Single-point macro high spectral system and method

Abstract

This invention discloses a single-point, large-field-of-view macro hyperspectral system and method, including an imaging component, a rotation module, a linear movement module, an object plane, and a control module. The imaging component includes a light source, a single-point spectrometer, a mirror module, and a lens module. The control module adjusts the rotation speed of the rotation module and the linear movement distance of the imaging component based on the initial input scanning range to collect spectra until the recorded spectral data covers the entire area of ​​the object to be imaged. Then, the collected data is synthesized into three-dimensional image data. The large-field-of-view imaging is achieved in one step through rotation, without the need for complex stitching steps. The single-point, single-light source design ensures uniform imaging light for each pixel, ensuring stable imaging quality. After the movement and rotation imaging, there is no need for image stitching or feature point matching, effectively solving the stitching difficulties caused by the lack of feature points in traditional technologies. This enables efficient and high-quality imaging of a large field of view.

Description

Technical Field

[0001] This invention relates to the field of hyperspectral technology, specifically to a single-point, large-field-of-view, micro-hyperspectral system and method. Background Technology

[0002] Hyperspectral imaging technology, as a core means of acquiring fine features of objects in fields such as materials testing and biological sample analysis, is an indispensable key technology in scientific research and industrial production. The inability to acquire high-quality, large-field-of-view spectral data in macro and microscopic scenarios will directly lead to biased detection results and inaccurate analytical conclusions, seriously affecting research and production decisions. Therefore, achieving efficient and accurate data acquisition is extremely crucial.

[0003] To address this need, existing hyperspectral imaging and data processing equipment has been upgraded through hardware and optimized through algorithms to enhance data acquisition and processing capabilities in macro and microscopic scenarios, aiming to provide reliable data support for various fields. However, traditional hyperspectral technology still has significant shortcomings in practical applications: First, the limitations of macro field of view and the challenges of data stitching coexist. Traditional hyperspectral macro lenses have extremely small single-shot ranges, requiring multiple shots and complex algorithms to stitch together complete large-field-of-view data. This not only significantly increases data processing costs but also leads to data discontinuity and misalignment in areas with uniform sample surface texture and similar spectral characteristics due to a lack of effective matching points. Second, image stitching quality severely affects data usability. Hyperspectral images generated by line scanning technology have obvious seams and significant differences in brightness between scan lines, resulting in poor consistency of the stitched image. When scanning by region, the blurred features of edge areas make it difficult for the equipment to accurately match, causing problems such as texture breaks and boundary misalignment in the stitched image. Third, the contradiction between scanning efficiency and positioning accuracy is prominent. Taking microscopic line scanning with a 20x objective lens as an example, scanning a 1.5cm × 1.5cm area takes 48 minutes, and it is difficult to accurately locate the sample area. Key data are often missed or repeated scanning is caused by scanning position deviation, which seriously restricts the timeliness of application. These technical bottlenecks have become the core problems hindering the widespread application of hyperspectral technology in macro and microscopic scenarios. Summary of the Invention

[0004] This invention provides a single-point large field-of-view macro hyperspectral system and method, which solves the problems of narrow field of view, obvious seams in image stitching, difficulty in feature point matching, and low scanning efficiency of traditional hyperspectral macro lenses, which make it difficult to quickly acquire high-quality large field-of-view hyperspectral data.

[0005] This invention provides the following technical solution: a single-point, large-field-of-view macro hyperspectral system, comprising: an imaging component, a rotation module, a linear motion module, an object plane, and a control module. The imaging component includes: a light source, a single-point spectrometer, a mirror module, and a lens module. The control module is electrically connected to the imaging component, the rotation module, and the linear motion module to achieve data communication. The light source is arranged at the beginning of the system's optical path to provide stable and suitable illumination for imaging. The single-point spectrometer is connected to the lens module to achieve spectral imaging. The mirror module is positioned between the lens module and the object plane carrying the object to be imaged. The lens module is installed on the optical path of the system on the side closest to the object plane, and the height of the lens module is adjustable. The imaging component is slidably connected to one side of the linear motion module, and the rotation module is fixedly connected to the linear motion module. Based on the initial input scanning range, the control module adjusts the rotation speed of the rotation module and the linear motion distance of the imaging component to collect spectra until the recorded spectral data covers the entire area of ​​the object to be imaged, and then synthesizes the collected data into three-dimensional image data.

