Dual-wedge plate modulated phase-shift detection autofocus apparatus and method

Abstract

This invention discloses a phase-shift detection autofocus device and method using dual-wedge plate modulation, relating to the field of image processing technology. The device includes a beam splitter, a teleconverter, dual wedge plates, a first tube mirror, a second tube mirror, a monochrome image sensor, and a color image sensor. The beam splitter divides the original sample optical path acquired by the microscope objective into an imaging optical path and a focusing optical path. The optical axis of the imaging optical path is aligned with the optical axis of the objective lens. The first tube mirror and the color image sensor are sequentially arranged along the imaging optical path. The teleconverter, dual wedge plates, second tube mirror, and monochrome image sensor are sequentially arranged along the focusing optical path. The two wedge plates in the dual wedge plates are equidistant from the optical axis of the focusing optical path. The dual wedge plates are used to modulate the light rays on the left and right halves of the focusing optical path, thereby providing a phase-shift detection autofocus device that combines focusing direction discrimination capability, high focusing accuracy, and a large focusing range.

Description

Technical Field

[0001] This invention belongs to the field of image processing technology, specifically relating to a phase shift detection autofocus device and method using dual wedge plate modulation. Background Technology

[0002] In the medical field, traditional pathological diagnosis typically relies on manual observation and analysis of tissue sections under a conventional optical microscope. Due to variations in tissue section morphology, pathologists need to manually refocus when observing different areas of the tissue section. This method has significant drawbacks due to the time constraints, accuracy limitations, and human factors involved in observation. Wholeslide imaging (WSI) systems have emerged to address this issue. These systems can scan the entire tissue section and digitize it, converting it into high-resolution whole-slide images that can be analyzed and processed by a computer. This makes the storage, transmission, archiving, sharing, and analysis of tissue sections more efficient and convenient. It has become an important tool for biomedical research and clinical diagnosis. A typical wholeslide imaging (WSI) system mainly consists of the following components: 1) a conventional microscope imaging system (light source, objective lens, and tube lens); 2) a sample stage with programmable displacement control; 3) a camera for image acquisition or calculating the sample's refocus position; and 4) management software for browsing, analyzing, and storing digital whole-slide images.

[0003] In WSI systems, autofocus technology is the most important factor affecting imaging speed and image quality. WSI systems control the displacement of the sample stage through programming and automatically find the focal point of the sample during scanning using autofocus technology, continuously acquiring hundreds of high-resolution images, which are then seamlessly stitched together to form a single high-resolution whole-slice digital image. Current autofocus technologies can be mainly divided into three categories: laser reflection-based autofocus, focus map-based autofocus, and image-based autofocus.

[0004] However, laser reflection-based autofocus technology struggles to handle morphological variations in tissue sections on slides, while focus-image-based autofocus is a non-real-time technique, resulting in long acquisition times for whole-section images. Therefore, image-based autofocus technology has garnered more attention due to its potential efficiency, but its own technical challenges still need to be overcome.

[0005] In image-based autofocus technology, phase-shift detection autofocus is an important branch:

[0006] Phase-shift detection autofocus technology based on monochromatic dual-aperture modulation places two pinholes on the Fourier surface of the objective lens in the autofocus optical path. This results in the image captured by the camera after modulation by the dual pinholes containing two overlapping sample images spaced a certain distance apart. Based on a pre-calibrated linear relationship between axial defocus displacement and image pixel offset, the corresponding axial defocus amount is inferred. This method can achieve real-time autofocus with only a single image. However, when the defocus displacement of the sample is equal in both positive and negative directions, the camera will display a ghost image with the same spatial translation, making it impossible to distinguish between positive and negative defocus directions. Furthermore, due to the presence of the dual-aperture on the conjugate Fourier surface, the sample information that the focusing camera can receive is significantly reduced, affecting the measurement of pixel offset distance.

[0007] Phase-shift detection autofocus methods based on pupil-segmented images physically separate the light beam using a wedge plate in the conjugate Fourier plane, causing additional lateral displacement in the image acquired by the focusing camera. The collimation position is obtained by pre-calibrating the quantitative relationship between the sample's axial defocus displacement and the image pixel offset. However, this phase-shift detection autofocus technique suffers from several drawbacks. The pixel offset in the overlapping images is small, and when the sample is at both positive and negative defocus positions, the camera may still display ghost images with identical spatial translations, making it impossible to distinguish between positive and negative defocus directions. Furthermore, the axial focusing range is relatively small.

