Laminated imaging system and method for generating images

Abstract

The present invention relates to a stacked imaging system, comprising a plurality of light sources adapted to emit light onto a sample position, wherein the light sources are arranged in a predetermined pattern; a controller adapted to control the operation of the plurality of light sources; wherein at least one of a) the predetermined pattern of the light sources and b) the controlled operation of the plurality of light sources is adapted to compensate for geometric effects caused by the arrangement of the light sources relative to the sample position.

Description

Technical Field

[0001] The present invention relates to a stacked imaging system and a method for generating an image using the stacked imaging system, and in particular to a Fourier stacked imaging system and a method for generating an image using the Fourier stacked imaging system. Background Art

[0002] Imaging small structures requires optical systems with high numerical aperture (NA). Digital imaging and fast computer systems offer the opportunity to generate high-resolution images by analytically merging low-resolution sub-images.

[0003] These computational methods of microscopic imaging include laminography, which generates images by processing coherent interference patterns scattered from a sample. The sample or wave field moves relative to a constant function such as the illumination field or the aperture stop. Fourier laminography systems are advantageous for having a large field of view and high resolution. A large number of individual images are illuminated by different numerical aperture settings. Quasi-collimated and therefore (spatially) coherent illumination is used at large angles of incidence. From the perspective of the imaging lens, this corresponds to an illumination system with a single asymmetric subset at a high numerical aperture setting.

[0004] Conventional Fourier stack imaging systems use a planar two-dimensional array of light emitting devices (LEDs). The light sources are arranged with the same spacing in the x and y directions, that is, the distance between adjacent LEDs in the x and y directions remains constant. One light source is turned on at a time and the corresponding low-resolution image is captured. The position of the light source relative to the sample determines the measured portion of the Fourier spectrum. For Fourier stack imaging reconstruction, it is required to collect Fourier spectra with at least 50% overlap.

[0005] Conventional uniform LED arrays have the effect that due to high angles of incidence a geometric reduction of the available power density at the sample plane becomes relevant. Said reduction is proportional to the cosine of the angle of incidence and has a significant effect at high angles of incidence.

[0006] Another effect is that light sources far from the central optical axis through the sample provide redundant information. As a possible solution, non-uniform sampling is proposed in Guo et al., "Optimization of sampling pattern and the design of Fourierptychographic illuminator", Opt. Express 23, 6171-6180, 2015.

[0007] Another approach involves a regular arrangement in polar coordinates as proposed in US 2017 / 0371141 A1.

[0008] In addition, it is known to increase throughput, i.e., reduce the number of images required for reconstruction, by capturing low-resolution images using multiple light sources simultaneously. For example, US 2017 / 0146788 A1 relates to Fourier ptychography microscopy with multiplexed illumination. US 2016 / 0088205 A1 relates to multiplexed Fourier ptychography systems and methods. Another approach is known from Tian et al., "Computational illumination for high-speed in vitro Fourier ptychographic microscopy", Optica 2(10), 904-911, 2015.

[0009] Another method to increase the throughput is content adaptive illumination, as known from Bian et al., "Content adaptive illumination for Fourier ptychography", Opt. Lett. 39, 6648-6651, 2014. According to this method, the relevant Fourier components are identified and the corresponding light source is selected for scanning. By using only a subset of the images, the number of required images can be reduced to a certain extent.

[0010] Because Fourier stacking systems need to generate multiple images at consecutive time points, there is always a need to further reduce the time required to generate all the sub-images combined using the stitching algorithm. At the same time, the image quality must remain high, i.e., the reduction in the time to generate the sub-images should not result in a significant degradation in image quality.

[0011] It is therefore an object of the present invention to provide a staggered imaging system and a method of generating images using the staggered imaging system, which staggered imaging system generates high quality images in a relatively short time. Summary of the invention

[0012] This object is solved by the subject-matter of the independent claims. The invention provides a stylus imaging system as recited in claim 1 and a method of generating images using a stylus imaging system as recited in claim 15. Advantageous embodiments are set forth in the dependent claims.

[0013] Thus, according to a first aspect, the present invention provides a stacked imaging system having a plurality of light sources and a controller. The light sources emit light onto a sample location. The light sources are arranged in a predetermined pattern. The controller controls the operation of the plurality of light sources. The predetermined pattern of the light sources is selected so as to at least partially compensate for geometric effects due to the arrangement of the light sources relative to the sample location that would otherwise occur in a planar uniform arrangement. Additionally or alternatively, the control operation of the light sources is adapted to compensate for geometric effects due to the arrangement of the light sources relative to the sample location.

[0014] According to a second aspect, there is provided a method for generating an image using a stacked imaging system, comprising a plurality of light sources arranged in a predetermined pattern. The plurality of light sources emit light onto a sample position including a sample. A controller controls the operation of the plurality of light sources. The predetermined pattern of light sources and / or the controlled operation of the plurality of light sources is adapted to compensate for geometric effects caused by the arrangement of the light sources relative to the sample position.

[0015] The optimized arrangement of light sources can increase the throughput of the stack imaging system compared to a conventional Fourier stack imaging system with uniform spacing and sequential acquisition. Compensation for geometric effects can include compensation for distance effects and angle effects, and can result in more constant and improved signal-to-noise performance across the entire angular spectrum. The overall noise characteristics in the reconstructed image will be improved. A preferred setting may be a constant noise and background level for all sub-images combined in each reconstruction.

[0016] An advantage of the present invention is that adaptive illumination and / or detection settings, such as the arrangement of light sources and / or light source brightness, provide an optimized signal-to-noise ratio in the sub-images. In addition, due to the optimized illumination and exposure control, an optimized image contrast in the sub-images can be achieved. The stacked imaging system provides higher variability and flexibility. In addition, different diffraction angles can be selectively weighted and contrasted. According to one embodiment, certain structures, in particular small structures and / or edges and / or transition areas can be enhanced, for example, by providing a higher contrast. This effect is particularly advantageous for taking pictures that are further processed by certain systems, for example using machine learning methods based on artificial intelligence or neural networks. These systems work better if the contrast is enhanced. In these cases, the natural appearance is less important or unimportant.

[0017] Preferably, larger diffraction angles will be enhanced relative to the central optical axis. Images of greater contrast may be particularly advantageous for applications such as laboratory diagnostics, such as hematology for samples that can be imaged multiple times, such as fixed cells on a slide. The present invention provides high resolution imaging with a high working distance and a large depth of field for reconstructed images typically used in stacked imaging systems.