[0006] The light source can flexibly adjust its luminescence characteristics, such as spectral range and light intensity, according to the application requirements of macro and microscopic scenes, providing a stable and suitable lighting foundation for imaging and ensuring the lighting conditions for object plane imaging. The control module, as the system's "control center," sends commands to coordinate the spectral acquisition, rotation, and linear motion modules of the single-point spectrometer, ensuring synchronous operation of all components and achieving large-field-of-view rapid imaging and stable spectral detection. The single-point spectrometer is connected to the lens module via optical fiber, receiving light signals from the object plane and performing spectral dispersion to obtain the spectral information of that pixel. It is connected to the control module via cable, receiving the acquired control information and transmitting the acquired data to the control module for subsequent processing. The reflector module is positioned between the lens and the object plane in the optical path. It can be adjusted to a fixed angle. When the imaging component moves to its closest position to the rotation module on the linear motion module's track, the optical reflector of the reflector module precisely reflects the light source processed by the lens to the center point of the object plane directly below the rotation module, and then reflects the light signal reflected from the center point back to the lens, which is then transmitted to the single-point spectrometer via optical fiber. By adjusting the optical path direction through a fixed reflector angle, the system can ensure a more comprehensive scanning range without missing any points, and it also makes the center of gravity of the moving components on the track more stable, without affecting data acquisition. The lens is mounted on the optical path closer to the object plane, and its height can be adjusted so that the lens optical axis passes through the center of the reflector, ensuring the integrity and stability of the reflected optical path.

[0007] Preferably, the rotating module is installed at the junction of the linear motion module's moving track and the main system structure. It receives commands from the control module and is driven to rotate by a built-in motor with an encoder, causing the entire device to rotate and acquire data. The encoder readings for each acquisition position are transmitted to the control module. The relative position of the rotating module with the moving track and the main system structure remains stable, ensuring accuracy during movement and imaging.

[0008] Preferably, a counterweight is slidably connected to the side of the linear motion module away from the imaging component. The counterweight is electrically connected to the control module. The rotation module is located between the imaging component and the counterweight. The counterweight balances the weight of the imaging-side structure with its own weight, ensuring the overall structure remains stable and without tilt or offset when the rotation module and the moving track move or remain stationary, maintaining optical path accuracy and imaging quality. The linear motion module and the rotation module together serve as the core components of the scanning process. The center of the linear motion module precisely coincides with the center of the rotation module, and the linear motion module rotates with the rotation module. It receives commands from the control module to control the linear movement of the imaging component and the counterweight on the moving track, and calculates the rotation radius based on the initial and current positions, transmitting the position information to the control module. Through precise transmission, in conjunction with the rotation module, rapid displacement control for large-field-of-view imaging is achieved, ensuring imaging area coverage and scanning efficiency.

[0009] An imaging method using a single-point, large-field-of-view, macro hyperspectral system includes:

[0010] S1: Place the object to be imaged on the object plane, input the initial scanning range into the control module. The initial scanning range is set based on the size of the object to be imaged. Calculate the initial scanning radius and the initial rotation speed of the rotating module based on the initial scanning range.

[0011] S2: Start the rotation module to rotate. After the rotation speed stabilizes, the single-point spectrometer begins to acquire spectra and stores the acquired data to the control module.

[0012] S3: After completing one cycle of spectral acquisition, the control module controls the imaging component to move and change the scanning radius, and continues to complete one cycle of spectral acquisition, repeatedly changing the scanning radius until the scanning radius is changed to 0.

