Multi-module parallax-free spectral imaging method based on Sammer lens

Abstract

The invention discloses a multi-module parallax-free spectral imaging method based on a Sammer lens, and relates to the technical field of spectral imaging, and the method mainly comprises the steps: determining a reference module as an image reference standard of a whole imaging module system; controlling the reference module and all the compensation modules to synchronously acquire images of the target area, wherein each module acquires image information of different spectral bands; directly performing pixel-level spectrum fusion on an image acquired by each compensation module and an image acquired by the reference module on the basis of a geometric mapping relation determined by a curing space position of each compensation module relative to the reference module and a preset fixed inclination angle of each compensation module; and outputting the fused parallax-free multispectral image. According to the invention, the parallax among multiple modules can be thoroughly eliminated from the optical structure level.

Description

Technical Field

[0001] This invention relates to the field of spectral imaging technology, and specifically to a multi-module parallax-free spectral imaging method based on a Sham lens. Background Technology

[0002] In existing technologies, a system architecture comprising multiple independent imaging modules is commonly used to achieve multi-band synchronous imaging. For a six-module configuration, a typical existing solution employs a combination of "all ordinary industrial lenses" and mechanical calibration with algorithmic compensation. In this approach, all six spectral modules are equipped with ordinary industrial lenses of the same specifications, physically fixed using high-precision mechanical supports, and manually calibrated one by one to ensure the optical axes of each lens remain parallel. After imaging, complex image registration algorithms are used to digitally correct parallax deviations caused by the physical spacing between modules. Although some improved solutions attempt to optimize optical axis parallelism by increasing machining precision, they still do not introduce special optical lens designs with passive parallax compensation characteristics, nor do they deeply consider the forward and backward tilt compensation requirements brought about by the common two-row, three-column asymmetrical layout. In addition, other technical approaches attempt to reduce the number of modules by using shared beam-splitting components, but this often introduces new problems such as spectral crosstalk and energy loss. Especially when imaging targets at distances of 3 meters or more, existing solutions generally lack a coordinated compensation design for horizontal, vertical, and axial three-dimensional parallax. Relying solely on mechanical adjustments and software algorithms, the parallax elimination effect of the system is heavily dependent on the initial calibration state and a stable external environment. It is sensitive to changes in ambient temperature and vibration, and prone to optical axis shift and parallax recurrence. At the same time, the cumbersome post-processing algorithms also bring a huge computational burden, and the correction accuracy is limited, making it difficult to simultaneously meet the application requirements of high precision and real-time imaging. Therefore, there is an urgent need for a multi-module spectral imaging solution that can achieve stable, passive three-dimensional parallax compensation from both optical and structural perspectives. Summary of the Invention

[0003] To achieve stable, passive, three-dimensional parallax-free multi-module spectral imaging, this invention proposes a multi-module parallax-free spectral imaging method based on a SAM lens. The method comprises an imaging module system arranged in two rows and three columns. The imaging module in the middle of the first row is a reference module equipped with a standard industrial lens. The remaining five imaging modules are compensation modules equipped with customized SAM lenses. Each compensation module's SAM lens undergoes customized optical design based on a predetermined offset position relative to the reference module, and fixed tilt angles around the X, Y, and Z axes are set for passive three-dimensional parallax compensation. The method specifically includes the following steps: S1: Establish the reference module as the image reference benchmark for the entire imaging module system; S2: Control the reference module and all compensation modules to synchronously acquire images of the target area, with each imaging module acquiring images of different spectral bands; S3: Based on the geometric mapping relationship determined by the fixed spatial position of each compensation module relative to the reference module and its preset fixed tilt angle, the images acquired by each compensation module are directly fused with the images acquired by the reference module at the pixel level. S4: Outputs a fused, disparity-free multispectral image.

[0004] Furthermore, the spacing between adjacent imaging modules is equal.

[0005] Furthermore, the fixed tilt angles of the rotation around the X-axis, Y-axis, and Z-axis have specific rotation requirements as follows: In the first row, the first and third imaging modules have a horizontal displacement relative to the reference module, and the horizontal parallax in the X-axis direction needs to be compensated. In the second row, the first, second, and third imaging modules have vertical and axial offsets relative to the reference module, requiring compensation for vertical parallax in the Y-axis direction and axial parallax in the Z-axis direction. At the same time, the first and third imaging modules also have horizontal displacements relative to the reference module, requiring compensation for horizontal parallax in the X-axis direction.