[0018] As used in the present invention, the geometric effects to be compensated may include at least one of the following:

[0019] a) Angle-dependent spatial emission characteristics of the light source, such as the angular emission characteristics of the profile based on the geometric arrangement of the light source relative to the sample position.

[0020] b) The cosine effect, i.e. the proportional relationship between the available power density at the sample position and the cosine of the angle of incidence. The angle of incidence or azimuth of the light source is measured between the central optical axis and the line from the light source to the sample position, i.e. the optical axis of the particular light source. For a planar arrangement of the light sources, the central optical axis of the light source arrangement is perpendicular to the plane and passes through the sample position. The arrangement of the light sources can be symmetrical with respect to the central optical axis, for example in the form of a spherical cap. In this case, the arrangement of the light sources selects the central optical axis.

[0021] c) The drop in power level caused by the light source illuminating the investigated target area is kept constant at least at a point on the optical axis by appropriate control of the light source and / or by its geometric arrangement close to a spherical arrangement of the light source.

[0022] d) Higher diffraction angles result in lower diffraction efficiency. Compensation for the lower diffraction efficiency can result in better resolution and / or higher contrast of small object features in the corresponding reconstructed image. In contrast to high numerical aperture microscopy, ptychography offers the opportunity to control this effect more effectively than illumination systems with flat illumination profiles used in standard microscopy with Kohler-type illumination.

[0023] As used in the present invention, a sub-image refers to an image generated using a specific light source or using a specific selection or subset of light sources of multiple light sources.

[0024] According to another embodiment of the stacked imaging system, the predetermined pattern of the plurality of light sources comprises the light sources being arranged in a plurality of concentric rings around a central optical axis passing through the sample location. The variation in azimuth angle is substantially consistent for the light sources in subsequent rings. The variation in polar angle is substantially consistent for adjacent light sources within the same concentric ring. According to this embodiment, the stacked imaging system comprises a uniform theta-phi illuminator, theta being the polar angle and phi being the azimuth angle. This particular arrangement of light sources eliminates any variable redundancy in the measurement compared to conventional illumination methods. Each subsequent light source provides a consistent variation in angular spacing in radial and azimuth directions. Preferably, the variation in azimuth angle and the variation in polar angle are selected to provide a consistent 50% overlap for all light sources throughout Fourier space. The arrangement reduces the number of images required for reconstruction, resulting in a higher throughput. The reconstructed image is free of artifacts generated by the regular grid arrangement of the light sources.

[0025] According to another embodiment of the stylized imaging system, the arrangement of the light sources is planar.

[0026] According to another embodiment of the stacked imaging system, the arrangement of the light sources is spherical. According to this arrangement, the distance between the light source and the sample position is preferably constant. This arrangement has the additional advantage of providing a high dynamic range image for light sources located at the edge. This makes the signal-to-noise ratio of the reconstructed image better. This arrangement can correct for geometric effects caused by directionality and the distance between the light source and the sample.

[0027] According to another embodiment of the stacked imaging system, the controller controls the operation of the plurality of light sources by operating the plurality of light sources simultaneously. The number of light sources to be operated simultaneously is limited by a given maximum number and / or minimum distance criterion in the spatial or angular coordinate space of the light sources. For example, if the stacked imaging system comprises a plurality of N light sources, then if a plurality of M light sources are used simultaneously, the number of images required is reduced from N to N / M. This results in an M-fold increase in the throughput.

[0028] Multiplexing, i.e. operating multiple light sources to reduce the number of frames required, can be advantageous as long as relevant information is not lost, i.e. if different diffraction patterns in the Fourier plane can still be distinguished. That is, it should be possible to identify structural areas in Fourier space and to limit these areas relative to each other to assign these areas to the corresponding light sources. Thus, multiple diffraction patterns can be separated and extracted from a single frame. In order to ensure separability, the signal or structure in Fourier space must decay quickly enough so that different signals or structures do not merge above the noise level. This can be achieved by a minimum angular distance, i.e. by a minimum distance criterion in the coordinate space of the angle. The minimum distance criterion can be selected based on the structural size of the object to be analyzed. Multiplexing is particularly advantageous for weakly diffractive structures with rapidly decaying amplitude distributions in Fourier space. Preferably, the distance is maximized, taking into account that diametrically opposed light sources may have overlapping diffractions.

[0029] According to another embodiment of the stack imaging system, the controller controls the operation of the plurality of light sources by selecting a subset of the light sources in the plurality of light sources and operating only the light sources in the subset. In Fourier stack imaging, instead of scanning the sample in the spatial domain, the scan is performed in the Fourier domain. Therefore, it is determined which Fourier components are present and the corresponding optical devices are identified. The sample is scanned using only the identified optical devices. Because all sub-images are generated in a sequential order, using only a subset of light sources significantly reduces the time required to generate a high-resolution image.

[0030] According to another embodiment of the stacked imaging system, a subset of light sources to be operated is selected based on the characteristics of the sample to be observed. It can be known that the sample only contains certain elements or structures. Which Fourier component is dominant depends on the size and material properties of the element or structure. Therefore, only relevant Fourier components can be selected based on the characteristics of the sample, that is, only relevant light sources can be selected. For example, if the sample is a blood sample, the controller can select the light source that is most relevant to the blood sample.

[0031] According to another embodiment of the stacked imaging system, a subset of light sources to be operated is selected based on user input. The user can choose between different types of samples. The light sources of each sample can be stored in a memory of the controller. The controller selects the subset of light sources accordingly. In order to determine the subset, the density of the Fourier spectrum can be determined. If the Fourier spectrum density is small or there is no Fourier data in some areas, the corresponding light source will not be used for measurement. Therefore, the stacked imaging system realizes content adaptive lighting.

[0032] According to another embodiment of the stacked imaging system, the subset of light sources to be operated is selected based on a previous calibration. The calibration includes generating a calibration image of the sample using all or most of the light sources. Furthermore, the subset of light sources is selected based on the contribution of the light sources to the calibration image.

[0033] According to another embodiment of the stacked imaging system, when capturing a set of sub-images, a subset of light sources to be operated is dynamically selected based on an evaluation of the signal content in Fourier space of sub-images already captured in the set of sub-images, by determining the area and / or region and / or direction of the substantial signal contribution relative to a quality criterion and / or based on signal intensity, and by selecting light sources for subsequent images that partially overlap or are adjacent to the area and / or region and / or direction of the substantial signal contribution already measured.

