A common field of view optical path adjustment method and system for a dual push-scan hyperspectral camera

Abstract

The present invention discloses a method and system for adjusting the common field of view optical path of dual push-scan hyperspectral cameras, comprising: fixing a collimator, adjusting the optical axes of two hyperspectral cameras to be parallel to the optical axis of the collimator, and pre-adjusting the slits of the two hyperspectral cameras to be parallel to the ground; arbitrarily selecting one hyperspectral camera and performing coarse and fine adjustments in sequence; aligning the first hyperspectral camera with the target image, fixing the position of the first hyperspectral camera, and coarse and fine adjusting the other hyperspectral camera in sequence based on the adjustment result of the first hyperspectral camera to achieve a common field of view for the two hyperspectral cameras, and performing multi-dimensional precise adjustment of the hyperspectral cameras through a step-by-step progressive angle optimization strategy, thereby effectively solving the problem of insufficient precision of traditional adjustment methods and achieving high-precision common field of view adjustment.

Description

Technical Field

[0001] The present invention relates to the technical field of common-field-of-view optical path adjustment for a hyperspectral camera, and in particular to a method and system for adjusting the common-field-of-view optical path for a dual-push-scan hyperspectral camera. Background Art

[0002] In the field of spectral detection, push-broom hyperspectral cameras covering the 400-2500nm spectral range play a key role in scientific research, industry, and environmental monitoring. However, due to current limitations in optical design and device performance, manufacturing a single camera that covers this wide spectral range is extremely difficult. Therefore, the solution of using two cameras to create a common field of view and stitch the spectral data is widely used.

[0003] Common field of view (COV) systems primarily include coaxial beam splitting and parallel field of view installation. Coaxial beam splitting relies on specialized beam splitting elements to ensure consistent field of view and precise image registration, facilitating spectral stitching. However, light passing through the beam splitting element suffers significant light loss due to reflection, scattering, and absorption, reducing the camera's ability to detect weak spectral signals. Furthermore, the large space required to accommodate the elements and the complex optical path limits its application in scenarios with strict device size requirements, such as drone-mounted imaging.

[0004] The parallel field-of-view mounting structure reduces optical design complexity, keeps costs manageable, and minimizes light loss. However, it presents significant challenges in assembly and adjustment. The tightly interconnected optical systems of this structure place extremely high demands on the spatial layout precision of the parallel cameras. Both horizontal and vertical deviations must be kept to extremely small limits; even the slightest inaccuracy can cause image misalignment in the stitched field-of-view area. Furthermore, during debugging, it can be difficult to determine whether the cameras are mounted perfectly parallel. Due to the lack of high-precision and convenient inspection methods, relying solely on traditional measurement tools can result in significant errors. Summary of the Invention

[0005] In response to the problems in the related art, the present invention proposes a common field of view optical path adjustment method and system for a dual push-scan hyperspectral camera to achieve high-precision common field of view adjustment of a parallel installation structure.

[0006] To this end, the specific technical solutions adopted in the present invention are as follows:

[0007] A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera, comprising:

[0008] S1: Fix the collimator, adjust the optical axes of the two hyperspectral cameras to be parallel to the optical axis of the collimator, and pre-adjust the slits of the two hyperspectral cameras to be parallel to the ground;

[0009] S2: Randomly select a hyperspectral camera and perform coarse and fine adjustments in sequence;

[0010] The coarse adjustment of the first hyperspectral camera includes: adjusting the initial pitch angle, initial rotation angle, and initial sweep angle of the first hyperspectral camera, preliminarily eliminating stray light interference, and making the first hyperspectral camera pre-aligned with the target image;

[0011] Fine tuning of the first hyperspectral camera includes:

[0012] S2.1: Fix the initial pitch angle and initial rotation angle, obtain the position corresponding to the central reference point of the target image, calculate the average offset of the first hyperspectral camera, and adjust the sweep angle based on the average offset;

[0013] S2.2: Fixedly adjust the pan angle and preliminary rotation angle, adjust the preliminary pitch angle again to eliminate stray light interference and obtain the adjusted pitch angle;

[0014] S2.3: Fixedly adjust the pan and tilt angles, adjust the initial rotation angle, and eliminate stray light interference to obtain the adjusted rotation angle.

