Single-point large-view micro-distance hyperspectral system and method

Abstract

The system comprises an imaging assembly, a rotating module, a linear moving module, an object plane and a control module, and the imaging assembly comprises a light source, a single-point spectrometer, a reflector module and a lens module; the control module adjusts the rotating speed of the rotating module and the linear movement distance of the imaging assembly based on the initially input scanning range to carry out spectrum collection until the recorded spectrum collection data covers the whole area of an object needing to be imaged, then the collected data is synthesized into three-dimensional image data, and one-time large-view imaging is achieved through rotation. A complicated splicing step is not needed, and a single-point single-light-source design ensures that imaging light of each pixel is uniform, and stable imaging quality is ensured; according to the method, image splicing is not needed after moving and rotating imaging, dependence on feature point matching is not needed, the problem that splicing is difficult due to the fact that feature points are few in the prior art is effectively solved, and efficient and high-quality imaging of a large-view-field area can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of hyperspectral technology, and in particular to a single-point, large-field-of-view, macro-distance hyperspectral system and method. Background Art

[0002] Hyperspectral imaging technology, 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. Failure to obtain high-quality, wide-field-of-view spectral data in macro and microscopic scenarios can directly lead to biased test results and inaccurate analytical conclusions, severely impacting research and production decisions. Therefore, achieving efficient and accurate data collection is crucial.

[0003] To address this demand, existing hyperspectral imaging and data processing equipment is striving to enhance data acquisition and processing capabilities in macro and microscopic scenarios through hardware upgrades and algorithm optimization, aiming to provide reliable data support for various fields. However, traditional hyperspectral technology still faces significant drawbacks in practical applications: First, the limited macro field of view coexists with data stitching challenges. Traditional hyperspectral macro lenses have an extremely small single-shot range, and obtaining complete large-field data requires multiple captures and complex stitching algorithms. This not only significantly increases data processing costs, but also lacks effective matching points in areas with uniform surface texture and similar spectral characteristics, resulting in discontinuities and misalignments in the stitched data. Second, image stitching quality severely impacts data usability. Hyperspectral images generated by line scanning technology exhibit noticeable seams and significant differences in brightness between scan lines, resulting in poor image consistency in the stitched image. When scanning in separate regions, the blurred features at the edges make it difficult for the equipment to accurately match the edges, resulting in texture discontinuities and boundary misalignments in the stitched image. Third, there is a significant conflict between scanning efficiency and positioning accuracy. Taking a 20x objective lens for example, scanning a 1.5cm x 1.5cm area takes 48 minutes. Furthermore, it is difficult to precisely locate the sample area, and scan position deviation often leads to missed or duplicate scans of key data, severely limiting the timeliness of applications. These technical bottlenecks have become the core issues hindering the widespread application of hyperspectral technology in macro and microscopic scenarios. Summary of the Invention

[0004] The present invention provides a single-point, large-field-of-view macro hyperspectral system and method, which solves the problem of difficulty in quickly acquiring high-quality, large-field-of-view hyperspectral data due to the narrow field of view of traditional hyperspectral macro lenses, obvious seams in image splicing, difficulty in matching feature points, and low scanning efficiency.

[0005] The present invention provides the following technical solutions: 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, wherein the imaging component comprises: a light source, a single-point spectrometer, a reflector 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 intercommunication; the light source is arranged at the starting end of the system optical path to provide stable and adaptive lighting for imaging; the single-point spectrometer is connected to the lens module to achieve spectral imaging; the reflector module is arranged between the lens module and the optical path of the object plane carrying the required imaging object; the lens module is installed on the optical path of the system close 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; the control module adjusts the rotation speed of the rotation module and the linear motion distance of the imaging component based on the scanning range initially input to perform spectral acquisition, until the recorded spectral acquisition data covers the entire area of ​​the required imaging object, and then synthesizes the acquired data into three-dimensional image data.

