particular advantage of taking multiple focus stacks at once and adjusting the focus during the scan is that it is possible to track the focus of the surface of a non-planar body, such as a tissue sample or a rock sample, if the two outer focus positions are aligned on the slope of the focus curve , it is possible to predict the best focus position and adjust the focus position to place the central focus position at the best focus position. This can be achieved by using the three focus positions C, D, E with the aid of the focus evaluation curve shown in FIG. 11 . In this way it is possible to monitor whether the scanner is in focus by looking at the relative ratings of the detector at all three positions. If the focus is moved from the focus position, it changes the relative focus value as shown in Figure 12.

Comparing Figures 11 and 12, it can be seen that the relative ratings for the outer focus positions (C and E) change but the center focus position (D) does not necessarily change. If this is monitored during scanning, the focus can be adjusted when this occurs to bring the focus position (eg D) back to the center of the focus range. This can reduce the number of focused stacks that need to be scanned because there are no areas where the focused stacks are not close to the image focus as shown in FIG. A big advantage is the time saved by not having to scan more piles of images, many of which have only little useful information. By adjusting the nominal focus during scanning to follow the sample, this will give fewer different focus values at a time. This is shown in more detail in Figure 13a.

The provision for separately adjusting the focus enables the scanner to place the outer focus level on the edge of focus as the range of good focus values changes. This is shown in FIGS. 14 and 15 , where FIG. 15 shows the evaluation curves at two positions A, B in the scan of FIG. 14 . This variable focus range condition occurs, for example, in cell layer scans in which some embodiments cells are stacked on top of other cells in multiple layers and the number of cell layers varies during the scan. The idea of varying the spacing between images in a stack, shown in Figures 10 and 14, offers a number of advantages. Many existing solutions cannot provide this functionality.

Another embodiment is to align the out-of-focus images on the edge of the focus range but not actually out of focus. The focus assessment is then monitored to ensure that all detectors are out of focus. Once an outer detector shows an out-of-focus rating, adjust the focus to bring all detectors back into focus. Fig. 16 shows three focus positions C, D, E (all in focus), while Fig. 17 shows a position which is in focus despite being on the edge of focus. This enables the user to focus to see if the sample is being scanned at a depth of focus greater than the scanner's focus range. It is also possible to use software to combine these images into one image with a device to increase the depth of focus.

Although so far we have only illustrated the use of 1-dimensional arrays, multiple detectors can be used to accomplish this task with increased throughput. As shown in Figure 18, one embodiment employs a 3 by x pixel array. The focus is shifted in a sawtooth fashion as before, but with a traverse speed n times greater, where n is the number of row detector arrays. Figures 19a, 19b and 19c show multiple adjacent row detectors with multiple focus positions for 3, 4 and 2 row arrays respectively.

Also it is not required that the plurality of row detectors have rows adjacent to each other. A scanning process with gaps between detectors can thus be established as shown in FIG. 20 . In this state there is a limit on the row spacing relationship of the detectors. For integer scan row spacing, the spacing between 1D detector arrays must be mn-1, where m is the number of 1D detector arrays and n is an integer not equal to zero (the zero case is the same as the adjacent row case). The scanning speed must then be set to m. This gives the scan patterns shown in Figures 21a to 21c and Figures 22a to 22c. The numbers listed in Fig. 21a are the row numbers of the 1-dimensional detector array, so that in this case the number of rows is 3 and their spacing is 2 (m=3, n=1). For Figure 21b, number of rows = 3, spacing = 5 (m = 3, n = 2). For Figure 21c, number of rows = 3, spacing = 8 (m = 3, n = 3). In Fig. 22a, the number of rows = 3, the pitch = 2 (m = 3, n = 1), and there are three focus positions. In Fig. 22b, the number of rows = 3, the pitch = 5 (m = 3, n = 2), and there are three focus positions. In Fig. 22c, number of rows = 4, pitch = 4 (m = 4, n = 1), and there are three focus positions. Although an integer line spacing between detectors may not always be required and may even be desirable to scan with overlapping or "narrowed" lines as shown in FIG. 2 and has three focus positions.

To produce a color image or a multi-channel image, the color of the illumination can be changed on a line-by-line basis and the traverse slowed down by a factor of the number of channels. For example, as shown in Figure 24, for red, green and blue three-color scanning, it should be required to be a traversing speed of 1/3 of the monochromatic speed. In Figure 24, red, green and blue scanning is performed by switching light, and the number of rows in the figure=3, the spacing =2 (m=3, n=1) and there are three focus positions.

Another way to generate red, green, and blue color information is to place red, green, and blue filters on individual rows of detectors. Such combining may include combining all rows of the same color together as shown in Figure 25a or combining color sequences together as shown in Figure 25b. It is important to note that if overlapping or narrowing of the individual colored lines is not required, then the "same-colored-same-colored lines" spacing constraints apply as monochrome line spacing does.