[0013] S4: The control module synthesizes the collected data into three-dimensional image data.

[0014] Preferably, the initial rotation speed is the ratio of the number of sampling points of the single-point spectrometer to the number of acquisitions per second of the single-point spectrometer, and the number of sampling points is the ratio of the circumference of the initial scanning range to the field of view of the lens module.

[0015] Preferably, S4: The control module synthesizes the acquired data into 3D image data including:

[0016] S4.1 Calculate the canvas pixel size and the pixel coordinates of the canvas pixel center;

[0017] S4.2 Convert the polar coordinates of each sampling point to pixel coordinates;

[0018] S4.3. Use interpolation sampling to reconstruct the image and synthesize three-dimensional image data.

[0019] Preferably, the interpolation sampling method includes: rounding the pixel coordinates of any sampling point to obtain four neighboring pixels, calculating the horizontal and vertical distances from the sampling point to any neighboring pixel, calculating a weight function based on the horizontal and vertical distances, distributing the spectral data of the sampling point to the four neighboring pixels according to the weights, and finally weighted summing to obtain the spectral data of each sampling point.

[0020] The present invention has the following beneficial effects:

[0021] 1. This single-point large field-of-view macro hyperspectral system achieves one-time large field-of-view imaging through rotation, eliminating the need for complex stitching steps and avoiding various problems caused by stitching in traditional technologies. Furthermore, the single-point, single-light-source design ensures uniform imaging light for each pixel, guaranteeing stable imaging quality. On the other hand, the motion + rotation imaging eliminates the need for image stitching and feature point matching, effectively solving the stitching difficulties caused by the scarcity of feature points in traditional technologies. This enables efficient and high-quality imaging of large field-of-view areas.

[0022] 2. The single-point spectrometer used in this single-point large field-of-view macro hyperspectral system has a low cost, and can even be achieved through sensors and spectroscopic modules, which greatly reduces hardware investment. At the same time, the linear array detector of the single-point spectrometer is less difficult to optimize than the linear scanning area array detector, which further saves costs from the technical optimization level and can help companies reduce investment while ensuring imaging effect.

[0023] 3. This single-point, wide-field-of-view macro hyperspectral system can significantly improve imaging speed. Its rotating structure can achieve rapid operation, which, combined with the rapid characteristics of single-point imaging itself, greatly shortens the imaging time. It is significantly faster than the time consumed by traditional high-magnification objective lens line scanning, which can meet the needs of rapid image data acquisition in practical applications and improve overall work efficiency. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the system structure of Embodiment 1 of this application.

[0025] Figure 2 This is a schematic diagram of the mounting angle structure of the lens module in Embodiment 1 of this application.

[0026] Figure 3 This is a schematic diagram of the installation angle structure of the reflector module in Embodiment 1 of this application.

[0027] Figure 4This is a schematic diagram of the imaging method steps in Embodiment 2 of this application.

[0028] Figure 5 This is a schematic diagram of point selection during the interpolation sampling and image reconstruction process in Embodiment 2 of this application.

[0029] In the diagram: 1. Light source; 2. Single-point spectrometer; 3. Mirror module; 4. Lens module; 5. Rotation module; 6. Linear movement module; 7. Object plane; 8. Counterweight. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Example 1

[0032] Please see Figure 1 One type of single-point large field-of-view macro hyperspectral system includes an imaging component, a rotation module 5, a linear motion module 6, an object plane 7, a counterweight 8, and a control module. The imaging component includes a light source 1, a single-point spectrometer 2, a reflector module 3, and a lens module 4.

