TDI camera, super-resolution image acquisition method

Abstract

The present invention provides a TDI camera and a method for acquiring super-resolution images. The TDI camera comprises: a plurality of rectangular pixel units, each including a photosensitive region and a circuit region, wherein the length of each rectangular pixel unit is twice its width; the plurality of rectangular pixel units are arranged in an array, with the rectangular pixel units in odd-numbered columns constituting a first pixel array and the rectangular pixel units in even-numbered columns constituting a second pixel array, and in two adjacent pixel arrays, the photosensitive regions in the first pixel array and the circuit regions in the second pixel array being aligned and closely arranged along the width direction of the pixel units; and a plurality of square microlenses inclined at 45 degrees, the microlenses covering the first and second pixel arrays. The present invention can improve the yield of super-resolution images while achieving the same resolution as conventional devices.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a TDI camera and a method for acquiring super-resolution images. Background Art

[0002] The structure of patterned wafer defect inspection equipment is as follows Figure 1 As shown, it is used to detect various defects during the IC chip manufacturing process. The chip manufacturing process involves hundreds of process steps such as oxidation, photolithography, etching, ion implantation, thin film deposition, and polishing. Each process step may introduce deviations from the chip design layout. These deviations are transmitted and accumulated during the chip manufacturing process, causing IC chip defects, including protruding defects, recessed defects, and isolated particle defects. The extreme condition of protruding defects is bridging defects, and the extreme condition of recessed defects is disconnection defects.

[0003] The Time Delay Integration (TDI) sensor is a novel photoelectric sensor capable of rapidly imaging high-speed patterned wafers. It is a key component for image acquisition in patterned wafer defect inspection equipment. TDI cameras expose the same target multiple times, utilizing delayed integration technology to significantly enhance the collection of reflected light. This results in advantages such as high responsiveness and a wide dynamic range. Due to the requirements for clear TDI imaging, the device scanning speed must meet the following requirements: the device scanning speed is equal to the product of the object-side pixel size and the TDI camera's line rate; the image width is equal to the product of the object-side pixel size and the number of pixels in the non-scanning direction; and the area scanned per unit time is equal to the product of the image width and the device scanning speed. In practice, the object-side pixel size is strongly correlated with the detection sensitivity of the TDI camera. Maintaining this sensitivity is crucial, and because the effective image width is no larger than the field of view (FOV), conventional solutions can only improve device yield by expanding the FOV or increasing the TDI camera's line rate, both of which are challenging.

[0004] Therefore, the present invention proposes a TDI camera and a method for acquiring super-resolution images, which can improve the yield of super-resolution images while achieving the same resolution as the equipment of traditional solutions. Summary of the Invention

[0005] The present invention provides a TDI camera and a method for acquiring super-resolution images to solve the technical problem in the prior art that it is difficult to improve the yield of super-resolution images.

[0006] In a first aspect, the present invention provides a TDI camera, comprising: a plurality of rectangular pixel units, wherein the rectangular pixel units include a photosensitive area and a circuit area, wherein the photosensitive area and the circuit area are of equal area and are arranged along the long side of the pixel unit, and the length of the rectangular pixel unit is twice the width; the plurality of rectangular pixel units are arranged in an array, wherein the rectangular pixel units in odd columns constitute a first pixel array, and the rectangular pixel units in even columns constitute a second pixel array, and in two adjacent columns of the pixel array, the photosensitive area in the first pixel array and the circuit area in the second pixel array are aligned and closely arranged along the wide side direction of the pixel unit; and a plurality of square microlens units inclined at 45 degrees, wherein the microlens units cover the first pixel array and the second pixel array. The microlens units are arranged on a pixel array, and each of the microlens units is corresponding to the photosensitive area of ​​each pixel unit, so that each microlens unit fully covers the photosensitive area of ​​the corresponding pixel unit; the microlens units are arranged in an array, the microlens units corresponding to the first pixel array constitute a first microlens array, and the microlens units corresponding to the second pixel array constitute a second microlens array; the first microlens array is used to collect signal light from a target object and converge the signal light onto the photosensitive area of ​​the first pixel array to form a first image of the target object, and the second microlens array is used to collect and converge the signal light onto the photosensitive area of ​​the second pixel array to form a second image of the target object.