[0006] Furthermore, in the first row, the first and third imaging modules compensate for the horizontal parallax in the X-axis direction, and in the second row, the first and third imaging modules compensate for the horizontal parallax in the X-axis direction, calculated using the following formula: In the formula, A fixed tilt angle is required to compensate for horizontal parallax in the X-axis direction. The horizontal spacing between adjacent imaging modules in the same row. For the lens field of view, For the lens focal length, For imaging distance, This is the field of view correction factor. This is the distortion correction factor.

[0007] Furthermore, in the second row, the first, second, and third imaging modules compensate for the vertical parallax in the Y-axis direction and the axial parallax in the Z-axis direction using the following formula: In the formula, A fixed tilt angle is required to compensate for vertical parallax in the Y-axis direction. A fixed tilt angle is required to compensate for axial parallax in the Z-axis direction. To optimize the synergistic amplification factor, The vertical spacing between the two rows of imaging modules. The vertical distance between the second row of imaging modules and the first row of imaging modules.

[0008] Furthermore, the customized optical design specifically means that the focal length of the SAM lens is the same as that of a regular industrial lens.

[0009] Furthermore, the customized optical design specifically means that the distortion coefficient of the SAM lens is consistent with that of a regular industrial lens.

[0010] Furthermore, in step S3, the pixel-level spectral fusion specifically involves: Image pixel coordinates acquired by the reference module Using the reference image, the first image is determined through a predetermined mapping function. The pixel coordinates of the corresponding object point in the image of each compensation module Pixel-level spectral fusion is performed between the coordinate-mapped images of each compensation module and the reference image. The parameters of the mapping function include: the distance between the current compensation module and the reference module, and the preset fixed tilt angles of the current compensation module around the X, Y, and Z axes.

[0011] Furthermore, in step S3, the pixel-level spectral fusion adopts a weighted fusion method, expressed by the formula: In the formula, For the fused multispectral image, As the reference image, For the first The image after coordinate mapping of each compensation module These are the preset weighting coefficients.

[0012] Compared with the prior art, the present invention has at least the following beneficial effects: (1) The present invention proposes a multi-module parallax-free spectral imaging method based on Sham lens. By constructing an imaging system with a reference module as the center and multiple customized Sham lens modules arranged around it, and based on the fixed spatial position and fixed tilt angle of each module relative to the reference, the parallax between multiple modules is completely eliminated from the optical structure level. (2) Utilizing the optical characteristics of the SAM lens, it can passively compensate for horizontal, vertical and axial parallax caused by the module spacing during the imaging process, based on its own layout position in the array. (3) All key optical parameters of the SAM lens are strictly matched with the reference module. Combined with the pixel-level spectral fusion method based on the preset geometric mapping relationship, the multi-band images have high consistency at the imaging source. There is no need to perform complex parallax correction calculations, which greatly improves the real-time performance of image fusion and the overall system response speed. Attached Figure Description

[0013] Figure 1 A flowchart illustrating the steps of a multi-module parallax-free spectral imaging method based on a Sham lens; Figure 2 This is a schematic diagram of a two-row, three-column imaging module device; Explanation of reference numerals in the attached drawings: 1-Imaging module, 2-Integrated support platform, 3-Mounting cavity. Detailed Implementation

[0014] Although existing technologies have attempted to suppress parallax in multi-module spectral cameras through high-precision machining and complex image registration algorithms, their fundamental limitation lies in failing to eliminate parallax at its physical source in optical imaging. This causes the system to always rely on the initial accuracy of calibration and the subsequent compensation capability of the algorithm, resulting in inherent deficiencies in long-term stability, environmental adaptability, and real-time performance. To address this, this invention proposes a multi-module parallax-free spectral imaging method based on Sham lenses. This method comprises an imaging module system arranged in two rows and three columns. The imaging module in the middle of the first row is a reference module equipped with a standard industrial lens. The remaining five imaging modules are compensation modules equipped with customized Sham lenses. The Sham lenses of each compensation module are customized optically based on their predetermined offset position relative to the reference module, and fixed tilt angles are set around the X, Y, and Z axes for passive three-dimensional parallax compensation. The specific steps include: S1: Establish the reference module as the image reference benchmark for the entire imaging module system; S2: Control the reference module and all compensation modules to synchronously acquire images of the target area, with each imaging module acquiring images of different spectral bands; S3: Based on the geometric mapping relationship determined by the fixed spatial position of each compensation module relative to the reference module and its preset fixed tilt angle, the images acquired by each compensation module are directly fused with the images acquired by the reference module at the pixel level. S4: Outputs a fused, disparity-free multispectral image.