[0015] S2.4: Determine whether the adjustment step sizes of the pitch angle, rotation angle, and sweep angle of the first hyperspectral camera are less than a set threshold. If not, loop through S2.1-S2.3 until the adjustment step sizes of the three are less than the set threshold.

[0016] S3: After the first hyperspectral camera is aligned with the target image, the position of the first hyperspectral camera is fixed, and based on the adjustment result of the first hyperspectral camera, the other hyperspectral camera is coarsely adjusted and finely adjusted in sequence to achieve a common field of view for the two hyperspectral cameras.

[0017] This method for adjusting the common field of view optical path of a dual-push-scan hyperspectral camera uses a step-by-step progressive angle optimization strategy to perform multi-dimensional precise adjustment of the hyperspectral camera, effectively solving the problem of insufficient precision of traditional adjustment methods and achieving high-precision common field of view adjustment.

[0018] Furthermore, the adjusted sweep angle is the difference between the initial sweep angle and the correction term. The correction term is the product of the sweep angle fixed coefficient, the sign function of the average offset and the natural logarithm function of the average offset. The sweep angle fixed coefficient is obtained through experimental calibration fitting and is used to control the amplitude of the correction. The sign function is used to control the adjustment direction. The natural logarithm function provides a nonlinear scaling factor to control the amplitude of the adjustment.

[0019] Furthermore, the focal length of the collimator is F, and the collimator includes a graticule. The graticule is circular and has two rectangular light holes and two square identification holes. The two light holes are symmetrically arranged on the graticule. The spacing between the two light holes is b, the length of the light hole is x, the width of the light hole is consistent with the side length of the identification hole, and the two identification holes are arranged on both sides of the middle of the light hole. The side length of the identification hole is d, and the spacing between the identification hole and the light hole is c, c=d, b=3d, x=5d, d=s×F / f, where s is the slit width of the hyperspectral camera, and f is the focal length of the hyperspectral camera lens;

[0020] During the coarse adjustment of the first hyperspectral camera, light passes through the reticle and generates two light beams through the light hole. The first hyperspectral camera is swung. When the double white stripe image formed by the two light beams is detected and the white signal-to-noise ratio is greater than the set threshold, the camera's position and posture are locked to obtain the preliminary pitch angle, preliminary rotation angle, and preliminary sweep angle.

[0021] Furthermore, in S2.1, the coordinates of the central reference point of the double white stripe image are used to calculate the average offset of the first hyperspectral camera, and the panning angle is adjusted based on the average offset. In S2.2, the total brightness of the two beams is defined as B = b1 + b2, where b1 and b2 are the brightness of the two beams, respectively. The brightness change of the two beams corresponding to the adjusted pitch angle and the preliminary pitch angle is ΔB, and the brightness contrast change of the two beams is ΔC. When adjusting the pitch angle, adjustment continues in the same direction as B and ΔC increase. When the brightness approaches the extreme value, the automatic step adjustment is decelerated. When the adjusted pitch angle and the preliminary pitch angle are less than a threshold, the adjustment is completed to obtain the adjusted pitch angle. In S2.3, the absolute value of the brightness difference between the two beams is Δb. The adjusted rotation angle is proportional to Δb, and the sign function of Δb is used to control the adjustment direction. When the brightness difference between the two beams increases, the rotation is rotated toward the beam with higher brightness. When Δb is less than a threshold, the adjustment is completed to obtain the adjusted rotation angle. In S2.4, during the loop S2.1-S2.3, when the adjustment step of the pitch angle, rotation angle, and pan angle of the first hyperspectral camera is less than the set threshold, the position and posture of the camera are locked to obtain the final pitch angle, rotation angle, and pan angle.