In all cases where the number of focal planes is less than the number of 1D arrays used, the time required for the detector to detect light is shorter than the time required to move to the next position number to prevent image blurring. For example, for a four-row detector with a single focal plane, the light detection time should be less than a quarter of the motion time. It is not necessary to have a single detector system as shown in Figures 18, 20 and 25a, 25b.

A single row of detectors such as that shown in Figure 3 can be combined in various light overlapping methods as shown in Figures 26a to 26c so that the detectors are all located on the same focal plane. Figure 26a shows the use of a device with two beam splitters 4,5 and three arrays 1,2,3. Virtual images of arrays 1, 3 are shown at 1', 3', respectively. The optical axis is shown at 6 . The pitch of the detector array is two scanning lines (m=3, n=1). Figure 26b shows a dual array and dual mirror 7, 8 scheme at a row spacing of 2 (m=3, n=1). Arrays 1 and 2 and chief rays are shown at 1" and 2", respectively. Fig. 26c shows a three-array 1, 2, 3 and two-mirror 7, 8 system at a row spacing of 8 (m=3, n=3).

If the layout is arranged so that the detectors are not placed in the same plane, when any adjustments to the relative focal planes are required during scanning, the detectors must move relative to each other and make it difficult to achieve variable non-uniform focal plane separation, or A large number of non-uniform focus stacks would require a large number of detectors to scan each focal plane simultaneously. Other methods of combining detectors include fiber optic bundles, physically bringing the detectors together and arrays of microprisms arranged on imaging lenses.

There are several possible methods of adjusting focus during scanning, some of which are shown in Figures 27a to 27g. Figure 27a shows the movement of the detector head 9 comprising a double beam splitter system with 3 arrays. The range of focus movement is shown at 10 . Figure 27b shows an alternative example in which a movement of the imaging lens 11 between the sample and the detector is used. Figure 27c shows the use of movable beam stack mirrors 12, 13, 14 to achieve focus. Figure 27d shows how focus modulation is achieved by moving the sample 15. Figure 27e shows that focus changes are achieved by moving elements 16 within the imaging lens 11. Figure 27f shows an alternative device in which a window 17 of adjustable thickness is provided on the sample side, for example an optoelectronically active quartz window. Focusing motion is caused by changing the optical thickness of the window. Figure 27g uses a rotating window 18 of variable optical thickness. The focus movement is caused by changing the optical thickness of the window. The window has a profile that varies the optical thickness as a function of rotation about its axis (ie, along the circumference) and/or varies along the sagittal direction. This is shown in more detail in FIG. 28 .

The approach described here is quite different from the prior art in which multiple detectors are utilized with each detector capturing an image at a different focus position. Instead, with the present invention we can use a single detector in many cases, as shown for example in FIG. 3 . We then vary the focus on a line-by-line basis to form multiple focus values. This is shown in FIGS. 5 to 10 . We could also use multiple detectors eg as in Figures 18 to 25, but at any moment these detectors are in the same focal position.

In summary, a line scanning method is thus provided to generate multi-focus stack images in one pass. A 1-dimensional (typically) detector array is traversed in a direction perpendicular to the axis of the array and is usually in the plane of the detector surface. The focus between scanlines is adjusted for each image of the focus stack image. This process is repeated cyclically as the traverse mechanism moves to the next scan line of the first focus stack image until all focus stack images are formed.

The detector array is typically traversed (relatively) in a direction perpendicular to the axis of the array and typically in the plane of the detector surface, the traverse speed is taken such that the next group of rows is an integer multiple of m rows on the image. The rows do not necessarily have to be of the same size as the sensitive areas of the detectors. The traverse speed relative to rows can be reduced by a factor of the number of color channels (if set), and the illumination color is changed for each channel of the image before or during each change of focus.

The focus level can advantageously be adjusted during scanning in order to track the focal zone. Focus ranges can be determined using these focus levels by looking at the relationship of the focus rating function at at least two levels on either side of the focal zone, where these levels are placed on the edge of the focal zone, e.g., between the in-focus plane and the edge of the focal plane. fixed relationship between them.

In a specific embodiment beam splitters/mirrors/microprisms (near the imaging lens) and arrays can be used to generate multiple confocal line 1D arrays.