[0015] Specific embodiments of the present invention provide a detailed implementation of a multi-module parallax-free spectral imaging method based on a Sham lens, clearly demonstrating how the technical concept of the present invention can be put into practice. It should be understood that the specific parameters, materials, dimensions, and angle values ​​described below are for better illustration of the present invention and not for limiting the scope of protection of the present invention. Those skilled in the art, based on their understanding of the core ideas of the present invention, can adapt and replace the following parameters according to different imaging distances, spectral ranges, resolution requirements, or volume constraints.

[0016] In an exemplary embodiment, the imaging system employs a compact and regular array layout, specifically two rows and three columns of six imaging modules 1, as shown below. Figure 2 As shown. This layout is advantageous for accommodating multiple spectral channels within a limited space and forming symmetrical or quasi-symmetrical geometric relationships, facilitating systematic parallax analysis and compensation design. As a preferred but not limiting option in this embodiment, the spacing between the optical centers of all adjacent imaging modules 1 is set to an equal value, for example, 20 mm. This equal spacing simplifies the symmetry analysis of machining and optical design, making it easier to calculate and manage the offset of each compensation module relative to the reference module. Of course, depending on actual design needs, such as to adapt to sensors of different sizes or to pursue a specific field-of-view overlap rate, the spacing can also be designed to other equal or unequal values, as long as the relationship is determined and can be measured in advance.

[0017] In the array, the middle position of the first row is selected as the reference module. The core function of the reference module is to provide a stable origin for coordinates, serving as a spatial and image reference for the entire system. It is equipped with a stable, standard industrial lens with known optical characteristics. For example, a fixed-focus lens with a medium focal length (e.g., 20mm to 35mm) can be chosen; in this example, for simplicity, a 25mm lens with a field of view of approximately 75 degrees is selected. This lens should possess good image quality and controllable distortion, and be equipped with a specific spectral filter, such as a narrowband filter with a center wavelength near the visible green light band (e.g., 550nm), to acquire image information in the reference band.

[0018] The remaining five imaging modules 1 serve as compensation modules, arranged around the reference module. At the core of each compensation module is a customized Sham lens. The Sham lens, designed according to Sham's law, achieves unique compensation for non-coplanar imaging by constructing its lens plane, imaging plane, and subject plane in a special relationship where they intersect on the same straight line. In this invention, this characteristic of the Sham lens is applied to the passive elimination of parallax in multi-module systems. The design of each Sham lens is not uniform but highly customized. This customization is based on the predetermined spatial offset position (including horizontal, vertical, and axial components) of the compensation module relative to the reference module in the array, and the fixed tilt angle preset to achieve passive parallax compensation. These fixed tilt angles are defined as rotation angles around the X, Y, and Z axes of a three-dimensional Cartesian coordinate system with the optical center of the reference module as the origin. The X-axis is typically defined as the horizontal direction, parallel to the row direction of the module arrangement; the Y-axis is defined as the vertical direction, parallel to the column direction of the module arrangement; and the Z-axis is along the optical axis of the reference module. Through a precise integrated support platform, these preset fixed rotation angles around the X, Y, and Z axes, unique to each mounting position, are processed and solidified on the module's mounting interface. This ensures that after installation, the optical axis of each compensation module naturally points in a calculated direction, thereby intersecting with the optical axis of the reference module at a distant target.