[0022] Furthermore, the coarse adjustment of the second hyperspectral camera includes: adjusting the center height of the second hyperspectral camera to be consistent with the center height of the collimator, swinging the second hyperspectral camera, and when the double white stripe image formed by the two light beams is detected and the white signal-to-noise ratio is greater than a set threshold, locking the camera's position and posture, and obtaining the preliminary pitch angle, preliminary rotation angle, and preliminary pan angle of the second hyperspectral camera. The fine adjustment of the second hyperspectral camera includes: adjusting the center line of the double white stripes in the image and the spacing between the center lines of the double white stripe image, adjusting the pitch angle, rotation angle, and pan angle of the second hyperspectral camera using the same method as S2.1-S2.4 until the spacing is less than the set threshold, and then locking the camera's position and posture.

[0023] The present application also provides a dual-push-scan hyperspectral camera common field of view optical path system for realizing the above-mentioned dual-push-scan hyperspectral camera common field of view optical path adjustment method, comprising: a collimator, a first hyperspectral camera, a second hyperspectral camera and a display system, the collimator comprising a graticule, the graticule is circular, and has two rectangular light holes and two square identification holes, the two light holes are symmetrically arranged on the graticule, the spacing between the two light holes is b, the length of the light hole is x, the width of the light hole is consistent with the side length of the identification hole, and the two identification holes are symmetrically arranged on the graticule. The identification holes are arranged on both sides of the middle part of the light-through hole. The side length of the identification hole is d. The distance between the identification hole and the light-through hole is c, c=d, b=3d, x=5d, d=s×F / f, where s is the slit width of the hyperspectral camera, and f is the focal length of the hyperspectral camera lens. The first hyperspectral camera and the second hyperspectral camera are symmetrically arranged on one side of the optical path of the reticle. The height, pitch angle, rotation angle and pan angle of the first hyperspectral camera and the second hyperspectral camera are adjustable. The display system is used to display the images of the first hyperspectral camera and the second hyperspectral camera.

[0024] The beneficial effects of the present invention are:

[0025] 1. This dual-push-scan hyperspectral camera common field of view optical path adjustment method uses a step-by-step progressive angle optimization strategy to perform multi-dimensional precise adjustment of the hyperspectral camera, effectively solving the problem of insufficient accuracy of traditional adjustment methods and achieving high-precision common field of view adjustment.

[0026] 2. This common field of view optical path adjustment method for a dual-push-scan hyperspectral camera uses small steps to fine-tune the pitch, rotation, and pan angles, determining the optimal angle based on light intensity changes and set thresholds. For example, in pitch angle fine-tuning, the angle corresponding to maximum light intensity is determined based on the light intensity and angle curve. This effectively solves the problem of insufficient accuracy of traditional adjustment methods and achieves high-precision common field of view adjustment.

[0027] 3. To address the difficulty of aligning parallel field-of-view structures, this dual-push-scan hyperspectral camera's common-field optical path system utilizes a custom reticle and collimator to create a precise reference system. The reticle's unique pattern and aperture design, combined with the collimator's stable parallel beam, provide a clear basis for adjustment. Deep learning algorithms automatically identify characteristic areas on the reticle, enabling rapid coarse adjustment. During fine adjustment, the adjustment components are driven by quantitative data, avoiding the traditional over-reliance on manual experience and simple tools, significantly reducing difficulty and improving efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0029] Figure 1 is a flow chart of the method steps of embodiment 1 of the present invention;

[0030] Figure 2 A schematic structural diagram of a reticle according to a first embodiment of the present invention;

[0031] Figure 3 2 is a schematic diagram of an image after light passes through a reticle according to a first embodiment of the present invention;

[0032] Figure 4 Schematic diagram of the structure of the common field of view optical path system of the dual push-scan hyperspectral camera according to the second embodiment of the present invention.