[0034] According to another embodiment of the stacked imaging system, sub-images are captured by operating multiple light sources in parallel, by assigning substantial content in Fourier space to the operated individual light sources before selecting the light sources to be operated in subsequent frames and determining which of these light sources can be operated simultaneously subject to a given maximum number and / or minimum distance criterion in the spatial or angular coordinate space of the light sources.

[0035] According to another embodiment of the stylus imaging system, the controller controls the operation of the plurality of light sources by adjusting illumination parameters of the light sources according to their positions within the arrangement of light sources. Controlling the illumination parameters may be based on the distance of the light sources to the sample position and / or the tilt angle of the light sources.

[0036] According to another embodiment of the stacked imaging system, adjusting the illumination parameters comprises adjusting at least one of:

[0037] a) the brightness of the light source,

[0038] b) the duration of operation of the light source,

[0039] c) Attenuation filter,

[0040] d) color filters,

[0041] e) the exposure time of the detector of the stack imaging system, and

[0042] f) Gain setting of the detector of the stack imaging system.

[0043] The illumination parameters can be adjusted to a given signal-to-noise ratio of the detector of the stack imaging system. This means controlling the effective light flux, i.e. the amount of light, to the sample and / or from the sample to the detector. For example, the controller can control the drive current amplitude, the on-time of the drive current, the setting of filters in the beam path; or an exposure control of the detector can be performed by controlling the gain and / or the exposure time. Preferably, the illumination parameters, e.g. at least one of the parameters a) to e), are controlled to achieve a given signal-to-noise ratio. Preferably, for each illumination angle, i.e. for each light source, a specific setting of the illumination parameters is determined.

[0044] According to another embodiment of the stylus imaging system, at least some of the control may be achieved by gain control, ie control of the sensitivity of a detector of the stylus imaging system.

[0045] According to another embodiment of the stylized imaging system, the controller adjusts the illumination parameters according to the azimuth angle of the light source.

[0046] The power density is proportional to the cosine of the incident angle (i.e. the azimuth of the light source). This means that a light source for a high angle requires more power. Therefore, the controller can adjust the lighting parameters depending on the azimuth of the light source, in particular using the inverse of the cosine of the azimuth of the light source to compensate for this angle effect.

[0047] According to another embodiment of the stacked imaging system, the sub-images generated by the stacked imaging system are rescaled to a predetermined set of standard settings for the illumination parameters. The rescaling takes into account the respective scaling behavior of each illumination parameter, for example, linear scaling of exposure time, non-linear scaling of LED driver current of luminous flux, etc.

[0048] According to another embodiment of the stacked imaging system, compensation is used in an adaptive manner according to a predetermined pattern of light sources. That is, the compensation depends on the geometrical setting of the light sources. For example, the light sources may be arranged in a plane with parallel optical axes or with a single tilted optical axis for at least one light source, wherein the optical axis is ideally directed to the center of the field of view of the sample, i.e., the sample position. This arrangement reduces the amount of geometric correction of the geometric emission profile of the corresponding light source. In addition, the light sources may be arranged in a spherical geometry to minimize the effects caused by the distance from the light source to the sample position. In addition, at least one of the light sources may be used and moved from one image capture to a different position of the next image capture so as to cover the entire aperture range of the desired image setting. This movement may be operated directly by physically moving the distal output end of each light source such as an LED or an optical fiber; or indirectly by redirecting the light beam from the corresponding light source by, for example, a scanning unit such as a galvanometer scanner. Based on the respective geometrical settings of the light sources relative to the correct position, the controller determines and uses an adaptive setting for correction.

[0049] According to another embodiment of the stylus imaging system, a controller controls the operation of the multiple light sources to obtain different intensities at different diffraction angles (ie, illumination apertures), thereby inherently providing contrast enhancement during the reconstruction process.

[0050] According to another embodiment of the method, controlling the operation of the plurality of light sources comprises the step of operating the plurality of light sources simultaneously. The number of light sources to be operated simultaneously is limited by a given maximum number and / or minimum distance criterion in the spatial or angular coordinate space of the light sources. For example, a minimum distance in theta-phi space between an operating light source and its adjacent operating light sources must be met.

[0051] According to another embodiment of the method, controlling the operation of the plurality of light sources comprises selecting a subset of light sources in the plurality of light sources and operating only the light sources in the subset. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The invention will be explained in more detail with reference to exemplary embodiments depicted in the drawings.

[0053] The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments of the invention and together with the description serve to explain the principles of the invention.

[0054] Other embodiments of the present invention and many of the expected advantages of the present invention will become better understood by reference to the following detailed description. The same reference numerals represent corresponding similar parts. It should be understood that the method steps are numbered for easier reference, but unless otherwise explicitly or implicitly described, the numbering does not necessarily imply that the steps are performed in that order. In particular, the steps may also be performed in an order different from the order indicated by their numbers. Some steps may be performed simultaneously or in an overlapping manner.

[0055] Figure 1 Schematically showing a block diagram illustrating a stacked imaging system according to an embodiment of the present invention;

[0056] Figure 2 A stacked imaging system according to an embodiment of the present invention is schematically shown;

[0057] Figure 3 Show the azimuth and spacing of the light source arrangement;

[0058] Figure 4 shows the polar angle of the light source;

[0059] Figure 5 shows the illumination numerical aperture of a planar array of light sources as a function of position along the x-axis;

[0060] Figure 6 is a top view of an arrangement of light sources according to an embodiment of the present invention;

[0061] Figure 7 The side view schematically shows Figure 6 The arrangement of light sources;

[0062] Figure 8 schematically showing a test object used for simulation;

[0063] Fig. 9 A low-resolution image showing a simulation for conventional grid-based lighting;

[0064] Fig.10 shows a reconstructed high-resolution image for conventional grid-based lighting;

[0065] Fig.11 shows a reconstructed high-resolution image obtained by a stacked imaging system according to an embodiment of the present invention;

[0066] Fig.12 showing an exemplary Fourier spectrum of a test object;

[0067] Fig.13 An exemplary subset of light sources for stack imaging is shown;

[0068] Fig.14shows a reconstructed high-resolution image obtained by a stacked imaging system according to an embodiment of the present invention;

[0069] Fig.15 shows a reconstructed high-resolution image obtained by a stacked imaging system according to another embodiment of the present invention;

[0070] Fig.16 shows a reconstructed high-resolution image obtained by a stacked imaging system according to yet another embodiment of the present invention; and

[0071] Fig.17 A flow chart showing a method of generating an image using a stacked imaging system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0072] Figure 1 A block diagram illustrating a staggered imaging system 1 is shown. The staggered imaging system 1 comprises a plurality of light sources 2, in particular light emitting devices (LEDs), which emit light onto a sample arranged at a sample position of the staggered imaging system 1. The light sources 2 are arranged in a predetermined pattern. The overall shape of the arrangement may comprise a planar array or a spherical cap.