[0019] To illustrate the logic of this angle preset more specifically, this embodiment uses a typical two-row, three-column layout with a spacing of 20 mm as an example (with the reference module as the origin, the left side of the same row is the negative X-axis direction, the right side of the same row is the positive X-axis direction, and the extension direction of the compensation module in the second column of the second row is the positive Y-axis direction). This is an example optimized by optical simulation, but it is by no means the only solution. For the compensation module located in the first row (i.e., with the same Y-axis coordinate) and in the first column as the reference module, its main offset relative to the reference module is 20 mm in the negative X-axis direction. To compensate for the resulting horizontal parallax and ensure that it sees the same far-distance field of view center as the reference module, the mounting interface of this compensation module is preset with a fixed tilt angle around the X-axis, for example, -5° (the negative sign represents a specific rotation direction according to the right-hand rule). Similarly, the compensation module in the third column of the first row is preset with a tilt angle of +5° around the X-axis. Since they have no relative offset from the reference module in the Y and Z directions (in this layout model), their tilt angles around the Y and Z axes are preset to 0°. The situation is more complex for the compensation modules in the second row. Relative to the reference module, they may not only be offset in the X-direction (different positions on the left, center, and right), but also have a downward offset in the Y-direction (e.g., -20 mm). Furthermore, considering lens occlusion and to maintain focal plane consistency, there may be a slight depth difference in the Z-direction due to the two-row layout, which assists the optical system's focusing and image quality. Therefore, compensation for them requires coordinated three-dimensional angles. For example, after optimization, all modules in the second row can be uniformly preset with a tilt angle around the Y-axis (e.g., -3.5°) to compensate for vertical offset, and uniformly preset with a tilt angle around the Z-axis (e.g., +4.2°) to coordinately compensate for the effects of depth difference and angle. Based on this, the modules on the left, center, and right of the second row are further tilted around the X-axis by -5°, 0°, and +5° respectively to address their respective horizontal offset differences. These specific angle values ​​are optimized solutions obtained through simulation under specific boundary conditions such as an object distance of 3 meters, a focal length of 25 mm, and a spacing of 20 mm. They clearly demonstrate the core idea of ​​"customized angle presets based on relative positions". In practical applications, if the object distance range, focal length, or layout changes, these angle values ​​will inevitably change accordingly, but the basic principle of determining and fixing them based on geometric relationships remains unchanged.

[0020] Specifically, in the first row, the first and third column modules need to compensate for horizontal parallax in the X-axis direction, and in the second row, the first and third column modules need to compensate for horizontal parallax in the X-axis direction. This can be calculated using the following formula: , In the formula, The tilt angle required to compensate for horizontal parallax in the X-axis direction. The horizontal spacing between adjacent modules For the lens field of view, For the lens focal length, For imaging distance, This is the field of view correction factor. This is the distortion correction factor.

[0021] In the second row, the first, second, and third column modules need to compensate for the vertical parallax in the Y-axis direction and the axial parallax in the Z-axis direction, which can be calculated using the following formula: , , In the formula, The tilt angle required to compensate for vertical parallax in the Y-axis direction The tilt angle required to compensate for axial parallax in the Z-axis direction. This is for the layout of the synergistic amplification factor.

[0022] The customized optical design of the SAM lens is another key to achieving "passive compensation." The design goal of each SAM lens is to match the overall optical characteristics of the reference lens when mounted at a specific angle. This primarily involves matching the focal length. For example, if the reference lens has a focal length of 25mm, then all SAM lenses should also be designed with a focal length of 25mm to ensure consistent image scale. However, this does not mean that the focal length of this invention must be 25mm; it can be any focal length suitable for long-distance imaging, such as 16mm, 35mm, or 50mm. The important thing is that the focal lengths of all modules match each other. Secondly, and more importantly, is the matching of aberration characteristics, especially distortion characteristics. Lens distortion causes non-linear shifts in pixel positions in the image. If the distortion patterns of the lenses differ significantly, even with correct geometric projection relationships, precise pixel-level alignment cannot be achieved. Therefore, during the design phase of the SAM lens, its imaging distortion (including barrel or pincushion distortion) under a preset tilted mounting state needs to be optimized to make it as consistent as possible with the distortion curve of the reference lens under normal mounting state. For example, it can be required that the distortion coefficient (a parameter for quantifying distortion) of all lenses be controlled within the same small range. This active matching of optical characteristics is a prerequisite for ensuring the effectiveness of subsequent simple mapping algorithms. SAM lenses can be manufactured using high-precision glass molding or composite grinding processes, and the lens material can be selected according to the required spectral range, such as ordinary optical glass, quartz glass, or sapphire glass.