[0033] The imaging component and counterweight 8 are symmetrically slidably connected at both ends of the linear motion module 6, and the rotation module 5 is located in the middle of the linear motion module 6. The control module is electrically connected to the imaging component, the rotation module 5, and the linear motion module 6 to achieve data communication. The light source 1 is arranged at the beginning of the system's optical path to provide stable and suitable illumination for imaging. The single-point spectrometer 2 is connected to the lens module 4 to achieve spectral imaging. The reflector module 3 is set between the lens module 4 and the object plane 7 carrying the object to be imaged. The installation angle of the reflector module 3 is fixed. The installation angle of the reflector module 3 is such that when the imaging component slides to the closest point to the rotation module 5, the reflector module 3 reflects the light source processed by the lens module 4 to the center point of the object plane 7 located directly below the rotation module 5, and reflects the light signal reflected at the center point to the lens module 4. By fixing the reflector angle of the reflector module 3 and adjusting the optical path, the system's scanning range can be ensured to be more comprehensive without any missed points, and the center of gravity of the imaging component and counterweight 8 on the moving track of the linear motion module 6 is more stable and will not affect data acquisition.

[0034] It should be noted that, as Figure 2 As shown, after the height of the lens module 4 is adjusted, the optical axis of the lens module 4 passes through the center of the mirror module 3, and the rotation module 5 is rotated by a motor with a built-in encoder.

[0035] Due to equipment and scanning mirror installation requirements, the optical mirror of mirror module 3 needs to have a certain angle to achieve complete imaging. The angle is calculated as follows: Relevant parameter definitions: w is the minimum distance between the mirror and the center of rotation; h is the height from the imaging plane; the dashed line in the diagram represents the normal, and the mirror is perpendicular to the normal; angle b is the angle between the mirror and the water surface; and angle a is the angle between the incident light and the reflected light. The formula for calculating angle a is a = arctan(h / w), where the angle of incidence equals the angle of reflection. Figure 3 As shown, the angle b between the optical mirror and the water surface can be calculated as 90 + 0.5 * a, and the optical mirror of mirror module 3 is fixed at this angle. It is connected to light source 1 and single-point spectrometer 2 via a Y-type optical fiber, receiving the light source signal and transmitting it to mirror module 3. It also initially converges the light reflected from the object plane 7 collected by mirror module 3 and transmits it to single-point spectrometer 2 via the Y-type optical fiber. Models and other specifications can be selected according to the application scenario.

[0036] Example 2

[0037] This embodiment describes the imaging method of the single-point large-field-of-view macro hyperspectral system applied in Embodiment 1. Please refer to [link to previous document]. Figure 4 .

[0038] First, perform hardware initialization: initialize the connection of the single-point spectrometer 2 to ensure stable communication between the single-point spectrometer 2 and the control module; initialize the scanning device to put the rotation module 5 and the linear motion module 6 into a ready-to-work state; initialize the encoder connection in the rotation module 5 to ensure accurate transmission of angle signals; and initialize the control of the light source 1 to ensure that the light source 1 provides a stable light source output to the system.

[0039] Then, the control parameters are set. After hardware initialization, the object to be imaged and the reference plate are placed on the object plane 7, and the control parameters are input according to the size of the acquisition area. The scanning range is set to define the area for subsequent scanning operations; the field of view (i.e., pixel physical size) of the lens module 4 is input as the basis for calculating the relevant scanning parameters; the exposure time of the single-point spectrometer 2 is set to obtain clear spectral data. After the control module receives the command, the light source 1 emits light and preheats, the single-point spectrometer 2 loads the acquisition configuration, the rotation module 5 and the linear movement module 6 are initialized, the angle of the reflector module 3 is fixed, and the system enters the acquisition standby state. The control module generates control commands according to the acquisition range and the lens field of view.

[0040] After the system is debugged, data acquisition is performed. The initial scanning radius R = 0.5L (i.e., the farthest acquisition radius) is calculated based on the input scanning range L. The linear movement module 6 drives the imaging component to this radius and locks it.