[0008] Therefore, how to provide a phase-shift detection autofocus device that combines the ability to determine focus direction, high focus accuracy, and a large focus range has become an important issue. Summary of the Invention

[0009] To address the aforementioned problems in the prior art, the present invention provides a phase shift detection autofocus device and method based on dual wedge plate modulation.

[0010] The technical problem to be solved by this invention is achieved through the following technical solution:

[0011] In a first aspect, the present invention provides a phase shift detection autofocus device modulated by a double wedge plate, the phase shift detection autofocus device comprising a beam splitter, a teleconverter, a double wedge plate, a first tube lens, a second tube lens, a monochrome image sensor, and a color image sensor;

[0012] The beam splitter divides the original optical path of the sample acquired by the microscope objective into an imaging optical path and a focusing optical path; the optical axis of the imaging optical path is consistent with the optical axis of the objective.

[0013] The first tube lens and the color image sensor are arranged sequentially along the imaging optical path;

[0014] The teleconverter, the double wedge plate, the second tube lens, and the monochrome image sensor are arranged sequentially along the focusing optical path; the two wedge plates in the double wedge plate are equidistant from the optical axis of the focusing optical path; the double wedge plate is used to modulate the light rays on the left and right halves of the focusing optical path.

[0015] Optionally, the double wedge plate comprises two completely symmetrical wedge plates at the same angle.

[0016] Optionally, the teleconverter uses a 4 f system.

[0017] In a second aspect, the present invention provides a phase-shift detection autofocus method based on dual wedge plate modulation, the phase-shift detection autofocus method comprising:

[0018] Starting from the current field of view, control the monochrome image sensor to acquire an overlapping image, and extract the left half and right half of the overlapping image;

[0019] The relative pixel offset is determined based on the left half image and the right half image;

[0020] The axial defocus distance is calculated based on the relative pixel offset according to the predetermined linear equation between the sample axial defocus distance and the relative pixel offset.

[0021] After the electric displacement platform reaches the designated focus position according to the axial defocus distance, the focus image under the current field of view is acquired by the color image sensor.

[0022] The electric displacement platform is controlled to move to the next field of view and then returns to execute the step of controlling the monochrome image sensor to acquire an overlapping image until all in-focus images under all fields of view are acquired, thus completing the phase shift detection autofocus of the dual wedge plate modulation.

[0023] Optionally, the linear equation for the predetermined sample axial defocus distance and relative pixel offset includes:

[0024] ;

[0025] in, Indicates the axial defocus distance of the sample; This represents the first constant obtained by fitting a relation curve beforehand; Indicates relative pixel offset; This represents the second constant obtained by fitting a relation curve beforehand.

[0026] Optionally, determining the relative pixel offset based on the left half of the image and the right half of the image includes:

[0027] A cross-correlation image registration algorithm is used to calculate the horizontal pixel offset based on the left half image and the right half image;

[0028] Convert the horizontal pixel offset to a relative pixel offset.

[0029] Optionally, the step of employing a cross-correlation image registration algorithm to calculate the lateral pixel offset based on the left half image and the right half image includes:

[0030] A first frequency domain representation is obtained by performing a Fourier transform on the left half of the image, and a second frequency domain representation is obtained by performing a Fourier transform on the right half of the image.

[0031] The cross-power spectrum of the left half image and the right half image is calculated based on the first frequency domain representation and the second frequency domain representation;

[0032] Perform an inverse Fourier transform on the cross-power spectrum to obtain the cross-correlation function;

[0033] The horizontal pixel offset is determined by finding the horizontal coordinate of the main peak position in the cross-correlation function.

[0034] Optionally, the cross-power spectrum is:

[0035] ;

[0036] in, This represents the cross-power spectrum; Represents spatial frequency in the frequency domain; This represents the first frequency domain representation; This represents the complex conjugate of the second frequency domain representation; This represents the phase difference between the left half of the image and the right half of the image. It represents the imaginary unit.

[0037] Optionally, converting the lateral pixel offset into a relative pixel offset includes:

[0038] The total offset is calculated based on the horizontal pixel offset and the segmentation offset of the overlapping image;

[0039] Based on the total offset, the relative pixel offset is determined.