[0023] like Figure 2As shown, the integrated support platform 2, which supports the entire module array, is the physical basis for achieving the preset angle. This platform is typically made of a material with high rigidity and low coefficient of thermal expansion, such as aluminum alloy, stainless steel, or engineering ceramics. Using a high-precision multi-axis CNC machining center, the mounting cavities 3 for all modules are machined onto the platform substrate in a single operation. The bottom of each cavity is not a simple plane, but a composite inclined plane calculated based on the aforementioned preset tilt angle. Using the table surface of the integrated support platform 2 as the platform reference plane, the spatial orientation of this inclined plane relative to the platform reference plane needs to be precisely controlled during machining. This specific orientation is defined by a set of three independent rotation angle parameters (e.g., pitch angle, yaw angle, and roll angle) determined in advance through optical simulation and oriented around the platform's fixed coordinate system X-axis, Y-axis, and Z-axis. High-precision CNC machining ensures that the actual orientation of the inclined plane is consistent with this set of design angle parameters. The core of this method of defining the spatial orientation of the mounting positions by fixing the three axial tilt angles lies in directly converting the ideal posture calculated for passive parallax compensation for each SAM lens module in optical simulation into a machinable geometric feature. Once these angular and spacing parameters are precisely machined and solidified, it means that the optical axis of each compensation module will automatically achieve the required orientation after assembly, without subsequent adjustment. After machining, a high-precision coordinate measuring machine or laser interferometer must be used to detect the actual angle of each mounting position to ensure that the error between it and the design value is within the allowable range. In addition, the relative position (spacing) between the mounting positions must also be strictly controlled.

[0024] Image fusion is the final step in the process, and the fusion algorithm is based on a preset geometric mapping relationship. This relationship is determined during the system design phase and stems from the fixed spatial transformation of each compensation module relative to the reference module (determined by the spacing and preset tilt angle). For a given object distance (e.g., the average distance of the target scene), the coordinates from the reference image can be derived. To the Image coordinates of each compensation module The mapping function. The specific form of this function depends on the camera imaging model and typically includes perspective projection terms and correction terms that consider lens distortion. To describe this mapping relationship more precisely, we introduce the following mathematical expression. Let the focal length of the reference module be... , No. The center distance between each compensation module and the reference module is (In an equally spaced layout, all spacing can be 20mm), the first The preset fixed tilt angles around the X, Y, and Z axes for each compensation module are as follows: The pixel size of the imaging sensor is The object distance to the target point is Under ideal optical conditions, neglecting lens distortion, the target point is at the [missing information]. The pixel coordinates of each compensation module on the imaging plane The coordinates of the reference image can be obtained through the following mapping formula. The calculation shows that: , The physical meaning of this formula is that the second term on the right side of the formula represents the compensation for the module spacing. and tilt angle The resulting parallax necessitates adjustments in pixel coordinates. Through customized SAM lens optical design and angle presets, the system has ensured that the optical axis orientation of each module satisfies a specific geometric relationship at the hardware level, thus naturally achieving parallax compensation during imaging. This mapping formula is a mathematical description of this geometric relationship.

[0025] In practical optical systems, lens distortion is unavoidable. To achieve higher pixel-level alignment accuracy, distortion correction needs to be considered during the mapping process. This assumes the reference lens has the same distortion coefficient as all custom SAM lenses. Then the corrected exact mapping formula can be expressed as: , The terms within square brackets represent the distortion correction for the reference image coordinates. Since all lenses exhibit consistent distortion characteristics, this correction can be uniformly applied to the reference image. Then, the images from each compensation module are mapped to the corrected reference image coordinates using the aforementioned formula, thereby achieving precise alignment even while considering distortion.

[0026] In actual fusion, the image processing unit (which can be an embedded CPU, GPU, or FPGA) uses the reference image as a reference for each of its output pixel positions. The source pixel coordinates in each compensation module image can be quickly calculated using the mapping function described above. Since coordinates may be sub-pixel accurate, bilinear interpolation is typically used to obtain pixel values. Finally, according to predetermined weighting coefficients, the pixel values ​​of the reference image and the mapped pixel values ​​of each compensation module image are weighted and synthesized to obtain the final disparity-free multispectral image. This weighted fusion process can be expressed by the following formula: , in, For the fused multispectral image, As the reference image, For the first The image after coordinate mapping of each compensation module These are the weighting coefficients for the reference image and the images of each compensation module. These coefficients can be preset according to the importance of each spectral band and satisfy the following conditions: The entire fusion process avoids time-consuming image feature matching, searching, and other disparity correction algorithms, resulting in short processing latency and meeting real-time processing requirements.

[0027] In summary, the present invention proposes a multi-module parallax-free spectral imaging method based on SAM lenses. By constructing an imaging system with a reference module as the center and multiple customized SAM lens modules arranged around it, and based on the fixed spatial position and fixed three-dimensional tilt angle of each module relative to the reference, the parallax between multiple modules is completely eliminated from the optical structure level.