[0033] In the picture:

[0034] 1. Light source; 2. Frosted glass; 3. Reticle; 4. Collimating lens; 5. First hyperspectral camera; 6. Second hyperspectral camera; 7. Lifting platform. DETAILED DESCRIPTION

[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0036] Example 1

[0037] A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera, such as Figure 1 Shown, including:

[0038] S1: Fix the collimator and adjust the optical axes of the two hyperspectral cameras to be roughly parallel to the optical axis of the collimator, and as close as possible within the manually adjustable range. The lenses of the two hyperspectral cameras are located near the center of the collimator's illumination area, and the collimator is located near the center of the field of view of the two hyperspectral cameras. The slits of the two hyperspectral cameras are pre-adjusted to be parallel to the ground.

[0039] S2: Randomly select a hyperspectral camera and perform coarse and fine adjustments in sequence;

[0040] The coarse adjustment of the first hyperspectral camera includes: adjusting the initial pitch angle, initial rotation angle, and initial sweep angle of the first hyperspectral camera, preliminarily eliminating the stray light interference caused by the reflection of the inner wall of the collimator, and making the first hyperspectral camera pre-aligned with the target image. Specifically:

[0041] The light passes through the reticle 3 of the collimator, as shown in Figure 2 As shown, the collimator includes a graticule 3, which is circular and has two rectangular light holes and two square identification holes. The two light holes are symmetrically arranged on the graticule 3, with a spacing of b between the two light holes, a length of x, and a width of the light hole consistent with a side length of the identification hole. The two identification holes are arranged on both sides of the middle of the light hole, with a side length of d, and a spacing of c between the identification holes and the light holes, c=d, b=3d, x=5d, and d=s×F / f, where s is the slit width of the hyperspectral camera, f is the focal length of the hyperspectral camera lens, and the focal length of the collimator is F.

[0042] A dual-screen real-time monitoring system is established to display the projection features of the reticle 3 collected by the spectral camera in real time. When the first hyperspectral camera is coarsely adjusted, the light passes through the reticle 3 and generates two beams through the light hole, such as Figure 3 As shown, the dual-screen real-time monitoring system displays the double white stripe image captured by the first hyperspectral camera. At this time, the first hyperspectral camera is swung. When the double white stripe image formed by the two light beams is detected and the white signal-to-noise ratio is greater than the set threshold, the camera's position and posture are locked to obtain the preliminary pitch angle, preliminary rotation angle, and preliminary sweep angle. This is because when light passes through the reticle 3, the hyperspectral camera is pitched up and down, and a large area of ​​bright light and two white stripe areas appear as the target bright area. Therefore, the reticle 3 can be used to eliminate stray light interference, so that the first hyperspectral camera is pre-aligned with the target image.

[0043] Fine tuning of the first hyperspectral camera includes:

[0044] S2.1: Fine-tune the sweep angle, fix the initial pitch angle and initial rotation angle, and obtain the coordinates of the central reference point of the double white stripe image, such as Figure 3As shown, the horizontal direction is the x-axis (number of pixel columns), and the vertical direction is the y-axis (number of spectral bands). If the number of image pixel columns is n, then the center reference point is x0=n / 2. The average offset Δx of the camera can be calculated through the coordinates x1 and x2 corresponding to the center reference point. , calculate and adjust the sweep angle according to the average offset Δx, the calculation process is as follows:

[0045] , when the stripe deviates from the center, the horizontal angle is adjusted according to the logarithmic function of the offset, where, To adjust the sweep angle, is the initial sweep angle, k h It is a fixed coefficient between pixels and sweep angles, obtained through experimental calibration, such as taking k h = 0.1 rad / pixel; sng(Δx) is the sign function, and the direction of adjustment is determined by the offset. This is an optimization term to avoid excessive adjustments when Δx is too large. When Δx < 1, the distance between the center reference point of the double white stripe image and the image sensor coordinate center is considered to be less than 1 pixel, and the current sweep angle is locked as the adjustment sweep angle.