[0073] The light sources 2 may be arranged in a plurality of concentric rings around a central optical axis. All light sources 2 within the same concentric ring have the same azimuth angle measured between the central optical axis and a line from the light source 2 to the sample position. The azimuth angles of the light sources 2 may exhibit a constant variation between adjacent concentric rings. In other words, the differences between the azimuth angles of the light sources 2 in different adjacent concentric rings are independent of the concentric ring under consideration.

[0074] Likewise, the variation of the polar angle may be consistent for every two light sources 2 selected within the same concentric ring. In other words, the difference between the polar angles of two adjacent light sources in the same concentric ring is independent of the selected light sources 2. In particular, the light sources 2 in different concentric rings may have substantially the same variation of the polar angle, except for the deviation due to the fact that there may only be an integer number of light sources 2 in each concentric ring.

[0075] The controller 3 of the stacked imaging system 1 controls the operation of the light source 2. The controller 3 may include at least one of a central processing unit (CPU) or a graphics processing unit (GPU) (such as a microcontroller (μC)), an integrated circuit (IC), an application-specific integrated circuit (ASIC), an application-specific standard product (ASSP), a digital signal processor (DSP), a field programmable gate array (FPGA), etc. The controller 3 may also include or access a memory. The memory may be a volatile or non-volatile data storage, such as a solid state disk, a memory card, etc. In a preferred embodiment, the controller may store a sequence of previously defined subsets of selected LEDs to be operated at a specific time, and call these subsets of the sequence during the capture of a sub-image of the entire sub-image set.

[0076] The controller 3 activates at least one light source 2 at a time to generate a corresponding sub-image. The controller 3 may simultaneously operate a subset of the plurality of light sources 2. In particular, for each sub-image, the controller 3 may select at least two light sources 2 to operate. The light sources 2 may be selected according to a predetermined pattern.

[0077] The controller 3 may also select a subset of the plurality of light sources 2. Only the selected light sources 2 are operated to generate the corresponding sub-images. One or more light sources 2 may be selected to generate each sub-image. The unselected light sources 2 are not used to generate the sub-images. The controller 3 may select the light sources 2 according to instructions stored in the memory.

[0078] The controller 3 can select the subset to be operated according to the characteristics of the sample to be observed. The stack imaging system 1 may include a user interface for receiving instructions from the user. The user can select one or more options about the sample type. For each type of sample, the controller 3 selects the corresponding subset of the light source 2.

[0079] The controller 3 may also select the light source 2 based on a previous calibration. During calibration, the controller 3 generates a low-resolution sub-image. To generate the sub-image, all light sources 2 are used. In other words, no selection is made during the calibration phase. The controller 3 may then generate a high-resolution image by combining the low-resolution sub-images. The high-resolution image is considered a calibration image. Next, the controller 3 may analyze the high-resolution image using Fourier analysis to determine the most contributing Fourier components. For example, only those components whose average value exceeds a certain threshold may be considered "most contributing" or relevant. A specific light source 2 among the multiple light sources 2 corresponds to a corresponding Fourier component. During operation, the controller 3 selects only the light source 2 corresponding to the Fourier component with the largest contribution identified during the calibration phase.

[0080] The controller 3 may be adapted to control illumination parameters of the light source 2. The value of the illumination parameter may depend on the position of the light source 2 within the arrangement of light sources 2. In particular, the illumination parameter may be selected solely in dependence on the azimuth of the light source 2. In other words, the same illumination parameters are selected for light sources 2 with the same azimuth. For light sources 2 with different azimuths, at least some of the illumination parameters are different. The illumination parameters may include the brightness of the light source 2, the operation duration of the light source 2, an attenuation filter, a color filter, a gain value of a detector of the stylized imaging system 1 and / or an exposure time.

[0081] Figure 2An exemplary stacked imaging system 1 is shown. The stacked imaging system 1 comprises an array of light sources 2. The light sources 2 emit (laser) light onto a sample plane 4 having samples located at sample positions 5. The sample positions are spread out. However, since the size of the sample positions is small relative to the size of the distances between adjacent light sources 2, the sample positions can be considered to be point-like in a good approximation. The sample may comprise a solid structure or a fluid, such as a sample of blood.

[0082] The light emitted by the light source 2 interacts with the sample, for example via absorption, deflection or reflection. At least a portion of the light emitted by the light source 2 passes through the sample at the sample position 5 and enters the microscope objective 6. The imaging lens 7 focuses the light onto a real plane imaging detector 8. The detector 8 generates a sub-image and stores the data in a memory, for example, of the controller 3. The detector 8 can also provide the data to an external processing unit, such as an external computer for analyzing the data.

[0083] although Figure 2 A uniform array of light sources 2 is shown, but the array of light sources 2 may generally have different arrangements, such as non-uniform spacing between the light sources 2. In particular, the arrangement may include consistent variations in azimuthal and / or polar angles. Furthermore, different arrangements may also include that different types of light sources may be mounted at different positions within the stack illuminator. The different types of light sources may be selected from, but not limited to, the following types: monochromatic or single-bandwidth emitting or multi-bandwidth emitting light emitting devices (LEDs), such as RGB LEDs, S-LEDs, lasers (especially semiconductor lasers), thermal emitters, fiber-based light sources, etc. The light sources may differ in at least one of the following characteristics, for example: wavelength, spectral bandwidth, spatial emission characteristics, temporal emission characteristics such as continuous or pulsed operation, coherence parameters such as temporal and / or spatial coherence degree, brightness or etendue.

[0084] Figure 3 The azimuth angle of the arrangement of the light source 2 is shown and spacing x. Azimuth is the angle measured between a central optical axis passing through the central light source 21 and extending to the sample position 5 and a line extending between the respective light source 22 and the sample position 5. The spacing x corresponds to the distance between two adjacent light sources 21,22.