[0028] Utilizing the optical characteristics of SAM lenses, horizontal, vertical, and axial parallax caused by module spacing can be passively compensated during imaging based on their placement within the array. All key optical parameters of the SAM lenses are strictly matched to those of the reference lens. Combined with a pixel-level spectral fusion method based on a preset geometric mapping relationship, multi-band images exhibit high consistency at the imaging source, eliminating the need for complex parallax correction calculations and significantly improving the real-time performance of image fusion and the overall system response speed.

[0029] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0030] Furthermore, in this invention, descriptions involving terms such as "first," "second," and "a" are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0031] In this invention, unless otherwise explicitly specified and limited, the terms "connection," "fixed," etc., should be interpreted broadly. For example, "fixed" can mean a fixed connection, a detachable connection, or an integral part; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0032] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible for those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

Claims

1. A multi-module parallax-free spectral imaging method based on a Sham lens, characterized in that, The system comprises imaging modules arranged in two rows and three columns. The middle imaging module in the first row is a reference module equipped with a standard industrial lens. The remaining five imaging modules are compensation modules equipped with customized SAM lenses. The SAM lenses of each compensation module are optically customized according to their predetermined offset position relative to the reference module, and are set with fixed tilt angles around the X, Y, and Z axes to perform passive three-dimensional parallax compensation. The specific steps include: S1: Establish the reference module as the image reference benchmark for the entire imaging module system; S2: Control the reference module and all compensation modules to synchronously acquire images of the target area, with each imaging module acquiring images of different spectral bands; S3: Based on the geometric mapping relationship determined by the fixed spatial position of each compensation module relative to the reference module and its preset fixed tilt angle, the images acquired by each compensation module are directly fused with the images acquired by the reference module at the pixel level. S4: Outputs a fused, disparity-free multispectral image.

2. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 1, characterized in that, The spacing between adjacent imaging modules is equal.

3. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 1, characterized in that, The fixed tilt angles for rotation around the X-axis, Y-axis, and Z-axis, and the specific rotation requirements are as follows: In the first row, the first and third imaging modules have a horizontal displacement relative to the reference module, and the horizontal parallax in the X-axis direction needs to be compensated. In the second row, the first, second, and third imaging modules have vertical and axial offsets relative to the reference module, requiring compensation for vertical parallax in the Y-axis direction and axial parallax in the Z-axis direction. At the same time, the first and third imaging modules also have horizontal displacements relative to the reference module, requiring compensation for horizontal parallax in the X-axis direction.

4. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 3, characterized in that, In the first row, the first and third imaging modules compensate for the horizontal parallax in the X-axis direction, and in the second row, the first and third imaging modules compensate for the horizontal parallax in the X-axis direction, using the following formula: In the formula, A fixed tilt angle is required to compensate for horizontal parallax in the X-axis direction. The horizontal spacing between adjacent imaging modules in the same row. For the lens field of view, For the lens focal length, For imaging distance, This is the field of view correction factor. This is the distortion correction factor.

5. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 4, characterized in that, In the second row, the first, second, and third imaging modules compensate for the vertical parallax in the Y-axis direction and the axial parallax in the Z-axis direction using the following formula: In the formula, A fixed tilt angle is required to compensate for vertical parallax in the Y-axis direction. A fixed tilt angle is required to compensate for axial parallax in the Z-axis direction. To optimize the synergistic amplification factor, The vertical spacing between the two rows of imaging modules. The vertical distance between the second row of imaging modules and the first row of imaging modules.

6. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 1, characterized in that, The customized optical design specifically means that the focal length of the SAM lens is the same as that of a regular industrial lens.

7. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 1, characterized in that, The customized optical design specifically means that the distortion coefficient of the SAM lens is consistent with that of a regular industrial lens.

8. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 1, characterized in that, In step S3, the pixel-level spectral fusion specifically involves: Image pixel coordinates acquired by the reference module Using the reference image, the first image is determined through a predetermined mapping function. The pixel coordinates of the corresponding object point in the image of each compensation module Pixel-level spectral fusion is performed between the coordinate-mapped images of each compensation module and the reference image. The parameters of the mapping function include: the distance between the current compensation module and the reference module, and the preset fixed tilt angles of the current compensation module around the X, Y, and Z axes.

9. The multi-module parallax-free spectral imaging method based on a Sham lens as described in claim 8, characterized in that, In step S3, the pixel-level spectral fusion adopts a weighted fusion method, expressed by the formula: In the formula, For the fused multispectral image, As the reference image, For the first The image after coordinate mapping of each compensation module These are the preset weighting coefficients.

Images