[0046] S2.2: Fine-tune the pitch angle. Fixedly adjust the sweep angle and the initial rotation angle. Adjust the initial pitch angle to eliminate the stray light interference caused by the reflection of the inner wall of the collimator and obtain the adjusted pitch angle. Specifically:

[0047] The total brightness of the two beams is defined as B=b1+b2, where b1 and b2 are the brightness of the two beams respectively. The brightness change of the two beams corresponding to the pitch angle is ΔB, ΔB=B current -B previous , where B current is the total brightness of the two beams corresponding to the current pitch angle adjustment, B previous is the total brightness of the two beams corresponding to the previous pitch angle adjustment, and the brightness contrast of the two beams is C. , the change in brightness contrast between the two beams is ΔC, ΔC=C current -C previous , C current is the current brightness contrast between the two beams, C previous To improve the brightness contrast between the two beams, adjust the pitch angle proportional to ΔB. Specifically:

[0048] ,in, To adjust the pitch angle, is the initial pitch angle, k pC and k pBis the adjustment coefficient, which is obtained through experimental calibration and fitting; the overall meaning of the formula is to continue adjusting in the same direction when the total brightness increases and the brightness contrast of the two light beams increases. If the total brightness and the light beam increase and the other decreases, the adjustment amount in the increasing direction and the decreasing direction is obtained according to the corresponding weight coefficients. After such an adjustment, the system will eventually stay in an area where both light beams exist and the overall brightness is higher. If only the total brightness of the stripes is used for control, it may happen that one stripe disappears but the brightness of the other stripe is higher. Adding contrast change as a feedback amount and assigning a suitable coefficient can increase the robustness of the control method without sacrificing accuracy. Coefficient k pB The unit is rad / (gray level). Coefficient k pC The unit is 1 / (gray level). When the threshold is reached, it is considered that the current optimal pitch angle has been reached and the current pitch angle is locked as the adjusted pitch angle.

[0049] S2.3: Fine-tune the rotation angle. Fixedly adjust the pan and tilt angles. Adjust the initial rotation angle to eliminate stray light interference and obtain the adjusted rotation angle. Specifically:

[0050] The absolute value of the brightness difference between the two beams is Δb, , adjust the rotation angle proportional to Δb, specifically, ,in, To adjust the rotation angle, is the initial rotation angle, k r is the adjustment coefficient, k r The value of Δb is 0.02-0.05 rad / (gray level), obtained through experimental calibration and fitting. sng(b1-b2) is the sign function, indicating the direction of the difference. The general meaning of the formula is that when the brightness difference between the two beams increases, the rotation is directed toward the brighter beam. When Δb < the threshold, the brightness difference between the two stripes is considered negligible, and the current rotation angle is locked as the adjustment rotation angle.

[0051] It should be noted that the adjustment order is horizontal, pitch, and rotation. Only one dimension is adjusted at a time, and the other dimensions are locked. If it is detected , pause and recalibrate the level.

[0052] S2.4: Determine whether the adjustment step sizes of the pitch angle, rotation angle, and sweep angle of the first hyperspectral camera are less than the set threshold. If not, loop through S2.1-S2.3 until the adjustment step sizes of the three are less than the set threshold. Specifically:

[0053] During the loop S2.1-S2.3, when the adjustment step of the pitch angle, rotation angle, and sweep angle of the first hyperspectral camera is less than the set threshold (such as 0.01°), the position and posture of the camera are locked to obtain the final pitch angle, rotation angle, and sweep angle. At this time, it is determined that the first hyperspectral camera is aligned with the target image.