[0006] The light source can flexibly adjust its luminous properties, such as spectral range and intensity, based on application requirements such as macro and microscopic scenes. This provides a stable and adaptable illumination foundation for imaging and ensures optimal lighting conditions for imaging the object plane. The control module, serving as the system's "control center," issues commands to coordinate the spectral acquisition of the single-point spectrometer, the operation of the rotation module, and the linear motion module, ensuring synchronized operation of all components, enabling rapid imaging with a wide field of view and stable spectral detection. The single-point spectrometer is connected to the lens module via an optical fiber, receiving light signals from the object plane and performing spectroscopic analysis to obtain spectral information for each pixel. It is connected to the control module via a cable, receiving collected control information and transmitting the collected data to the control module for subsequent processing. The reflector module is positioned between the lens and the object plane and can be adjusted to a fixed angle. When the imaging assembly moves on the linear motion module's track to its closest position to the rotation module, the reflector module's optical mirror precisely reflects the light processed by the lens to the center point of the object plane, directly below the rotation module. The reflected light signal from this center point is then reflected back to the lens and transmitted via the optical fiber to the single-point spectrometer. Adjusting the optical path by fixing the reflector angle ensures a more comprehensive scanning range without missing any points, and also stabilizes the center of gravity of the moving components on the track, without affecting data acquisition. The lens is mounted on the optical path near the object plane, and its height can be adjusted so that the optical axis passes through the center of the reflector, ensuring the integrity and stability of the reflected light path.

[0007] The rotation module is preferably mounted at the junction of the linear motion module's track and the main system structure. It receives commands from the control module and, driven by a built-in motor with an encoder, rotates the entire device to collect data. The encoder readings corresponding to each collected position are then transmitted to the control module. The rotation module maintains a stable position relative to the track and the main system structure, ensuring accuracy during movement and imaging.

[0008] Preferably, the linear moving module is slidably connected to a counterweight on the side away from the imaging component, the counterweight is electrically connected to the control module, and the rotating module is located between the imaging component and the counterweight. The counterweight balances the weight of the imaging side structure by its own weight, ensuring that the rotating module and the moving track are stable and have no tilt when the components are moving and stationary, maintaining the accuracy of the optical path and the imaging quality. The linear moving module and the rotating module serve as the core components of the scanning, and the center precisely coincides with the center of the rotating module. They rotate with the rotating module and receive instructions from the control module to control the imaging component and the counterweight to move linearly on the moving track. The rotation radius is calculated based on the initial position and the current position, and the position information is transmitted to the control module. Through precise transmission, the rotating module is used to achieve rapid displacement control of large-field imaging, ensuring the coverage of the imaging area and the 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, and the initial scanning radius and the initial rotation speed of the rotation module are calculated based on the initial scanning range;

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

[0012] S3: After completing one week of spectrum acquisition, the control module controls the imaging component to move and change the scanning radius, and continues to complete one week of spectrum 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 synthesizing the collected data into three-dimensional image data includes:

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

[0017] S4.2, converting the polar coordinates of each sampling point into pixel coordinates;

[0018] S4.3. Use interpolation sampling method 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 distance and the vertical distance, distributing the spectral data of the sampling point to the four neighboring pixels according to the weight, and finally performing weighted accumulation 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, without the need for complex stitching steps, thus avoiding the various problems caused by stitching in traditional technologies. The single-point, single-light source design ensures uniform imaging light for each pixel, ensuring stable imaging quality. On the other hand, mobile + rotation imaging does not require image stitching and does not rely on feature point matching, effectively solving the stitching difficulty caused by the small number of feature points in traditional technologies, and can achieve efficient, 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 is relatively low in cost and can even be implemented through sensors and spectroscopic modules, greatly reducing hardware investment. At the same time, the linear array detector of the single-point spectrometer is less difficult to optimize than the line-scan area array detector, further saving costs from the technical optimization level, which can help companies reduce investment while ensuring imaging effects.

[0023] 3. This single-point, large-field-of-view, macro-hyperspectral system can significantly improve imaging speed. Its rotating structure enables rapid operation. Combined with the rapid characteristics of single-point imaging itself, it greatly shortens imaging time. It is significantly faster than the time taken by traditional high-magnification objective line scanning, meeting the demand for rapid acquisition of image data in practical applications and improving overall work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0025] Figure 2 This is a schematic diagram of the installation angle structure of the lens module in Example 1 of the present application.

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

[0027] Figure 4This is a schematic diagram of the steps of the imaging method of Example 2 of the present application.

[0028] Figure 5 This is a schematic diagram of points taken during the interpolation sampling and image reconstruction process in the second embodiment of the present application.