[0040] The present invention provides a phase shift detection autofocus device with dual wedge plate modulation, which introduces dual wedge plates to realize linear deflection of light rays on the left and right halves of the focusing optical path. By analyzing the pixel offset of the image overlap area determined by the dual wedge plates, the positive and negative directions of defocus can be accurately determined. At the same time, by introducing pixel deviation closely related to the defocus direction, the autofocus range of the imaging system is expanded.

[0041] Furthermore, the dual wedge modulation does not directly block sample information on the Fourier surface. Compared with the dual aperture modulation method, this invention can still achieve spatial optical path offset modulation while preserving the complete information of the image, thus improving the robustness of phase correlation calculation and focusing accuracy.

[0042] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the structure of a phase shift detection autofocus device with dual wedge plate modulation provided in an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the autofocus principle provided in an embodiment of the present invention;

[0045] Figure 3 This is a schematic flowchart of a phase shift detection autofocus method with dual wedge plate modulation provided in an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the overlapping image and the left and right halves of the image at the focus position provided in the embodiment of the present invention;

[0047] Figure 5 This is a schematic diagram comparing pixel offset at +35μm and -35μm defocus positions provided in an embodiment of the present invention;

[0048] Figure 6 This is a schematic diagram showing the relationship between the sample axial defocus distance and the relative pixel offset provided in the embodiments of the present invention. Detailed Implementation

[0049] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0050] To address the shortcomings of existing phase-shift detection autofocus technology, such as lack of focus direction discrimination capability, low focusing accuracy, and small focusing range, this invention provides a phase-shift detection autofocus device with dual wedge plate modulation. (See also...) Figure 1 , Figure 1 This is a schematic diagram of a phase-shift detection autofocus device modulated by a double wedge plate according to an embodiment of the present invention. The double wedge plate modulated phase-shift detection autofocus device includes a beam splitter, a teleconverter, double wedge plates, a first tube lens, a second tube lens, a monochrome image sensor, and a color image sensor.

[0051] In this embodiment of the invention, the beam splitter divides the original optical path collected by the objective lens of the microscope into an imaging optical path and a focusing optical path; the optical axis of the imaging ray is consistent with the optical axis of the objective lens.

[0052] Specifically, the original optical path consists of the light reflected or projected from the sample collected by the microscope's objective lens; the condenser lens located below the sample is used to focus the light emitted by the LED light source to ensure uniform illumination and minimize light energy loss. A beam splitter can separate the original optical path into an imaging path that propagates towards the color sensor and a focusing path that propagates towards the monochrome sensor.

[0053] By splitting the light beam in two with a beam splitter, creating two separate paths for imaging and focusing, the interference of focusing on image quality can be avoided.

[0054] In this embodiment of the invention, the optical axis of the imaging optical path is aligned with the optical axis of the objective lens.

[0055] In this embodiment of the invention, the first tube lens and the color image sensor are arranged sequentially along the imaging optical path.

[0056] The first tube lens can be used to focus the image of the optical system onto the surface of the color image sensor; the color image sensor is responsible for acquiring the focused image.

[0057] In this embodiment of the invention, the teleconverter, the double wedge plate, the second tube lens, and the monochrome image sensor are arranged sequentially along the focusing optical path.

[0058] The teleconverter's function is to delay the image from the Fourier plane of the microscope objective to outside the objective.

[0059] Specifically, the teleconverter uses 4 f System. 4 f The system is an optical system consisting of four basic elements, specifically two lenses and two Fourier transform planes.

[0060] In this embodiment of the invention, the two wedge plates in the double wedge plate are equidistant from the optical axis of the focusing optical path; the double wedge plate is used to modulate the light rays on the left and right halves of the focusing optical path.

[0061] Specifically, the double wedge plate consists of two completely symmetrical wedge plates at the same angle, located on the back focal plane of the second lens in the teleconverter. The double wedge plates are equidistant from the central optical axis, achieving linear deflection of the left and right halves of the light rays in the focusing optical path, thereby introducing pixel deviation in the image that is closely related to the defocus direction.

[0062] In this embodiment of the invention, the second tube lens is used to focus the image of the front-path optical system onto a monochrome image sensor. The monochrome image sensor is responsible for acquiring images phase-modulated by a double wedge for autofocus.