Claims (38)

1. A method of producing a multi-focus stack image of an object, the stack image comprising a plurality of images of the object, each image having a corresponding focus range or position, the method comprising: causing relative scanning motion between the object and the photodetector array for repeatedly receiving image information from the object during scanning in scan lines; and The relative focus between the target and the array is caused to vary between said focal ranges or positions of the individual images during the scan, whereby the images in the stack are obtained, wherein each image is made of a The image information obtained by range or position is formed. 2. The method according to claim 1, wherein the method is repeated for a plurality of swaths of the object. 3. A method according to claim 1 or 2, wherein the multifocus stack image is obtained within a single scan of the object. 4. A method according to any one of claims 1 to 3, wherein after a scan line is obtained from the target for a particular image in the stack, the relative focus is changed such that after another scan line is obtained for said particular image A scan line is obtained for at least one other image in the stack before the line. 5. A method according to claim 4, wherein during scanning, scan lines are sequentially obtained for each image to cycle focus repeatedly between images. 6. A method according to claim 5, wherein the relative motion is suspended during each cycle. 7. A method according to any one of claims 1 to 5, wherein the scanning motion is substantially continuous. 8. A method according to claim 7, further comprising interpolating the image information obtained for each image to form corresponding images in the stack having different focus positions or ranges. 9. A method according to any preceding claim, wherein the array is a one-dimensional array defining the first direction. 10. The method according to claim 9, wherein the relative scanning motion is substantially perpendicular to the first direction. 11. A method according to any preceding claim, wherein the focal positions or ranges are focally evenly spaced from each other. 12. A method according to any one of claims 1 to 10, wherein the focus positions or ranges are focally non-uniformly spaced from each other. 13. A method according to claim 12, wherein when using focus ranges, the focus ranges do not overlap. 14. A method according to claim 12, wherein when using focus ranges, the focus ranges overlap. 15. A method according to any preceding claim, further comprising, when using a focus range, using image information during scanning to modify the focus so that successive image information is obtained for each region of each image. 16. The method according to claim 15, further comprising obtaining a rating curve as a function of focus position. 17. A method according to claim 16, further comprising controlling focus for each image in the stack so as to cover a desired focus position for a particular region. 18. A method according to claim 17, wherein the focus of the most central image (group) in the stack is set to correspond to the ideal focus position of these regions. 19. A method according to any preceding claim, further comprising combining the images in the stack to produce an output image having a depth of focus according to the focus range or position of the images from which it was constructed. 20. A method according to any preceding claim, wherein the array comprises a plurality of pixels arranged in sub-arrays substantially along the scan direction. 21. A method according to claim 20, wherein the sub-arrays are spaced substantially along the scan direction. 22. A method according to claim 21, wherein each sub-array is adapted to receive light of a corresponding colour. 23. A method according to claim 22, wherein image information is obtained from adjacent regions for each image and wherein m subarrays are provided, and wherein the subarray spacing seen from the array in units of region width is m x n- 1, where n is a non-zero integer, thereby interleaving image information from different sections obtained at different times. 24. A method according to any preceding claim, wherein the velocity of the relative movement is determined from the dimensionality and orientation of the array. 25. An apparatus for producing a multi-focus stack image of an object, the image stack comprising a plurality of images of the object, each image having a corresponding focus range or position, the apparatus comprising: an array of photodetectors for receiving image information from the target in scan lines; scanning means for providing relative motion between the array and the target; focusing means for controlling the relative focus between the array and the target, and a control system for operating the scanning device to cause relative scanning motion between the target and the photodetector array; further adapted to control the array to repeatedly receive image information from the target during scanning; and also for operating the focusing means to cause a change in the relative focus between the target and the array during scanning between said focal ranges or positions of individual images, whereby each image in the stack is obtained by a respective focal range during scanning Or the image information obtained at the position is formed. 26. Apparatus according to claim 25, wherein the array comprises a one-dimensional array. 27. Apparatus according to claim 25 or 26, wherein the array comprises a plurality of sub-arrays arranged in a direction substantially perpendicular to the scanning direction. 28. Apparatus according to claim 27, wherein m sub-arrays are provided and spaced such that corresponding image information derived from each sector is spaced in the image by substantially an integer number of sectors in the scan direction. 29. Apparatus according to claim 27 or 28, wherein the sub-arrays each include a filter to receive light corresponding to a particular colour. 30. Apparatus according to claim 28 or 29, further comprising a beam splitter to provide the physical separation of the sub-arrays according to claim 28 and the imaginary spacing of the sub-arrays. 31. Apparatus according to any one of claims 25 to 30, wherein the focusing means achieves focusing by movement of the array. 32. Apparatus according to any one of claims 25 to 31, further comprising an imaging lens and wherein the focusing means achieves focusing by movement of the imaging lens or imaging lens members. 33. Apparatus according to any one of claims 25 to 32, further comprising a stack mirror and wherein the focusing means achieves focusing by movement of the stack mirror. 34. Apparatus according to any one of claims 25 to 33, wherein the focusing means achieves focusing by movement of the object. 35. Apparatus according to any one of claims 25 to 32, further comprising a window of controllable optical thickness, wherein the focusing means operates the window to achieve focusing. 36. Apparatus according to claim 35, wherein the window is formed of a photovoltaic material. 37. Apparatus according to claim 36, wherein the window is rotatable and has a variable optical thickness as a function of the angle of rotation. 38. Apparatus according to any one of claims 25 to 37, wherein the apparatus forms part of a microscope system for imaging microscopic objects.

Images