[0007] Its beneficial effect is that the first microlens array and the second microlens array respectively focus the incident light onto the photosensitive areas of the corresponding pixel arrays below. According to the imaging principle of the TDI camera, two independent images are obtained. Then, the two images are synthesized through the super-resolution reconstruction algorithm, which can improve the yield of super-resolution images while achieving the same resolution as the equipment of traditional solutions.

[0008] Optionally, the first image and the second image have a positional deviation of the width of the rectangular pixel unit in both the long and short sides of the rectangular pixel unit. This advantageously provides two images separated by a certain distance at corresponding positions in the horizontal and vertical directions by the positional deviation of the first image and the second image in both the long and short sides of the rectangular pixel unit.

[0009] Optionally, the TDI camera performs scanning integration along the long side direction of the rectangular pixel unit to acquire an image of the target object.

[0010] Optionally, the scanning of the TDI camera is achieved by uniformly moving the target object at a preset moving speed to acquire the first image and the second image.

[0011] Optionally, the preset movement speed is set based on a first product, where the first product is the product of the length of the rectangular pixel elements along the direction of scanning integration by the TDI camera and the line frequency of the TDI camera, and the optical magnification of the TDI camera. This advantageously allows the TDI camera to scan an area twice as large per unit time as in conventional techniques when moving a target object at the preset movement speed.

[0012] In a second aspect, the present invention provides a method for acquiring a super-resolution image, comprising: S1, acquiring a first image and a second image obtained by a TDI camera as described in any one of the first aspects; S2, combining and arranging the pixels in the first image and the second image according to the arrangement of sensors corresponding to the pixels in the first image and the second image to acquire a third image; S3, rotating the third image by 45 degrees, and filling the vicinity of the rotated third image with pixels having a pixel value of 0 to acquire a fourth image, wherein the fourth image is a minimum rectangular image that includes the rotated third image; S4, dividing the fourth image by every two pixels in both the row direction and the column direction to acquire a fifth image, a sixth image, a seventh image, and an eighth image; the fifth image , the resolutions of the sixth image, the seventh image and the eighth image are all smaller than those of the fourth image; S5, according to the fifth image, the sixth image, the seventh image and the eighth image, obtaining the pixel values ​​of the center positions of the intersections of the pixels in the fourth image to obtain a ninth image; S6, rotating the ninth image 45 degrees to obtain a tenth image, and the rotation direction of the ninth image is opposite to the rotation direction of the third image; S7, deleting the virtual pixels in the tenth image to obtain a super-resolution image, wherein the virtual pixels are pixels whose pixel values ​​are 0 that fill the vicinity of the rotated third image.

[0013] The beneficial effect is that: due to technical limitations, the existing technology needs to start four TDI cameras and obtain four low-resolution images to synthesize a super-resolution image, while the super-resolution image acquisition method provided by the present invention only needs two low-resolution images to synthesize a super-resolution image with the same resolution as the existing technology, saving resources and improving efficiency.

[0014] Optionally, in S4, it includes: dividing the fourth image every two pixels in the row direction and the column direction to obtain several sub-image parts containing four pixels; extracting the four pixels in the several sub-image parts in sequence according to the pixels occupying the same relative position in the sub-image parts, and synthesizing the pixels occupying the same relative position in the several sub-image parts into a fifth image, a sixth image, a seventh image and an eighth image in sequence according to the corresponding positions in the fourth image.