[0063] The phase shift detection autofocus device with dual wedge plate modulation provided in this invention is applicable to microscopic imaging equipment such as digital pathology slide scanners and has a wide range of applications.

[0064] In this embodiment of the invention, a double wedge plate is used to modulate the light rays on the left and right halves of the focusing optical path, introducing a pixel deviation closely related to the defocusing direction, causing the acquired image information to overlap at different locations. See also Figure 2 , Figure 2 This is a schematic diagram of the autofocus principle provided in an embodiment of the present invention. When the sample is in the near-focus position, the overlapping image acquired by the monochrome image sensor produces a moderate positive shift; when the sample is in the negative defocus position, the overlapping image produces a small positive shift; and when the sample is in the positive defocus position, the overlapping image produces a large positive shift. By pre-calibrating the linear relationship between the sample's axial defocus displacement and the image pixel shift, the axial defocus position of the sample can be determined by calculating the pixel shift between the overlapping images in real time, thus achieving autofocus.

[0065] In this embodiment of the invention, a double wedge plate is introduced into the phase shift detection autofocus device, which realizes the linear deflection of the light rays on the left and right halves of the focusing optical path. By analyzing the pixel offset of the image overlap area determined based on the double wedge plate, the positive and negative directions of defocus can be accurately determined. At the same time, by introducing a pixel deviation that is closely related to the defocus direction, the autofocus range of the imaging system is expanded.

[0066] Furthermore, the dual wedge modulation does not directly block sample information on the Fourier surface. Compared with the dual aperture modulation method, the embodiments of the present invention can still achieve spatial optical path offset modulation while preserving the complete information of the image, thereby improving the robustness of phase correlation calculation and focusing accuracy.

[0067] Based on the same inventive concept, this invention also provides a phase-shift detection autofocus method using dual-wedge plate modulation, see [link to relevant documentation]. Figure 3 , Figure 3 This is a flowchart illustrating a phase-shift detection autofocus method using dual-wedge modulation provided in an embodiment of the present invention, specifically including the following steps:

[0068] Step S301: Starting from the current field of view, control the monochrome image sensor to acquire an overlapping image and extract the left half image and the right half image of the overlapping image.

[0069] In this embodiment of the invention, starting from the current field of view, the monochrome image sensor is controlled to acquire a complete overlapping image, the pixel width of which is... The height is .

[0070] Specifically, the overlapping image is obtained by modulating the left and right halves of the field of view with a double wedge, causing the light rays to overlap and form an image on a monochrome image sensor.

[0071] In this embodiment of the invention, extracting the left and right halves of the overlapping image includes:

[0072] The acquired overlapping images are arranged along the horizontal center line ( Divide the image into two halves: the left half and the right half.

[0073] Step S302: Determine the relative pixel offset based on the left half image and the right half image.

[0074] In this embodiment of the invention, determining the relative pixel offset based on the left half image and the right half image includes:

[0075] A cross-correlation image registration algorithm is used to calculate the horizontal pixel offset based on the left and right half images;

[0076] Convert the horizontal pixel offset to a relative pixel offset.

[0077] In this embodiment of the invention, the left half of the image is used as the reference image and the right half of the image is used as the search area. By calculating the cross-correlation function of the two parts, the horizontal pixel offset that maximizes the correlation between the two parts is found.

[0078] In one implementation, a cross-correlation image registration algorithm is used to calculate the lateral pixel offset based on the left and right halves of the image, including:

[0079] The first frequency domain representation is obtained by performing a Fourier transform on the left half of the image, and the second frequency domain representation is obtained by performing a Fourier transform on the right half of the image.

[0080] The cross-power spectrum of the left and right halves of the image is calculated based on the first and second frequency domain representations.

[0081] Perform an inverse Fourier transform on the cross-power spectrum to obtain the cross-correlation function;

[0082] The horizontal pixel offset is determined by finding the x-coordinate of the main peak position in the cross-correlation function.

[0083] The specific process is as follows:

[0084] First, examine the left half of the image. and right half image Perform Fourier transforms on each side to obtain the first frequency domain representation of the left half of the image. The second frequency domain representation of the right half of the image :

[0085] ;

[0086] in, This indicates the Fourier transform operation; Indicates the spatial domain pixel position; It represents the spatial frequency in the frequency domain.