[0041] Calculate the optimal scanning speed: The number of times the single-point spectrometer 2 collects data per second is denoted as F. When the scanning radius is R, the circumference length C = 2 × π × R (mm). According to the lens field of view φ mm, the circumference can be divided into P = C / φ sampling points. The optimal rotation speed V = P / F (revolutions / s).

[0042] The encoder-equipped motor built into the rotation module 5 starts, driving the imaging component to move in a circle around the current radius R. The counterweight 8 balances the structural weight to ensure the stability of the optical path during rotation. The encoder monitors the real-time rotational speed: it records the angle change θ within a time period T, and calculates the real-time rotational speed Vc = θ / T. If the optimal rotational speed V is not reached, it continues to wait until the optimal speed is reached and enters a constant speed state.

[0043] During data acquisition, light from light source 1 is transmitted through lens module 4 to reflector module 3, and after reflection, illuminates the sample / reference plate surface. The reflected light returns along the original path, is converged by lens module 4, and then transmitted to single-point spectrometer 2 via Y-shaped optical fiber. The encoder divides the angle evenly according to the number of sampling points P and records the real-time angle value. The linear motion module 6 synchronously records the current radius and the acquisition timestamp of each sampling point. The spectral data is correlated with the timestamp, angle, and radius parameters and then transmitted back to the control module. Data is synthesized, saved, and displayed according to the scanning radius and rotation angle until all spectral data is recorded. The data format is shown in Table 1 below.

[0044] Table 1:

[0045] .

[0046] After each complete ring-shaped acquisition, the control module sends a displacement command to the linear motion module 6, driving the imaging component and counterweight to move one step (equal to φ mm) towards the center and locking the new radius R' = R - φ mm. The parameters under the new radius R' are calculated: circumference C' = 2 × π × R (mm), number of sampling points Pn = C' / φ, and the frame rate of the single-point spectrometer 2 is adjusted to Fn = F × Pn / P. The ring-shaped acquisition is repeated until the radius decreases to 0, covering the entire area. The control module synchronously integrates the timestamped datasets from each ring for data synthesis.

[0047] The data synthesis steps include:

[0048] Canvas parameter calculation: The scanning imaging area is a circle with a diameter of L. The canvas pixel size is calculated based on the lens field of view φ and the scanning range. ,in This indicates rounding up to the nearest integer to ensure complete coverage of the circular area. The canvas pixel center coordinates are xc=W / 2, yc=H / 2.

[0049] Coordinate mapping transformation: For each sampling point, the transformation from polar coordinates to Cartesian coordinates and then to pixel coordinates is performed sequentially. Polar coordinates to Cartesian coordinates: x = R * cosθ, y = R * sinθ. Cartesian coordinates to image pixel coordinates: xp = xc + X / φ, yp = yc + Y / φ.

[0050] Image reconstruction via interpolation sampling: Since (xp, yp) are mostly non-integer values, image reconstruction can be achieved using methods such as interpolation sampling. For example, methods like neighborhood gray-level weighted allocation can be used to obtain the integer parts of the pixel coordinates i={xp}, j={yp}, and thus the coordinates of surrounding pixels. Figure 5 As shown: Top left A(i,j), top right B(i+1,j), bottom left D(i,j+1), bottom right C(i+1,j+1). The weighting function can be calculated using horizontal and vertical distances. The spectral data DN of the current sampling point is distributed to the above four neighboring pixels according to the weights, and finally, the weighted sums are used to obtain the spectral data for each pixel position, outputting the 3D image data. Subsequent processing can include dark noise subtraction, denoising, and other conventional spectral data preprocessing.