[0015] Optionally, in S5, it includes: S51, performing a discrete Fourier transform on the fifth image to obtain a first frequency domain image; performing a discrete Fourier transform on the sixth image to obtain a second frequency domain image; performing a discrete Fourier transform on the seventh image to obtain a third frequency domain image; performing a discrete Fourier transform on the eighth image to obtain a fourth frequency domain image; S52, obtaining sampling results of continuous Fourier transform corresponding to the first frequency domain image, the second frequency domain image, the third frequency domain image, and the fourth frequency domain image, and setting a truncation parameter of the sampling results of the continuous Fourier transform to 1; S53, according to the sampling results of the continuous Fourier transform, obtaining 8 sub-images that have a preset sub-pixel offset relationship with any one of the fifth image, the sixth image, the seventh image, or the eighth image, and the pixel values ​​of all pixels of the 8 sub-images are the pixel values ​​of the center position of the junction between the microlenses; S54, synthesizing the 8 sub-images into a high-resolution image to obtain a ninth image.

[0016] Optionally, in S53, the preset sub-pixel shift relationship includes: a first sub-pixel shift relationship, a second sub-pixel shift relationship, a third sub-pixel shift relationship, a fourth sub-pixel shift relationship, a fifth sub-pixel shift relationship, a sixth sub-pixel shift relationship, a seventh sub-pixel shift relationship and an eighth sub-pixel shift relationship; the first sub-pixel shift relationship is a shift of 0.25 pixels in the row direction and 0 pixels in the column direction; the second sub-pixel shift relationship is a shift of 0.75 pixels in the row direction and 0 pixels in the column direction; the third sub-pixel shift relationship is a shift of 0 pixels in the row direction and 0 pixels in the column direction. The fourth sub-pixel translation relationship is 0.5 pixel translation in the row direction and 0.25 pixel translation in the column direction; the fifth sub-pixel translation relationship is 0.25 pixel translation in the row direction and 0.5 pixel translation in the column direction; the sixth sub-pixel translation relationship is 0.75 pixel translation in the row direction and 0.5 pixel translation in the column direction; the seventh sub-pixel translation relationship is 0 pixel translation in the row direction and 0.75 pixel translation in the column direction; the eighth sub-pixel translation relationship is 0.5 pixel translation in the row direction and 0.75 pixel translation in the column direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of the structure of an existing TDI camera;

[0018] Figure 2 A schematic diagram of a partial structure embodiment of a TDI camera provided by the present invention;

[0019] Figure 3 A schematic diagram of the working principle of a TDI camera provided by the present invention;

[0020] Figure 4 It is an existing pixel structure;

[0021] Figure 5 A schematic diagram of an embodiment of a pixel structure provided by the present invention;

[0022] Figure 6 A schematic diagram of an embodiment of a fourth image provided by the present invention;

[0023] Figure 7 A schematic diagram of a sub-graph embodiment provided by the present invention;

[0024] Figure 8 A schematic diagram of an embodiment of a ninth image provided by the present invention;

[0025] Figure 9 A schematic diagram of an embodiment of a tenth image provided by the present invention;

[0026] Figure 10 This is a schematic diagram of an embodiment of a super-resolution image provided by the present invention. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application are described below in conjunction with the accompanying drawings. The terms used in the following embodiments are for the purpose of describing specific embodiments only and are not intended to limit the present application. As used in the specification and claims of this application, the singular expressions "a," "the," "the," "the," and "the" are intended to include expressions such as "one or more," unless the context clearly indicates otherwise. It should also be understood that in the following embodiments of this application, "at least one" and "one or more" refer to one or more (including two). The term "and / or" is used to describe an association relationship between associated objects, indicating that three relationships can exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0028] References to "one embodiment" or "some embodiments" etc. described in this specification mean that the specific features, structures or characteristics described in conjunction with the embodiment are included in one or more embodiments of the present application. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. appearing in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in another way. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in another way. The term "connected" includes direct and indirect connections, unless otherwise stated. "First" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated.