[0054] S3: After the first hyperspectral camera is aligned with the target image, the position of the first hyperspectral camera is fixed. Based on the adjustment result of the first hyperspectral camera, the other hyperspectral camera is coarsely and finely adjusted in sequence to achieve a common field of view for the two hyperspectral cameras. Specifically:

[0055] The coarse adjustment of the second hyperspectral camera includes: while the image on the dual-screen real-time monitoring system remains stable as the two white stripes image, the center height of the second hyperspectral camera is adjusted to be consistent with the center height of the parallel light tube; when the double white stripe image formed by the two light beams is detected and the white signal-to-noise ratio is greater than the set threshold, the position and posture of the camera are locked to obtain the preliminary pitch angle, preliminary rotation angle and preliminary sweep angle of the second hyperspectral camera; the fine adjustment of the second hyperspectral camera includes: adjusting the center line of the double white stripes in the image and the spacing between the center lines of the double white stripe image; the pitch angle, rotation angle and sweep angle of the second hyperspectral camera are adjusted using the same method as S2.1-S2.4 until the spacing is less than the set threshold (such as 0.01°), which means that the second hyperspectral camera has been adjusted and the two cameras have completed the common field of view adjustment.

[0056] Example 2

[0057] like Figure 4As shown, the present invention provides a dual-push-scan hyperspectral camera common field of view optical path system, comprising: a collimator, a first hyperspectral camera 5, a second hyperspectral camera 6, and a display system. The collimator comprises a light source 1, a frosted glass 2, a reticle 3, a collimating lens 4, the first hyperspectral camera 5, the second hyperspectral camera 6, and a lifting platform. The frosted glass 2, reticle 3, collimating lens 4, the first hyperspectral camera 5, and the second hyperspectral camera 6 are sequentially arranged along the optical path to one side of the light source 1. The light source 1, frosted glass 2, reticle 3, collimating lens 4 form the collimator, which has a focal length of F = 550 mm and an aperture of 50 mm. Light source 1 provides stable light and serves as the starting light source for the entire optical path system, providing basic illumination for subsequent optical operations. Frosted glass 2, located after light source 1, diffusely reflects the light emitted by light source 1, converting the concentrated light into uniform diffuse light. This ensures a more uniform light distribution in the subsequent optical path and reduces imaging errors caused by uneven light. The reticle 3 is placed behind the frosted glass 2 and features a specific pattern that serves as a reference for optical path adjustment. The reticle is circular with a diameter of 24.5 mm. It features two rectangular apertures and two square identification apertures. The two apertures are symmetrically positioned on the reticle 3, with a spacing of b and a length of x. The width of the apertures matches the side length of the identification apertures. The two identification apertures are positioned on either side of the central portion of the aperture, with a side length of d. The spacing between the identification apertures is c, where c = d, b = 3d, and x = 5d. D = s × F / f, where s is the slit width of the hyperspectral camera and f is the focal length of the hyperspectral camera lens. The collimating lens 4 collimates the light passing through the reticle 3, converting the divergent light into parallel light. Parallel light has better stability and predictability in optical path propagation, which is beneficial for subsequent optical path adjustment and camera imaging. The lifting platform 7 is mainly used to adjust the vertical height position of the hyperspectral camera. By adjusting the lifting platform 7, the hyperspectral camera can reach a suitable height in the vertical direction, thereby ensuring the consistency of the optical paths of the two hyperspectral cameras in the vertical direction. The first hyperspectral camera 5 and the second hyperspectral camera 6 are symmetrically arranged on one side of the optical path of the graticule 3. The height, pitch angle, rotation angle and pan angle of the first hyperspectral camera 5 and the second hyperspectral camera 6 are adjustable. The first hyperspectral camera 5 and the second hyperspectral camera 6 are connected to the lifting platform 7. Through this system, the height, pitch angle, rotation angle and pan angle of the first hyperspectral camera 5 and the second hyperspectral camera 6 can be adjusted to accurately align the two cameras with the common field of view. The display system is used to display the imaging of the first hyperspectral camera and the second hyperspectral camera.