[0085] Figure 4 The polar angle θ of the light source 2 is shown. The polar angle θ of a particular light source 22 corresponds to the angle between the line between that light source 22 and the central light source 21 and another line between a predetermined light source 23 and the light source 21. The predetermined light source 23 is arbitrarily chosen as a reference. What is important is the variation of the polar angle, i.e. the difference between the polar angles of two adjacent light sources 22, 23 within the same concentric circle, which is independent of the reference.

[0086] Figure 5The numerical aperture A of illumination for a conventional uniform planar array of light sources 2 is shown depending on the position B along the x-axis. Each light source illumination provides separate measurements in Fourier space. Such multiple separate measurements are used for reconstruction of high resolution, high field of view images. For accurate reconstruction of the image, it is necessary to have at least 50% overlap between the Fourier spectra measured by adjacent light emitting devices 2. In the center, the overlap between adjacent light emitting devices 2 can be about 55%. From Figure 5 It can be seen that the illumination numerical aperture does not increase linearly. This causes an increase in the overlap between the Fourier spectra of adjacent light sources 2 towards the periphery. In order to solve this problem, a specific arrangement of the light sources 2 can be chosen.

[0087] Figure 6 The arrangement of the light sources 2 is schematically shown in a top view.

[0088] Figure 7 Schematically showing the Figure 6 A side view of the arrangement of the light source 2. The light sources 2 are arranged in concentric circles around the central optical axis. Azimuth The variation of is chosen to be uniform and the variation of the polar angle θ between the light sources 2 in the same concentric circle is also chosen to be substantially uniform. In other words, the stacked imaging system 1 comprises a uniform theta-phi illuminator.

[0089] Preferably, the angular orientation of each concentric ring is statistically or randomly selected. In other words, additional symmetries are preferably avoided to prevent artifacts. For example, such artifacts may appear if there is a great circle (or meridian) on a sphere including the light sources 2 for each concentric ring. Preferably, the light sources 2 are not arranged along a meridian.

[0090] although Figure 6 and Figure 7 A spherical uniform theta-phi illuminator is shown, but the light source 2 can also be arranged in a plane, thereby providing a planar uniform theta-phi illuminator. In the planar case, additional symmetries are preferably also avoided. Therefore, the LEDs of the different rings are not arranged on straight radial lines corresponding to the meridians of the spherical shell.

[0091] To determine this arrangement, first, the polar angles can be determined so that in the radial direction, the continuous light sources 2 have approximately 50% overlap in the Fourier spectrum. These polar angles and the expected maximum numerical aperture are used to find the number of rings of light sources 2. Each ring will be arranged in a consistent polar angle step to ensure that there is consistent overlap in the radial direction.

[0092] For example, the ideal spacing for a numerical aperture of 0.1 may be about 5 mm. Most high power LEDs have a footprint of about 2 to 3 mm. If the height (i.e. the distance between the central light source 21 and the sample position 5) is chosen to be 70 mm, there is about 55% overlap in the Fourier spectrum. From the spacing x and the height, the azimuth angle can be calculated according to the following formula:

[0093]

[0094] Next, the position of the light source 2 along the axis is determined based on the synthetic aperture requirement. For example, if the synthetic aperture needs to be 0.6, the illumination numerical aperture needs to be at least about 0.5, using the numerical aperture of the lens of 0.1 by diffraction. The maximum azimuth angle can be calculated according to the following formula:

[0095]

[0096] In the following, we will use The radius of each circle can be calculated using the following formula:

[0097]

[0098] Where n=1, 2, ... The maximum radius depends on the required effective numerical aperture. In order to obtain an illumination numerical aperture of 0.5, a central light source 2 surrounded by 7 rings with additional light sources 2 is required.

[0099] The radius of the circle is given by the following value (in mm):

[0100] 4.8949,9.8379,14.8790,20.0722,25.4779,31.1660,37.2197.

[0101] Light sources 2 located at these radii will maintain consistent overlap in the radial direction.

[0102] Next, it is described how to find the azimuthal position of the light source 2 on each circle so that there is a consistent overlap along the azimuthal direction.

[0103] First, for the change of the polar angle θ between adjacent light sources 2, the number of light sources 2 in the first circle is calculated using the following formula:

[0104] Δθ=tan -1 (pitch / radius) = 45.6°.

[0105] Therefore, the first circle requires 360 / 45=8 light sources 2 .

[0106] Likewise, the azimuth changes are calculated for the corresponding circles to have the following values:

[0107] 45.6086,26.9414,18.5746,13.9877,11.1031,9.1144,7.6512

[0108] The corresponding number for light source 2 is calculated as

[0109] 7.8933,13.3623,19.3813,25.7368,

[0110] 32.4234,39.4981,47.0517,55.

[0111] In order not to violate the criterion of overlap in the pupil of at least 50%, rounding to the nearest higher integer gives the number M of light sources 2 per circle n:

[0112] 8,14,20,26,33,40,47,55.

[0113] Angular step length Δθ of loop n n Calculated by the following formula:

[0114] Δθ n =360 / M n

[0115] Among them, M n represents the corresponding number M of light sources in ring n.

[0116] The position of each light source 2 can be found using the angular coordinates and the radius using the following formula:

[0117] [x,y]=[r·cos(m·Δθ n ),r·sin(m·Δθ n )].

[0118] In this formula, r represents the radius of each light source 2, and m varies from 1 to the maximum number M of light sources 2 on the circle n. n According to a preferred embodiment for avoiding additional symmetries, for each ring n, a (possibly different) statistical angle Δθ may be added n0 , the statistical angle can be determined according to the following formula:

[0119] Δθ n0 =Rn·Δθ n

[0120] where Rn represents at least one random number in the range [0; 1] for all rings, or preferably a different random number from the range [0; 1] for each ring. n The value of is different, even a random number will help to avoid additional symmetries. The position of light source 2 will be found from the angular coordinates, the angular variation and the corresponding radius of ring n by the following formula:

[0121] [x,y]=[r·cos(m·Δθ n +Δθ n0 ),r·sin(m·Δθ n +Δθ n0 )]

[0122] For the design of a spherical uniform theta-phi array, a similar procedure is used to find the polar and azimuthal angles. The only difference compared to a planar array is that the coordinates of the light source are determined using the following equations:

[0123]

[0124]

[0125]

[0126] According to this formula, n varies from 1 to the maximum radius. Δθ n Indicates the azimuth step size in ring number n. m n Indicates the number of corresponding light sources 2 in ring number n, ranging from 1 to m n For the preferred embodiment, the statistical angle Δθ can be added accordingly n0 .