[0029] In the embodiments of this application, words such as "exemplarily" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplarily" or "for example" in the embodiments of this application should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplarily" or "for example" is intended to present the relevant concepts in a concrete manner.

[0030] The present invention provides a TDI camera, comprising: a plurality of rectangular pixel units, wherein the rectangular pixel units include a photosensitive area and a circuit area, wherein the photosensitive area and the circuit area are of equal area and arranged along the long side of the pixel unit, and the length of the rectangular pixel unit is twice the width; the plurality of rectangular pixel units are arranged in an array, wherein the rectangular pixel units in odd-numbered columns constitute a first pixel array, and the rectangular pixel units in even-numbered columns constitute a second pixel array, and in two adjacent columns of the pixel array, the photosensitive area in the first pixel array and the circuit area in the second pixel array are aligned and arranged closely along the wide side of the pixel unit; and a plurality of square microlens units inclined at 45 degrees, wherein the microlens units cover the first pixel array and the second pixel array. The microlens units are arranged on a pixel array, and each of the microlens units is corresponding to the photosensitive area of ​​each pixel unit, so that each microlens unit fully covers the photosensitive area of ​​the corresponding pixel unit; the microlens units are arranged in an array, the microlens units corresponding to the first pixel array constitute a first microlens array, and the microlens units corresponding to the second pixel array constitute a second microlens array; the first microlens array is used to collect signal light from a target object and converge the signal light onto the photosensitive area of ​​the first pixel array to form a first image of the target object, and the second microlens array is used to collect and converge the signal light onto the photosensitive area of ​​the second pixel array to form a second image of the target object.

[0031] For example, Figure 2 As shown in (a) in Figure 2 The plurality of sensors 301 shown in (a) constitute a first pixel array, as shown in FIG. Figure 2 The plurality of sensors 302 shown in (a) constitute a second pixel array. The first pixel array and the second pixel array output two independent images through two independent sets of readout circuits. The first pixel array and the second pixel array work independently. Figure 2 As shown in (b), a square represents a microlens, such as Figure 2 The plurality of micro lenses 303 shown in (b) constitute a first micro lens array, as shown in FIG. Figure 2 The plurality of microlenses 304 shown in (b) constitute a first microlens array.

[0032] The first microlens array covers the first pixel array, and the second microlens array covers the second pixel array.

[0033] like Figure 3 As shown in (a) of the figure, a microlens array arranged in a diamond structure is covered on the pixel array, so that all the light information passing through the microlenses can enter the pixels behind the microlenses, as shown in the figure. Figure 3 As shown in (b) in the figure. This allows most of the incident light to be collected by the photosensitive area, and due to the imaging characteristics of the TDI camera, any position in the field of view can be imaged by the two sensor arrays in sequence. Figure 3 As shown in (a), the thick black frame is a single pixel, the length of the pixel is twice the width, the shaded area is the photosensitive area, and the white area is the transmission area. The first microlens array and the second microlens array are used to focus the incident light onto the photosensitive area of ​​the corresponding sensor array below. According to the imaging principle of the TDI camera, two images with a certain size interval at corresponding positions in the horizontal and vertical directions are obtained. Then, the two images are synthesized through the super-resolution reconstruction algorithm, which can improve the yield of super-resolution images while achieving the same super-resolution of the equipment of the traditional solution. Optionally, the shape of the microlens is square, and the width of the microlens is 1 / 4 of the width of the pixel. times. Figure 3 The “TDI sensor” in the figure refers to the TDI camera in this application.

[0034] Optionally, the first image and the second image have a positional deviation of the width of the rectangular pixel unit in both the long and short sides of the rectangular pixel unit. This advantageously provides two images separated by a certain distance at corresponding positions in the horizontal and vertical directions by the positional deviation of the first image and the second image in both the long and short sides of the rectangular pixel unit.