[0051] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0052] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A single-point, large-field-of-view, macro hyperspectral system, characterized in that, include: The imaging component includes an imaging module (5), a rotation module (6), an object plane (7), and a control module. The imaging component includes a light source (1), a single-point spectrometer (2), a mirror module (3), and a lens module (4). The control module is electrically connected to the imaging component, the rotation module (5), and the linear motion module (6) to achieve data communication; The light source (1) is arranged at the beginning of the optical path of the system to provide stable and suitable illumination for imaging; The single-point spectrometer (2) is connected to the lens module (4) to achieve spectral imaging; The reflector module (3) is positioned between the lens module (4) and the object plane (7) on which the object to be imaged is mounted; The lens module (4) is installed on the optical path of the system on the side closest to the object plane (7), and the height of the lens module (4) is adjustable; The imaging component is slidably connected to one side of the linear motion module (6), and the rotation module (5) is fixedly connected to the linear motion module (6); Based on the initial input scanning range, the control module adjusts the rotation speed of the rotating module (5) and the linear movement distance of the imaging component to collect spectra until the recorded spectral data covers the entire area of ​​the object to be imaged, and then synthesizes the collected data into three-dimensional image data. A counterweight (8) is slidably connected to the side of the linear motion module (6) away from the imaging component. The imaging component and the counterweight (8) are symmetrically slidably connected at both ends of the linear motion module (6). The rotating module (5) is located in the middle of the linear motion module (6). The center of the rotating module (5) coincides with that of the linear motion module (6). The installation angle of the reflector module (3) is fixed. The installation angle of the reflector module (3) satisfies that when the imaging component slides to the closest point to the rotating module (5), the reflector module (3) will place the light source reflection guide plane (7) processed by the lens module (4) at the center point directly below the rotating module (5) and reflect the light signal reflected at the center point to the lens module (4).

2. The single-point large field-of-view macro hyperspectral system according to claim 1, characterized in that: After the height of the lens module (4) is adjusted, the optical axis of the lens module (4) passes through the center of the mirror module (3).

3. The single-point large field-of-view macro hyperspectral system according to claim 1, characterized in that: The rotating module (5) rotates via a motor with a built-in encoder.

4. A single-point large field-of-view macro hyperspectral imaging method, implemented using the single-point large field-of-view macro hyperspectral system as described in any one of claims 1-3, characterized in that: include: S1: Place the object to be imaged on the object plane (7), input the initial scanning range into the control module, the initial scanning range is set based on the size of the object to be imaged, and calculate the initial scanning radius and the initial rotation speed of the rotation module (5) based on the initial scanning range; S2: Start the rotation module (5) to rotate. After the rotation speed stabilizes, the single-point spectrometer (2) starts to collect spectra and stores the collected data to the control module. S3: After completing one cycle of spectral acquisition, the control module controls the imaging component to move and change the scanning radius, and continues to complete one cycle of spectral acquisition, repeatedly changing the scanning radius until the scanning radius is changed to 0. S4: The control module synthesizes the collected data into three-dimensional image data.

5. The single-point large field-of-view macro hyperspectral system imaging method according to claim 4, characterized in that: The initial rotation speed is the ratio of the number of sampling points of the single-point spectrometer (2) to the number of times the single-point spectrometer (2) collects data per second. The number of sampling points is the ratio of the circumference of the initial scanning range to the field of view of the lens module (4).

6. The single-point large field-of-view macro hyperspectral system imaging method according to claim 5, characterized in that: S4: The control module synthesizes the acquired data into 3D image data, including: S4.1 Calculate the canvas pixel size and the pixel coordinates of the canvas pixel center; S4.2 Convert the polar coordinates of each sampling point to pixel coordinates; S4.

3. Use interpolation sampling to reconstruct the image and synthesize three-dimensional image data.

7. The single-point large field-of-view macro hyperspectral system imaging method according to claim 6, characterized in that: The interpolation sampling method includes: rounding the pixel coordinates of any sampling point to obtain four neighboring pixels, calculating the horizontal and vertical distances from the sampling point to any neighboring pixel, calculating a weight function based on the horizontal and vertical distances, distributing the spectral data of the sampling point to the four neighboring pixels according to the weights, and finally weighted summing to obtain the spectral data of each sampling point.

Images