[0087] Then, based on the frequency domain representations of the left and right halves of the image, the cross-power spectrum of the left and right halves is calculated. :

[0088] ;

[0089] in, for The complex conjugate; This represents the phase difference between the left and right halves of the image. It represents the imaginary unit. It can be seen as being caused by Obtained by translation:

[0090] ;

[0091] in, express x The amount of translation in the direction; express y The amount of translation in the direction;

[0092] Then, in the frequency domain, their Fourier transforms satisfy the following:

[0093] ;

[0094] Substituting into the cross-power spectrum, we get:

[0095] ;

[0096] in, for The complex conjugate;

[0097] right Performing an inverse Fourier transform yields the initially estimated cross-correlation function. :

[0098] ;

[0099] in, This indicates the inverse Fourier transform operation; This represents a two-dimensional impulse function.

[0100] The obtained cross-correlation function It reaches its maximum value at its main peak position, and the relative pixel offset between images is uniquely determined by the main peak position. By finding By finding the horizontal coordinate of the main peak position in the image, we can obtain the horizontal pixel offset of the most correlated parts of the overlapping image, the left and right halves. :

[0101] ;

[0102] In this embodiment of the invention, converting the horizontal pixel offset to a relative pixel offset includes:

[0103] The total offset is calculated based on the horizontal pixel offset and the segmentation offset of the overlapping images;

[0104] The relative pixel offset is determined based on the total offset.

[0105] In this embodiment of the invention, the horizontal pixel offset is the starting position of the right half of the image relative to the left half of the image. The offset at that point, but the actual starting position of the right half of the image in the original image is... Therefore, the total offset in the overlapping images It consists of two parts. The first part is the cross-correlation offset, which is calculated by cross-correlation between the left and right halves of the image. The second part is: segmentation offset, which is half the width of the original image. .

[0106] Therefore, total offset as follows:

[0107] ;

[0108] Subtract the pixel offset of the overlapping image at the focus position from the total offset. This will give you the relative pixel offset. :

[0109] ;

[0110] in, This is the relative pixel offset at the 0μm position obtained during the pre-calibration process.

[0111] Step S303: Calculate the axial defocus distance based on the relative pixel offset according to the predetermined linear equation between the sample axial defocus distance and the relative pixel offset.

[0112] In this embodiment of the invention, the linear equation relating the sample axial defocus distance to the relative pixel offset includes:

[0113] ;

[0114] in, Indicates the axial defocus distance of the sample; Indicates relative pixel offset; This represents the first constant obtained by fitting a relation curve beforehand; This represents the second constant obtained by fitting a relation curve beforehand.

[0115] The following will explain the process of determining the linear equation between the sample axial defocus distance and the relative pixel offset provided in the embodiments of the present invention:

[0116] First, manually control the axial displacement platform to move the objective lens to the focusing position. The monochrome image sensor in the autofocus device acquires a complete overlapping image with a pixel width of [missing information]. W The height is H The acquired overlapping images are arranged along the horizontal center line. Divide the image into two halves, left and right, as shown in the diagram. Figure 4 , Figure 4 This is a schematic diagram of the overlapping image and the left and right half images at the focus position provided in the embodiment of the present invention.

[0117] After acquiring overlapping images at the collimation position using a monochrome image sensor, the objective lens was moved to the +35μm and -35μm defocus positions, respectively, and overlapping images were acquired again using the monochrome image sensor. 500 images with identical information on the left and right halves at the collimation position were selected respectively. A comparison was made between a 500-pixel area and pixel areas of the same size and position on the left and right halves of the screen at +35μm and -35μm out-of-focus locations. (See [reference]). Figure 5 , Figure 5 This is a schematic diagram comparing pixel offset at +35μm and -35μm defocus positions provided in an embodiment of the present invention.

[0118] Comparative analysis revealed that as the sample moves along the axial direction, the information on the left and right sides of the overlapping image exhibits a regular displacement. Specifically, in a negative defocus state, as the sample moves further away from the focal plane, the information on the left and right sides of the overlapping image moves inward synchronously; while in a positive defocus state, as the sample moves further away from the focal plane, the information on the left and right sides of the overlapping image moves outward synchronously. This displacement characteristic reveals the correspondence between axial defocus displacement and image pixel offset, providing a clear characteristic basis for determining the defocus state of overlapping images at different axial defocus distances.