[0035] Optionally, the TDI camera performs scanning integration along the long side direction of the rectangular pixel unit to acquire an image of the target object.

[0036] Optionally, the scanning of the TDI camera is achieved by uniformly moving the target object at a preset moving speed to acquire the first image and the second image.

[0037] Optionally, the preset movement speed is set based on a first product, where the first product is the product of the length of the rectangular pixel elements along the direction of scanning integration by the TDI camera and the line frequency of the TDI camera, and the optical magnification of the TDI camera. This advantageously allows the TDI camera to scan an area twice as large per unit time as in conventional techniques when moving a target object at the preset movement speed.

[0038] Specifically, for traditional arrangements, such as Figure 4 As shown, each square represents a pixel. A pixel is a square with a side length of A. The optical magnification of the TDI sensor, i.e., the TDI camera, is M, and the number of pixels in the horizontal direction is W. The line frequency of the TDI camera is F. Therefore, the scanning area per unit time of the TDI camera in the prior art is: .

[0039] By arranging the internal components of the TDI camera provided by this application, if you want to obtain an image with the same resolution as the traditional arrangement, such as Figure 5 As shown, the optical magnification of the TDI camera provided by the present invention is M, the length of the pixel along the scanning direction is 2A, and the width is A. Considering the limitation of the field of view of the objective lens, the number of pixels in the horizontal direction of each sensor array is W / 2, and the line frequency of the TDI camera provided by the present invention is F. Then, the scanning area per unit time of the TDI camera provided by the present invention is: Since W is much larger than 1, S1 is approximately equal to 2 times S0. Figure 5 The “TDI sensor” in the figure refers to the TDI camera in this application.

[0040] Because the microlens array is densely packed in a diamond shape, while the pixels in the display are densely packed in a square shape, the display cannot directly display the first image obtained from the first microlens array and the second image obtained from the second microlens array simultaneously. To solve this problem, interpolation can be used to obtain an image that meets the display's display mode, but interpolation will cause the loss of high-frequency information, so it is not a good choice.

[0041] In order to reduce the loss of image information, we use super-resolution reconstruction (SRR) technology to synthesize the first image and the second image into a super-resolution image with higher resolution than the low-resolution image.

[0042] Super-resolution reconstruction (SRR) technology generally refers to methods for generating higher-resolution images from low-resolution images. It is used to reconstruct a high-resolution image based on a single or multiple low-resolution images through digital image processing. Super-resolution reconstruction is a crucial research topic in computer vision and image processing, with various implementation methods, including spatial, frequency, interpolation, and statistical approaches. In recent years, there has been significant progress in super-resolution reconstruction using machine learning.

[0043] While typical interpolation-based super-resolution reconstruction methods are fast, the improvement is generally limited. Deep learning methods, while more effective, are computationally intensive and unsuitable for real-time super-resolution reconstruction. To maximize the preservation of frequency domain features while maintaining high speed, frequency domain super-resolution algorithms can be used.

[0044] However, traditional frequency-domain image super-resolution techniques require at least four low-resolution images. Acquiring these four low-resolution images simultaneously requires four TDI cameras and their associated equipment, significantly increasing the difficulty and cost of installation. However, using this diamond-shaped, densely packed microlens array, two low-resolution images can be directly split into four equally scaled, lower-resolution images. These images can then be reconstructed using traditional frequency-domain image super-resolution techniques to obtain a super-resolution image. This is something that traditional square-shaped, densely packed microlens arrays cannot achieve.

[0045] Based on the TDI camera provided in any of the above embodiments, the present invention further provides a method for acquiring a super-resolution image, comprising:

[0046] S1. Acquire a first image and a second image obtained by the TDI camera described in any one of the above embodiments;

[0047] S2. Combining and arranging pixels in the first image and the second image according to an arrangement of sensors corresponding to the pixels in the first image and the second image to obtain a third image;

[0048] S3, rotating the third image by 45 degrees, and filling the area near the rotated third image with pixels of 0 to obtain a fourth image, wherein the fourth image is a minimum rectangular image that contains the rotated third image; the structure of the fourth image is as follows: Figure 6 shown.