[0058] The first and second hyperspectral cameras 5 and 6 serve as the core imaging components of the system. They receive light from the conditioned optical path, split it, and create an image on a CCD or CMOS device. The image information is presented via a display system connected to the CCD or CMOS device. This display system can be configured to display the information on two separate screens or on a single screen in split-screen format.

[0059] It should be noted that before adjusting the optical path, ensure that the reference plane of reticle 3 is perpendicular to the direction of gravity and the optical axis to establish the adjustment reference plane. When coarsely adjusting the first hyperspectral camera, secure the collimator and position the lifting platform 7 so that the center of the collimator's optical axis roughly overlaps, aligning the intersection of the optical axes of the two hyperspectral cameras with the collimator's optical axis.

[0060] In summary, this dual-push-scanner hyperspectral camera common field of view optical path system solves the problem of low installation efficiency caused by the complexity of the optical system of the parallel field of view installation structure and the extremely high requirements for spatial layout accuracy, which leads to high difficulty in installation and adjustment, strict requirements on the professional skills and practical experience of the installation and adjustment technicians.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera, characterized in that: include: S1: Fix the collimator, adjust the optical axes of the two hyperspectral cameras to be parallel to the optical axis of the collimator, and pre-adjust the slits of the two hyperspectral cameras to be parallel to the ground; S2: Randomly select a hyperspectral camera and perform coarse and fine adjustments in sequence; The coarse adjustment of the first hyperspectral camera includes: adjusting the initial pitch angle, initial rotation angle, and initial sweep angle of the first hyperspectral camera, preliminarily eliminating stray light interference, and making the first hyperspectral camera pre-aligned with the target image; Fine tuning of the first hyperspectral camera includes: S2.1: Fix the initial pitch angle and initial rotation angle, obtain the position corresponding to the central reference point of the target image, calculate the average offset Δx of the first hyperspectral camera, and adjust the sweep angle based on the average offset Δx; S2.2: Fixedly adjust the pan angle and preliminary rotation angle, adjust the preliminary pitch angle again to eliminate stray light interference and obtain the adjusted pitch angle; S2.3: Fixedly adjust the pan and tilt angles, adjust the initial rotation angle, and eliminate stray light interference to obtain the adjusted rotation angle. S2.4: Determine whether the adjustment step size of the pitch angle, rotation angle, and pan angle of the first hyperspectral camera is less than a set threshold. If not, loop through S2.1-S2.3 until the adjustment step size of the three is less than the set threshold. S3: After the first hyperspectral camera is aligned with the target image, the position of the first hyperspectral camera is fixed, and based on the adjustment result of the first hyperspectral camera, the other hyperspectral camera is coarsely and finely adjusted in sequence to achieve a common field of view for the two hyperspectral cameras; The adjustment sweep angle is the difference between the initial sweep angle and the correction term, and the correction term is the fixed coefficient k between the pixel and the sweep angle. h , the sign function of the average offset Δx and The product of pixels and the fixed coefficient k between the sweep angle h It is obtained through experimental calibration fitting and is used to control the amplitude of the correction. The sign function is used to control the adjustment direction. The natural logarithm function provides a nonlinear scaling factor to control the amplitude of the adjustment.

2. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 1, characterized in that: The focal length of the collimator is F. The collimator includes a circular graticule with two rectangular light holes and two square identification holes. The two light holes are symmetrically arranged on the graticule. The spacing between the two light holes is b. The length of the light hole is x. The width of the light hole is consistent with the side length of the identification hole. The two identification holes are arranged on both sides of the middle of the light hole. The side length of the identification hole is d. The spacing between the identification hole and the light hole is c, c = d, b = 3d, x = 5d, and d = s × F / f, where s is the slit width of the hyperspectral camera and f is the focal length of the hyperspectral camera lens. During the coarse adjustment of the first hyperspectral camera, light passes through the reticle and generates two light beams through the light hole. The first hyperspectral camera is swung. When the double white stripe image formed by the two light beams is detected and the white signal-to-noise ratio is greater than the set threshold, the camera's position and posture are locked to obtain the preliminary pitch angle, preliminary rotation angle, and preliminary sweep angle.

3. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 2, characterized in that: In S2.1, the coordinates of the central reference point of the double white stripe image are used to calculate the average offset of the first hyperspectral camera, and the sweep angle is adjusted based on the average offset calculation.

4. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 3, characterized in that: In S2.2, the total brightness of the two light beams is defined as B=b1+b2, where b1 and b2 are the brightness of the two light beams, respectively. The brightness change of the two light beams corresponding to the pitch angle adjustment is ΔB, and the change in the brightness contrast of the two light beams is ΔC. When adjusting the pitch angle, as B and ΔC increase, adjustment continues in the same direction. When the brightness approaches the extreme value, the automatic step adjustment decelerates. When the difference between the adjusted pitch angle and the initial pitch angle is less than a threshold, the adjustment is completed and the adjusted pitch angle is obtained.

5. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 4, characterized in that: In S2.3, the absolute value of the brightness difference between the two light beams is Δb. The adjustment rotation angle is proportional to Δb, and the sign function of Δb is used to control the adjustment direction. When the brightness difference between the two light beams increases, the light beam rotates toward the light beam with higher brightness. When Δb is less than a threshold, the adjustment ends and the adjustment rotation angle is obtained.

6. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 5, characterized in that: In S2.4, during the loop S2.1-S2.3, when the adjustment step of the pitch angle, rotation angle, and pan angle of the first hyperspectral camera is less than the set threshold, the position and posture of the camera are locked to obtain the final pitch angle, rotation angle, and pan angle.

7. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 6, characterized in that: The coarse adjustment of the second hyperspectral camera includes: adjusting the center height of the second hyperspectral camera to be consistent with the center height of the collimator, swinging the second hyperspectral camera, and locking the camera's position and posture when a double white stripe image formed by the two light beams is detected and the white signal-to-noise ratio is greater than a set threshold, thereby obtaining the preliminary pitch angle, preliminary rotation angle, and preliminary sweep angle of the second hyperspectral camera.

8. A method for adjusting the common field of view optical path of a dual push-scan hyperspectral camera according to claim 7, characterized in that: Fine adjustment of the second hyperspectral camera includes adjusting the center line of the double white stripes in the image and the distance between the center lines of the double white stripes image. The same method as S2.1-S2.4 is used to adjust the pitch angle, rotation angle, and sweep angle of the second hyperspectral camera until the distance is less than the set threshold, and the position and posture of the camera are locked.

9. A dual-push-scan hyperspectral camera common field of view optical path system, used to implement the dual-push-scan hyperspectral camera common field of view optical path adjustment method according to any one of claims 1 to 8, characterized in that: include: A collimator, a first hyperspectral camera, a second hyperspectral camera, and a display system. The collimator includes a graticule. The graticule is circular and has two rectangular light holes and two square identification holes. The two light holes are symmetrically arranged on the graticule. The spacing between the two light holes is b, the length of the light hole is x, and the width of the light hole is consistent with the side length of the identification hole. The two identification holes are arranged on both sides of the middle of the light hole. The side length of the identification hole is d. The spacing between the identification hole and the light hole is c, c=d, b=3d, x=5d, and d=s×F / f, where s is the slit width of the hyperspectral camera, f is the focal length of the hyperspectral camera lens, and F is the focal length of the collimator. The first hyperspectral camera and the second hyperspectral camera are symmetrically arranged on one side of the optical path of the graticule. The height, pitch angle, rotation angle, and pan angle of the first hyperspectral camera and the second hyperspectral camera are adjustable. The display system is used to display the images of the first hyperspectral camera and the second hyperspectral camera.

Images