[0127] The light source in the arrangement described in the previous figures may comprise at least one of the following: a semiconductor based light source such as an LED, a SLED, a semiconductor laser such as a VCSEL, a gas laser, a white light laser such as a fiber crystal laser, a fluorescence based or stretched light source such as a white light LED, a thermal light source such as a halogen lamp, an arc lamp such as a Xe-Hg lamp, a fiber based light source, wherein any of the aforementioned light sources may be coupled into an optical fiber and the output end of the optical fiber may be used as the light source. Furthermore, these light sources may be coupled with beam forming optical elements such as collimators or focusing optics to optimize the amount of light radiated onto the sample area 5 under investigation. The collimator optics may be selected from, but not limited to, the following: a lens, a GRIN lens, a diffractive optical element (DOE), a refractive optical element such as a Fresnel lens, a computer generated hologram.

[0128] Figure 8 A test object for the simulated ground truth is schematically shown. For different stylus imaging systems, including a conventional stylus imaging system with a planar uniform array of light sources 2, images of the test object were simulated.

[0129] Fig. 9A simulated low-resolution image of such a grid-based conventional illumination with a 0.1 numerical aperture objective is shown. As can be observed, the quality is relatively poor and small structures are not resolved.

[0130] Fig.10 A reconstructed high-resolution image of a grid-based conventional illumination based on the stitching of multiple low-resolution images is shown. A total of 289 low-resolution images are combined to obtain Fig.10 High resolution image shown.

[0131] Fig.11 A reconstructed high-resolution image obtained by a stacked imaging system 1 with a uniform theta-phi illuminator according to the present invention is shown. Due to the specific arrangement, for the same numerical aperture as used by a conventional imaging system, only a smaller number of 244 images are required to reconstruct a high-resolution image.

[0132] In the following, a stylized imaging system 1 is described in which the controller 3 operates the plurality of light sources 2 by selecting a subset of the light sources 2 in the plurality of light sources 2 and by operating only the selected light sources 2 in the subset.

[0133] Fig.12 Show Figure 8 An exemplary Fourier spectrum of a test object is shown. It can be seen that the test object has significant Fourier components only along the x-axis and the y-axis. Therefore, the number of light sources 2 to be operated can be limited without losing important information.

[0134] Fig.13 An exemplary subset of light sources 2 for staggered imaging is shown. Only light sources 2 along the x-axis and along the y-axis are selected to generate respective low-resolution images.

[0135] Additionally or alternatively, multiple light sources 2 may be used simultaneously. For example, for each low-resolution image, two light sources 2 may be selected and turned on to capture the corresponding low-resolution image.

[0136] Fig.14 A reconstructed high-resolution image obtained by a stacked imaging system 1 with a uniform theta-phi illuminator according to the present invention is shown. A high-resolution image has been reconstructed from low-resolution images using multi-LED illumination with two light sources 2 for each low-resolution image. The high-resolution image has been obtained by merging a total of 122 low-resolution images in the stacked reconstruction process.

[0137] Fig.15 A reconstructed high-resolution image obtained by a stacked imaging system 1 with a uniform theta-phi illuminator according to the present invention is shown. To generate the low-resolution image, a content-adaptive selection of the light source 2 has been employed. Thus, only Fig.13The light sources 2 shown, ie only the light sources 2 along the x-axis and the y-axis, are used to generate the low-resolution image. The high-resolution image is obtained by stitching together a total of 31 low-resolution images.

[0138] Fig.16 1 shows a reconstructed high-resolution image obtained by a stacked imaging system 1 with a uniform theta-phi illuminator according to the present invention. In addition to the multi-LED illumination using two light sources 2 simultaneously, a content-adaptive selection of the light sources 2 is also employed. The content-adaptive selection can be based on a priori knowledge about the object. Thus, only Fig.13 The light sources 2 shown, i.e. only the light sources 2 along the x-axis and the y-axis, are used to generate the low-resolution image. Thus, the stacked imaging system 1 combines multi-LED illumination with content-adaptive selection of the light sources 2. The high-resolution image is obtained by reconstructing a total of 19 low-resolution images.

[0139] In general, for a given object, the orientation of the pattern of Fourier components in the pupil plane or in the Fourier transform of the image, respectively, strongly depends on the orientation or three-dimensional pose of the object in the object space of the imaging system.

[0140] exist Fig.16 In the embodiment, the axes of the test pattern are aligned with the x-axis and y-axis of the sensor array of the imaging system. In addition, the secondary grid of the light source is also aligned with respect to the x-axis and y-axis of the object, so that the Fourier pattern of the transformed image in the spatial frequency domain is aligned with the x-axis and y-axis of the detector, and therefore, the image frame and the light source to be selected are also selected to be aligned with the x-axis and y-axis of the grid of light source 2 that intersects on the optical axis of the imaging system.

[0141]

[0142] Table 1

[0143] Table 1 compares different stack imaging systems1. It can be seen that the use of uniform theta-phi illuminators, multi-LED illumination, and content adaptive illumination can each help reduce the number of images required for reconstruction. If all three measurements are used simultaneously, the number of images required for reconstruction is reduced from 289 to only 19. Therefore, the time required to generate the low-resolution images used to reconstruct the high-resolution image is significantly reduced.

[0144] Typically, if the light sources are to be selected based on a priori knowledge of the spatial frequency structure of the image of the object, it is necessary to first determine the spatial orientation or pose of the object or first determine the geometric principal axes of the coordinate system representing the object in the object space of the imaging system. In successive steps, a predetermined subset of the light sources 2 to be used in the imaging process is employed according to the angular orientation of their orientation or principal axes relative to the object when they are positioned in the object space of the imaging system.

[0145] Another alternative method for content-adaptive selection of light source 2 can be based on taking an image in a first step and on analyzing the Fourier components incorporated into the image in a second step. Based on the intensity of the contribution of the corresponding light source 2 to the signal when reconstructing the object, the light source 2 can be selected or not selected as part of a subset of the light sources 2 used in subsequent replications of the imaging of the same object or the same kind of object in the imaging system. In addition, it is important that when the light source 2 is selected to be adapted to the spatial frequency content of the object, the orientation of the measured object is the same as the orientation of the same or reference object. In any case of misalignment, the selection of the light source 2 as described in the previous section for the case of prior knowledge must be adopted.