[0049] S4, dividing the fourth image into every two pixels in both the row direction and the column direction to obtain a plurality of sub-image parts including pixel 1, pixel 2, pixel 3 and pixel 4, such as Figure 7 As shown in (a) in FIG; the pixels in the four pixel positions of the sub-image portion corresponding to the fourth image are extracted respectively and aggregated into four images, which are the fifth image, the sixth image, the seventh image and the eighth image respectively; the pixels in each of the four images correspond to the pixels in one of the four pixel positions of the sub-image portion corresponding to the fourth image, and the four images are as shown in FIG. Figure 7 As shown in (b), (c), (d) and (e), Figure 7 The fifth image shown in (b) contains only pixels 1, Figure 7 The sixth image shown in (c) contains only pixels 2, Figure 7 The seventh image shown in (d) contains only pixels 3, Figure 7 The eighth image shown in (e) contains only pixel 4;

[0050] S5. Obtain pixel values ​​at the center of the intersections of pixels in the fourth image based on the fifth image, the sixth image, the seventh image, and the eighth image to obtain a ninth image; the ninth image is as follows: Figure 8 As shown, the center position of each pixel junction in the fourth image is as shown in Figure 8 The resolution of the ninth image is higher than that of the first image and the second image.

[0051] S6, rotating the ninth image by 45 degrees to obtain a tenth image, wherein the direction in which the ninth image is rotated is opposite to the direction in which the third image is rotated; the tenth image is as follows: Figure 9 shown.

[0052] S7, deleting virtual pixels in the tenth image to obtain a super-resolution image, wherein the virtual pixels are pixels whose pixel values ​​are 0 in the vicinity of the rotated third image; Figure 10 As shown by the white dots in (a), only the Figure 10 The black dots in (a) are used to obtain Figure 10 The super-resolution image is composed of the black dots shown in (b).

[0053] The advantages of the super-resolution image acquisition method provided by the present invention are: the degree of improvement is superior to general interpolation methods, the calculation speed is faster, and it is superior to deep learning methods; it retains all frequency domain information and better preserves the information of the super-resolution image; and by utilizing the images captured by the TDI camera provided by the present invention, a frequency domain super-resolution algorithm that originally required four images is completed with only two images without almost any loss of frequency domain information, thereby obtaining a super-resolution image.

[0054] In some embodiments, in S4, it includes: dividing the fourth image every two pixels in the row direction and the column direction to obtain several sub-image parts containing four pixels; extracting the four pixels in the several sub-image parts in sequence according to the pixels occupying the same relative position in the sub-image parts, and synthesizing the pixels occupying the same relative position in the several sub-image parts into the fifth image, the sixth image, the seventh image and the eighth image in sequence according to the corresponding positions in the fourth image.

[0055] In some embodiments, S5 includes: S51, performing a discrete Fourier transform on the fifth image to obtain a first frequency domain image; performing a discrete Fourier transform on the sixth image to obtain a second frequency domain image; performing a discrete Fourier transform on the seventh image to obtain a third frequency domain image; performing a discrete Fourier transform on the eighth image to obtain a fourth frequency domain image; S52, obtaining sampling results of continuous Fourier transforms corresponding to the first frequency domain image, the second frequency domain image, the third frequency domain image, and the fourth frequency domain image, wherein a truncation parameter of the sampling results of the continuous Fourier transform is set to 1; S53, based on the sampling results of the continuous Fourier transform, obtaining eight sub-images having a preset sub-pixel offset relationship with any one of the fifth image, the sixth image, the seventh image, or the eighth image, wherein the pixel values ​​of all pixels in the eight sub-images are the pixel values ​​at the center position of the junction between the microlenses; and S54, synthesizing the eight sub-images into a high-resolution image to obtain a ninth image.