[0119] Next, to further obtain the correspondence between the pixel offset distance of the overlapping images and the axial defocus displacement of the sample, a cross-correlation image registration algorithm is used to calculate the pixel offset between the left and right halves of the overlapping images at different axial defocus distances. Using the left half of the image as the reference image and the right half as the search region, the cross-correlation function of the two parts is calculated to find the lateral pixel offset that maximizes the correlation between the two parts. Then, the total offset is calculated based on the lateral pixel offset and the segmentation offset of the overlapping images. Subtracting the pixel offset of the overlapping image at the focus position from the total offset of the overlapping images at different axial defocus distances yields the relative pixel offset of the overlapping images at different axial defocus distances. The specific calculation process can be found in step S102 and will not be repeated here.

[0120] Using the above data, a scatter plot of the relative pixel offset versus axial defocus distance of the overlapping images was plotted. Subsequently, curve fitting was performed to obtain a linear equation relating the sample's axial defocus distance to the relative pixel offset. (See [link to relevant documentation]). Figure 6 , Figure 6 This is a schematic diagram showing the relationship between the sample axial defocus distance and the relative pixel offset provided in the embodiments of the present invention. The horizontal axis represents the relative pixel offset in pixels, and the vertical axis represents the defocus distance in μm.

[0121] Therefore, the axial defocus distance can be calculated using the relative pixel offset, based on a linear equation between the predetermined sample axial defocus distance and the relative pixel offset of the overlapping image.

[0122] Step S304: After the electric displacement platform reaches the designated focusing position according to the axial defocus distance, the focusing image under the current field of view is acquired by the color image sensor.

[0123] In this embodiment of the invention, based on the axial defocus distance calculated above, the electric displacement platform can be controlled to reach the designated focus position, so that the objective lens can accurately return to the focal plane. After the electric displacement platform stabilizes, the focus image under the current field of view is acquired by the color image sensor.

[0124] Step S305: Control the electric displacement platform to move to the next field of view, and return to execute the step of controlling the monochrome image sensor to acquire an overlapping image, until all in-focus images under all fields of view are acquired, and complete the phase shift detection autofocus of the dual wedge plate modulation.

[0125] In this embodiment of the invention, the operation of acquiring a focused image is performed in each field of view, thereby realizing phase shift detection autofocus with dual wedge plate modulation.

[0126] In this embodiment of the invention, a double wedge plate is introduced into the phase shift detection autofocus device, which realizes the linear deflection of the light rays on the left and right halves of the focusing optical path. By analyzing the pixel offset of the image overlap area determined based on the double wedge plate, the positive and negative directions of defocus can be accurately determined. At the same time, by introducing a pixel deviation that is closely related to the defocus direction, the autofocus range of the imaging system is expanded.

[0127] Furthermore, the dual wedge modulation does not directly block sample information on the Fourier surface. Compared with the dual aperture modulation method, the embodiments of the present invention can still achieve spatial optical path offset modulation while preserving the complete information of the image, thereby improving the robustness of phase correlation calculation and focusing accuracy.

[0128] The embodiments of the present invention have high measurement accuracy for minute displacements and low accuracy requirements for the electric displacement platform, making them particularly suitable for applications such as digital pathological slide scanning and high-throughput imaging. They also have good system versatility and industrialization potential.

[0129] It should be noted that the terms "first," "second," etc., are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention.

[0130] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0131] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings and the disclosure in carrying out the claimed invention. In the description of the invention, the word "comprising" does not exclude other components or steps, "a" or "an" does not exclude a plurality, and "a plurality" means two or more, unless otherwise explicitly specified. Furthermore, while different embodiments may describe certain measures, this does not mean that these measures cannot be combined to produce good results.