[0146] The content-adaptive selection can also be based on iteratively checking the higher-order spatial frequency components in the object image. Preferably, the first image is taken with a low illumination aperture, such as axial illumination based on a central light source 2 or low-aperture illumination from a light source with a smaller number of rings n. When determining the spatial frequency content of the image or a subset of images with low-aperture illumination, preferably the light source in the next larger number n of rings is selected in those radial directions, where substantial Fourier components for lower spatial frequencies have been selected. Whether a radial direction has a substantial Fourier component can be judged based on an appropriate metric, such as a threshold of the absolute intensity of the Fourier component, or compared with the neighboring areas of the corresponding area, or relative to the average of other Fourier components with the same radial distance, or the expected background noise level for the radial distance. Since the method does not rely on reference information from a priori knowledge or taking images in the first step, no angle correction is required when selecting the light source 2 and applying them to consecutive images. On the other hand, the iterative inspection method may be beneficial for selecting the light source 2 on the first object and for consecutive measurements of comparable objects using the defined subset of light sources. Only when this subset of light sources 2 is applied to another object, the employed angles of the light sources 2 selected for the respective orientation of the second object may be applied.

[0147] Content adaptive selection can even be performed dynamically when taking images. At the beginning, a first image is captured with a low illumination aperture angle, preferably for on-axis illumination. The Fourier transform of the image is determined. In the Fourier transform, the area or region or direction of the peak sequence in the radial axis is determined, wherein the substantial signal contribution is determined. Due to the overlap criterion of the Fourier space selection of light sources 2, those adjacent light sources 2 whose corresponding illumination apertures overlap with the area, region or direction of the radial axis are selected. The next image is taken for the light source or multiple light sources, wherein the maximum substantial signal contribution is in the overlapping area of ​​the aperture space. When multiple light sources 2 are applied, the minimum distance criterion in the spatial or angular coordinate space of the light source 2 can be followed. When the next image is taken, an evaluation is performed similar to the first image.

[0148] Based on the determined area, region and / or direction of the radial axis of the two images, the next image is taken for the light source or light sources, where the maximum substantial signal contribution is in the overlapping area in the aperture space. This process is repeated for the other images until there is no more substantial signal contribution, or if all remaining areas, regions or directions of the radial axis do not have any substantial contribution for the additional light source 2 that has not been used so far. This situation can be considered as a termination criterion. After the image subset is completed or the termination criterion is reached, the overlay evaluation is applied on the image subset to determine the high-resolution image.

[0149] Fig.17 A flow chart showing a method of generating an image using a stylus imaging system 1. The method may be performed with any of the stylus imaging systems 1 described above.

[0150] The stack imaging system 1 includes a plurality of light sources 2 arranged in a predetermined pattern. The light sources 2 can be arranged in a plurality of concentric rings around a central optical axis of a sample position 5 passing through the stack imaging system 1. The changes in the azimuth angles of the light sources in the subsequent rings can be substantially consistent. In addition, the changes in the polar angles of adjacent light sources within the same concentric ring can be substantially consistent. The arrangement of the light sources 2 can be planar or spherical or any other geometric shape to support application-specific optimization. For example, in a planar arrangement, the light sources can be tilted relative to the plane according to their respective distances from the optical axis. Then, the setting provides a non-uniform grid of single light sources, each light source tilted at its respective angle relative to the optical axis, so that substantially all optical axes of all light sources intersect in a point or small area 5 in the sample plane. This arrangement helps to minimize the effects of spatially non-uniform emission characteristics of the light source 2, as typically seen in more directional emission profiles of, for example, LEDs, optical fibers, fiber bundles, lasers with slow and fast axes, or VCSELs.

[0151] In a first method step S1 , the controller 3 sends a control signal to the light source 2 in order to operate the light source 2 to emit (laser) light onto a sample at a sample position 5 of the stylus imaging system 1 .

[0152] The controller 3 may control the operation of the plurality of light sources 2 by simultaneously operating the plurality of light sources 2. In other words, each low-resolution image is generated by using the plurality of light sources 2. For example, each low-resolution image may be obtained by using two light sources 2.

[0153] The controller 3 may also control the operation of the plurality of light sources 2 by selecting a subset of the light sources 2 in the plurality of light sources 2 and by operating only the light sources 2 in the subset. For example, the controller 3 may select the light sources 2 only along the x-axis and the y-axis, such as Fig.13As shown. The selection of the light source can be based on the characteristics of the sample to be observed or measured. The selection can also be based on user input. For example, the user can choose between several types of samples, and the controller 3 selects the light source 2 corresponding to the selected type of sample. Which light sources 2 are selected for the corresponding types of samples can be stored in a lookup table in the memory of the stack imaging system 1. The controller 3 uses the lookup table to select the light source 2. The sample type specific lookup table can be provided to the user as a preset.

[0154] A subset of light sources 2 may also be selected based on a previous calibration. During calibration, a low-resolution image is generated using one or more light sources 2. All light sources 2, or at least a majority of light sources 2, are used to generate a corresponding low-resolution image. Finally, the low-resolution images are combined to generate a calibration image. The calibration image is analyzed, for example, using Fourier analysis to determine the most significant contribution of the light source 2. For example, only light sources 2 corresponding to Fourier components that contribute to the Fourier spectrum of the calibration image and exceed a predetermined threshold are selected. This selection of "useful" light sources may also be performed dynamically in real time during the image capture process of the low-resolution image to minimize and optimize image capture, and to skip illumination in areas adjacent to LEDs where no significant contribution in the Fourier component is measured, or where the corresponding intensity is very low relative to the background level.

[0155] The controller 3 may also control the operation of the plurality of light sources 2 by adjusting illumination parameters of the light sources 2 depending on their positions within the arrangement of light sources 2. In particular, the controller 3 may adjust the brightness of the light sources 2, the operation duration of the light sources 2, the attenuation filters, the color filters, the gain of the detector of the stacked imaging system 1 and / or the exposure time.

[0156] The illumination parameters may be adjusted to reduce or eliminate geometrical effects due to the arrangement of the light sources 2. For example, the illumination parameters may be adjusted depending on the azimuth angle of the light source 2, such as depending on the cosine of the azimuth angle of the light source 2.

[0157] In a further method step S2 , a plurality of low-resolution images generated by one or more light sources 2 are combined in order to generate a high-resolution image.

[0158] In a further method step, the high-resolution image can be further analyzed, for example in order to detect specific components or structures in the high-resolution image. This analysis can employ standard image analysis tools.

[0159] It will be appreciated that all advantageous options, modifications and variations described herein and the foregoing with respect to embodiments of the stack imaging system according to the first aspect may equally apply to embodiments of the method according to the second aspect, and vice versa.