[0056] In some embodiments, in S53, the preset sub-pixel shift relationship includes: a first sub-pixel shift relationship, a second sub-pixel shift relationship, a third sub-pixel shift relationship, a fourth sub-pixel shift relationship, a fifth sub-pixel shift relationship, a sixth sub-pixel shift relationship, a seventh sub-pixel shift relationship, and an eighth sub-pixel shift relationship; the first sub-pixel shift relationship is shifted 0.25 pixels in the row direction and 0 pixels in the column direction; the second sub-pixel shift relationship is shifted 0.75 pixels in the row direction and 0 pixels in the column direction; the third sub-pixel shift relationship is shifted 0 pixels in the row direction and 0 pixels in the column direction. The fourth sub-pixel translation relationship is 0.5 pixel translation in the row direction and 0.25 pixel translation in the column direction; the fifth sub-pixel translation relationship is 0.25 pixel translation in the row direction and 0.5 pixel translation in the column direction; the sixth sub-pixel translation relationship is 0.75 pixel translation in the row direction and 0.5 pixel translation in the column direction; the seventh sub-pixel translation relationship is 0 pixel translation in the row direction and 0.75 pixel translation in the column direction; the eighth sub-pixel translation relationship is 0.5 pixel translation in the row direction and 0.75 pixel translation in the column direction.

[0057] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A TDI camera, characterized in that: include: A plurality of rectangular pixel units, each comprising a photosensitive region and a circuit region, wherein the photosensitive region and the circuit region are of equal area and arranged along a long side of the pixel unit, and the length of the rectangular pixel unit is twice the width; The plurality of rectangular pixel units are arranged in an array, wherein the rectangular pixel units in odd-numbered columns constitute a first pixel array, and the rectangular pixel units in even-numbered columns constitute a second pixel array, and in two adjacent columns of the pixel array, the photosensitive areas in the first pixel array and the circuit areas in the second pixel array are aligned and closely arranged along the width direction of the pixel units, and the first pixel array and the second pixel array output two independent images through two independent sets of readout circuits; a plurality of square microlens units inclined at 45 degrees, the microlens units covering the first pixel array and the second pixel array, and each microlens unit being arranged corresponding to a photosensitive area of ​​each pixel unit so that each microlens unit fully covers the photosensitive area of ​​the corresponding pixel unit; The microlens units are arranged in an array, the microlens units corresponding to the first pixel array constitute a first microlens array, and the microlens units corresponding to the second pixel array constitute a second microlens array; The first microlens array is used to collect signal light from the target object and converge the signal light onto the photosensitive area of ​​the first pixel array to form a first image of the target object, and the second microlens array is used to collect and converge the signal light onto the photosensitive area of ​​the second pixel array to form a second image of the target object.

2. The TDI camera according to claim 1, wherein: The first image and the second image have a position deviation of the width of the rectangular pixel unit in both the long side direction and the short side direction of the rectangular pixel unit.

3. The TDI camera according to claim 1, characterized in that The TDI camera performs scanning integration along the long side direction of the rectangular pixel unit to acquire an image of the target object.

4. The TDI camera according to claim 3, characterized in that The target object is moved at a constant speed by a preset moving speed to achieve scanning by the TDI camera to obtain the first image and the second image.

5. The TDI camera according to claim 4, characterized in that Also includes: The preset moving speed is set according to a quotient of a first product and an optical magnification of the TDI camera, wherein the first product is a product of a length of the rectangular pixel unit in a direction of scanning integration by the TDI camera and a line frequency of the TDI camera.