[0132] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0133] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0134] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A phase-shift detection autofocus device with dual wedge plate modulation, characterized in that, The phase-shift detection autofocus device includes a beam splitter, a teleconverter, a double wedge plate, a first tube lens, a second tube lens, a monochrome image sensor, and a color image sensor. The beam splitter divides the original optical path of the sample acquired by the microscope objective into an imaging optical path and a focusing optical path; the optical axis of the imaging optical path is consistent with the optical axis of the objective. The first tube lens and the color image sensor are arranged sequentially along the imaging optical path; The teleconverter, the double wedge plate, the second tube lens, and the monochrome image sensor are arranged sequentially along the focusing optical path; the two wedge plates in the double wedge plate are equidistant from the optical axis of the focusing optical path; The double wedge plate is used to modulate the light rays on the left and right halves of the focusing optical path; The monochrome image sensor is used to acquire an overlapping image, which is then segmented into a left half and a right half, equally divided along a horizontal center line, to determine relative pixel offsets. ， The axial defocus distance is calculated based on a linear equation relating the predetermined sample axial defocus distance to the relative pixel offset; the overlapping image is obtained by the left and right half-field rays overlapping and imaging on the monochrome image sensor due to modulation by the double wedge plate; The linear equation relating the axial defocus distance of the sample to the relative pixel offset includes: ； ； ； in, Indicates the axial defocus distance of the sample; Indicates relative pixel offset; This represents the first constant obtained by fitting a relation curve beforehand; This represents the second constant obtained by pre-fitting a relation curve; Indicates the segment offset; Indicates cross-correlation offset; Indicates the total offset; This represents the pixel offset of the overlapping image at the focus position.

2. The phase-shift detection autofocus device according to claim 1, characterized in that, The double wedge plate comprises two completely symmetrical wedge plates at the same angle.

3. The phase shift detection autofocus device according to claim 1, characterized in that, The teleconverter uses 4 f system.

4. A phase-shift detection autofocus method using dual-wedge plate modulation, characterized in that, The phase shift detection autofocus method includes: Starting from the current field of view, the monochrome image sensor is controlled to acquire an overlapping image, and the overlapping image is divided equally along the horizontal center line to obtain a left half image and a right half image; the overlapping image is obtained by the double wedge plate modulating the light rays of the left and right halves of the field of view to overlap and image on the monochrome image sensor; The relative pixel offset is determined based on the left half image and the right half image; The axial defocus distance is calculated based on the relative pixel offset according to the predetermined linear equation between the sample axial defocus distance and the relative pixel offset. After the electric displacement platform reaches the designated focus position according to the axial defocus distance, the focus image under the current field of view is acquired by the color image sensor. The electric displacement platform is controlled to move to the next field of view and then returns to execute the step of controlling the monochrome image sensor to acquire an overlapping image until all quasi-focused images under all fields of view are acquired, thus completing the phase shift detection autofocus of the dual wedge plate modulation. The predetermined linear equation between the sample axial defocus distance and the relative pixel offset includes: ； ； ； in, Indicates the axial defocus distance of the sample; This represents the first constant obtained by fitting a relation curve beforehand; Indicates relative pixel offset; This represents the second constant obtained by pre-fitting a relation curve; Indicates the segment offset; Indicates the cross-correlation offset; Indicates the total offset; This represents the pixel offset of the overlapping image at the focus position.

5. The phase shift detection autofocus method according to claim 4, characterized in that, Determining the relative pixel offset based on the left half image and the right half image includes: A cross-correlation image registration algorithm is used to calculate the horizontal pixel offset based on the left half image and the right half image; Convert the horizontal pixel offset to a relative pixel offset.

6. The phase shift detection autofocus method according to claim 5, characterized in that, The method employs a cross-correlation image registration algorithm to calculate the lateral pixel offset based on the left and right halves of the image, including: A first frequency domain representation is obtained by performing a Fourier transform on the left half of the image, and a second frequency domain representation is obtained by performing a Fourier transform on the right half of the image. The cross-power spectrum of the left half image and the right half image is calculated based on the first frequency domain representation and the second frequency domain representation; Perform an inverse Fourier transform on the cross-power spectrum to obtain the cross-correlation function; The horizontal pixel offset is determined by finding the horizontal coordinate of the main peak position in the cross-correlation function.

7. The phase shift detection autofocus method according to claim 6, characterized in that, The cross-power spectrum is: ； in, This represents the cross-power spectrum; Represents spatial frequency in the frequency domain; This represents the first frequency domain representation; This represents the complex conjugate of the second frequency domain representation; This represents the phase difference between the left half of the image and the right half of the image. It represents the imaginary unit.

8. The phase shift detection autofocus method according to claim 5, characterized in that, The step of converting the horizontal pixel offset to a relative pixel offset includes: The total offset is calculated based on the horizontal pixel offset and the segmentation offset of the overlapping image; Based on the total offset, the relative pixel offset is determined.

Images