[0160] In the foregoing detailed description, for the purpose of simplifying the present disclosure, various features are combined together in one or more examples. It should be understood that the above description is illustrative rather than restrictive. It is intended to cover substitutions, modifications and equivalents. After reading the above description, many other embodiments will be apparent to those skilled in the art.

[0161] Reference numerals list

[0162] 1. Stacked Imaging System

[0163] 2 Light Source

[0164] 3 Controller

[0165] 4 Sample plane

[0166] 5 Sample location

[0167] 6 Microscope objective

[0168] 7 Imaging lens

[0169] 8 Imaging Detector

[0170] 21-23 Light Source

[0171] A Illumination numerical aperture

[0172] B Position

[0173] x Spacing

[0174] θ Polar angle

[0175] Azimuth

[0176] S1 First method step

[0177] S2 Second method step.

Claims

1. A stacked imaging system (1), comprising a plurality of light sources (2) adapted to emit light onto a sample location (5), wherein the light sources (2) are arranged in a predetermined pattern; and A controller (3) adapted to control the operation of the plurality of light sources (2); in, at least one of a) the predetermined pattern of the light sources (2) and b) the controlled operation of the plurality of light sources (2) is adapted to compensate for geometric effects caused by the arrangement of the light sources (2) relative to the sample position (5), wherein the controller (3) is adapted to control the operation of the plurality of light sources (2), the operation comprising the steps of selecting a subset of light sources (2) of the plurality of light sources (2) and operating only the light sources (2) in the subset, wherein the subset of light sources (2) to be operated is selected based on a previous calibration, wherein the calibration comprises the following steps: using all of the light sources (2) to generate a calibration image of a sample; and selecting the subset of light sources (2) based on their contribution to the calibration image, wherein during the calibration, one or more of the light sources are used to generate a low-resolution image, and the low-resolution images are combined to generate a calibration image, and the calibration image is analyzed using Fourier analysis to determine the most contributing Fourier components whose average value exceeds a threshold.

2. The stacked imaging system (1) according to claim 1, wherein: The predetermined pattern of the plurality of light sources (2) comprises the light sources (2) being arranged in a plurality of concentric rings around a central optical axis passing through the sample location (5); wherein the azimuth shift is consistent for the light sources (2) in subsequent rings, wherein the azimuth of the light source (2) is measured between the central optical axis and a line from the light source (2) to the sample position (5); and Among them, for adjacent light sources (2) in the same concentric ring, the offset of the polar angle is consistent.

3. The stacked imaging system (1) according to claim 1 or 2, wherein: The arrangement of the light sources (2) is planar.

4. The stacked imaging system (1) according to claim 1 or 2, wherein: The arrangement of the light sources (2) is spherical.

5. The stacked imaging system (1) according to claim 1 or 2, wherein: The controller (3) is adapted to control the operation of the plurality of light sources (2), the operation comprising the step of operating the plurality of light sources (2) simultaneously, wherein the number of light sources (2) to be operated simultaneously is limited by a given maximum number and / or minimum distance criterion in the spatial or angular coordinate space of the light sources (2).

6. The stacked imaging system (1) according to claim 1, wherein: The subset of light sources (2) to be operated is selected based on characteristics of the sample to be observed.

7. The stacked imaging system (1) according to claim 6, wherein: The subset of light sources (2) to be operated is selected based on user input.

8. The stacked imaging system (1) according to claim 1, wherein: When capturing a set of sub-images, based on an evaluation of the signal content in Fourier space of sub-images already captured in the set of sub-images, by determining the area and / or region and / or direction of the substantial signal contribution relative to a quality criterion and / or based on signal strength, the subset of light sources (2) to be operated is dynamically selected by selecting the light sources (2) of subsequent images that partially overlap or are adjacent to the area and / or region and / or direction of the substantial signal contribution already measured.

9. The stacked imaging system (1) according to claim 8, wherein: The sub-images are acquired by operating a plurality of light sources (2) in parallel, by assigning substantial contents in Fourier space to the respective light sources operated before selecting the light source (2) to be operated in the subsequent frame, and by determining which of these light sources (2) can be operated simultaneously subject to a given maximum number and / or minimum distance criterion in the spatial or angular coordinate space of the light sources (2).

10. The stacked imaging system (1) according to claim 1 or 2, wherein: The controller (3) is adapted to control the operation of the plurality of light sources (2), the operation comprising the step of adjusting lighting parameters of the light sources (2) depending on the position of the light sources (2) within the arrangement of the light sources (2).

11. The stacked imaging system (1) according to claim 10, wherein: Adjusting the illumination parameters comprises adjusting at least one of the following: the brightness of the light source (2), the operation duration of the light source (2), a dimmer, a color filter, a gain and an exposure time of a detector of the stacked imaging system (1).

12. The stacked imaging system (1) according to claim 10, wherein: The controller (3) is adapted to adjust the lighting parameters depending on the azimuth angle of the light source (2).

13. A method for generating an image using a staggered imaging system (1), the staggered imaging system comprising a plurality of light sources (2) arranged in a predetermined pattern, the method comprising: The plurality of light sources (2) emit light onto a sample position (5) including a sample; as well as The operation of the plurality of light sources (2) is controlled by a controller (3); wherein at least one of a) the predetermined pattern of the light sources (2) and b) the controlled operation of the plurality of light sources (2) is adapted to compensate for geometric effects caused by the arrangement of the light sources (2) relative to the sample position (5), wherein the controller (3) is adapted to control the operation of the plurality of light sources (2), the operation comprising the steps of selecting a subset of light sources (2) of the plurality of light sources (2) and operating only the light sources (2) in the subset, wherein the subset of light sources (2) to be operated is selected based on a previous calibration, wherein the calibration comprises the following steps: using all of the light sources (2) to generate a calibration image of a sample; and selecting the subset of light sources (2) based on their contribution to the calibration image, wherein during the calibration, one or more of the light sources are used to generate a low-resolution image, and the low-resolution images are combined to generate a calibration image, and the calibration image is analyzed using Fourier analysis to determine the most contributing Fourier components whose average value exceeds a threshold.

14. The method according to claim 13, wherein: Controlling the operation of the plurality of light sources (2) comprises the step of operating the plurality of light sources (2) simultaneously, wherein the number of light sources (2) to be operated simultaneously is limited by a given maximum number and / or minimum distance criterion in the spatial or angular coordinate space of the light sources (2).

Images