6. A method for acquiring a super-resolution image, characterized in that: include: S1. Acquire a first image and a second image obtained by the TDI camera according to any one of claims 1 to 5; S2. Combining and arranging pixels in the first image and the second image according to an arrangement of sensors corresponding to the pixels in the first image and the second image to obtain a third image; S3, rotating the third image by 45 degrees, and filling the area near the rotated third image with pixels having a pixel value of 0 to obtain a fourth image, wherein the fourth image is a minimum rectangular image that contains the rotated third image; S4. Divide the fourth image into sub-image portions every two pixels in both the row and column directions to obtain a plurality of sub-image portions each containing four pixels. Extract the four pixels in the sub-image portions sequentially according to the pixels occupying the same relative position in the sub-image portions. Combine the pixels occupying the same relative position in the sub-image portions into a fifth image, a sixth image, a seventh image, and an eighth image in sequence according to their corresponding positions in the fourth image. The resolutions of the fifth image, the sixth image, the seventh image, and the eighth image are all smaller than those of the fourth image. S5. Obtaining pixel values ​​at center positions of intersections of pixels in the fourth image based on the fifth image, the sixth image, the seventh image, and the eighth image to obtain a ninth image, where the pixel values ​​at center positions of the intersections of pixels are pixel values ​​of all pixels in eight sub-images that have a preset sub-pixel offset relationship with any one of the fifth image, the sixth image, the seventh image, or the eighth image. S6. Rotate the ninth image by 45 degrees to obtain a tenth image, wherein the rotation direction of the ninth image is opposite to the rotation direction of the third image; S7. Deleting virtual pixels in the tenth image to obtain a super-resolution image, wherein the virtual pixels are pixels whose pixel values ​​are 0 and are filled in the vicinity of the rotated third image.

7. The method for obtaining a super-resolution image according to claim 6, wherein: In S5, it includes: S51. Perform a discrete Fourier transform on the fifth image to obtain a first frequency domain image; perform a discrete Fourier transform on the sixth image to obtain a second frequency domain image; perform a discrete Fourier transform on the seventh image to obtain a third frequency domain image; and perform a discrete Fourier transform on the eighth image to obtain a fourth frequency domain image. S52: Acquire sampling results of continuous Fourier transform corresponding to the first frequency domain image, the second frequency domain image, the third frequency domain image, and the fourth frequency domain image, wherein a truncation parameter of the sampling results of the continuous Fourier transform is set to 1; S53. Acquire, based on the sampling results of the continuous Fourier transform, eight sub-images having a preset sub-pixel offset relationship with any one of the fifth image, the sixth image, the seventh image, or the eighth image, where the pixel values ​​of all pixels in the eight sub-images are the pixel values ​​at the center of the intersection between the microlenses. S54 , synthesizing the eight sub-images into a high-resolution image to obtain a ninth image.

8. The method for obtaining a super-resolution image according to claim 7, wherein: In S53, the preset sub-pixel shift relationship includes: a first sub-pixel shift relationship, a second sub-pixel shift relationship, a third sub-pixel shift relationship, a fourth sub-pixel shift relationship, a fifth sub-pixel shift relationship, a sixth sub-pixel shift relationship, a seventh sub-pixel shift relationship, and an eighth sub-pixel shift relationship; The first sub-pixel translation relationship is 0.25 pixels in the row direction and 0 pixels in the column direction; the second sub-pixel translation relationship is 0.75 pixels in the row direction and 0 pixels in the column direction; the third sub-pixel translation relationship is 0 pixels in the row direction and 0.25 pixels in the column direction; the fourth sub-pixel translation relationship is 0.5 pixels in the row direction and 0.25 pixels in the column direction; the fifth sub-pixel translation relationship is 0.25 pixels in the row direction and 0.5 pixels in the column direction; the sixth sub-pixel translation relationship is 0.75 pixels in the row direction and 0.5 pixels in the column direction; the seventh sub-pixel translation relationship is 0 pixels in the row direction and 0.75 pixels in the column direction; the eighth sub-pixel translation relationship is 0.5 pixels in the row direction and 0.75 pixels in the column direction.

Images