[0029] In the figure: 1. Light source; 2. Single-point spectrometer; 3. Reflector module; 4. Lens module; 5. Rotation module; 6. Linear motion module; 7. Object plane; 8. Counterweight. DETAILED DESCRIPTION

[0030] 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.

[0031] Example 1

[0032] See also Figure 1 A 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 the counterweight 8 are symmetrically slidably connected and arranged at both ends of the linear moving module 6, and the rotating module 5 is located in the middle of the linear moving module 6. The control module is electrically connected to the imaging component, the rotating module 5 and the linear moving module 6 to realize data intercommunication; the light source 1 is arranged at the starting end of the system optical path to provide stable and adaptive lighting for imaging; the single-point spectrometer 2 is connected to the lens module 4 to realize spectral imaging; the reflector module 3 is arranged between the lens module 4 and the optical path of the object plane 7 carrying the required imaging object. 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 to the rotating module 5, the reflector module 3 will reflect the light source processed by the lens module 4 to the center point of the object plane 7 located directly below the rotating module 5, and reflect the light signal reflected from the center point to the lens module 4. By adjusting the direction of the optical path by fixing the reflector angle of the reflector module 3, the system scanning range can be guaranteed to be more comprehensive without missing points, and the center of gravity of the imaging component and the counterweight 8 on the moving track of the linear moving module 6 is more stable as a whole, which will not affect data acquisition.

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

[0035] Due to the installation of the equipment and scanning mirror, in order to complete the imaging, the optical reflector of the reflector module 3 needs to have a certain angle, and the angle is calculated as follows. The relevant parameter definitions are: w is the minimum distance between the reflector and the rotation center, h is the height from the imaging plane, the dotted line in the figure is the normal line, the reflector is perpendicular to the normal line, angle b is the angle between the reflector and the water surface, and angle a is the angle between the incident light and the reflected light. The degree calculation formula of angle a is a=arctan (h / w), and the incident angle is equal to the reflection angle. Figure 3 As shown, the angle b = 90° + 0.5*a between the optical reflector and the water surface can be calculated. The optical reflector of reflector module 3 is fixed at this angle. A Y-type optical fiber connects to light source 1 and single-point spectrometer 2. The light source signal is received and transmitted to reflector module 3. The light reflected from object plane 7 collected by reflector module 3 is initially converged and transmitted to single-point spectrometer 2 via the Y-type optical fiber. Models and other parameters can be selected based on the usage scenario.

[0036] Example 2

[0037] This embodiment is an imaging method of a single-point, large-field-of-view, macro-distance hyperspectral system using the first embodiment. Figure 4 .

[0038] First, perform hardware initialization and 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 movement module 6 in a standby state; initialize the encoder connection in the rotation module 5 to ensure accurate transmission of the angle signal; initialize the light source 1 control to enable the light source 1 to provide stable light output for the system.

[0039] Then, set the control parameters. After hardware initialization, place the desired imaging object and reference plate on object plane 7 and enter control parameters based on the acquisition area size. Set the scanning range to define the area for subsequent scanning operations; enter the field of view (i.e., physical pixel size) of lens module 4 as the basis for calculating scanning parameters; and set the exposure time of single-point spectrometer 2 to obtain clear spectral data. After receiving the instructions, the control module starts emitting light to preheat, loads the acquisition configuration to single-point spectrometer 2, initializes the rotation module 5 and linear motion module 6, fixes the angle of the reflector module 3, and enters the acquisition standby state. The control module generates control instructions based on the acquisition range and lens field of view.

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

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

[0042] The motor with an encoder built into the rotation module 5 starts, driving the imaging assembly in circular motion at the current radius R. Counterweight 8 balances the structural weight, ensuring a stable optical path during rotation. The encoder monitors the real-time rotational speed: the angle change θ over a time period T is recorded and converted to the real-time rotational speed Vc = θ / T. If the optimal speed V is not reached, the system waits until it reaches the optimal speed 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, where it is reflected and illuminates the surface of the sample / reference plate. The reflected light returns along its original path, converges through lens module 4, and is transmitted to single-point spectrometer 2 via a Y-shaped optical fiber. The encoder evenly divides the angle according to the number of sampling points P and records the real-time angle value. The linear motion module 6 simultaneously records the current radius and the acquisition timestamp of each sampling point. The spectral data is associated with the timestamp, angle, and radius parameters and then transmitted back to the control module. Data synthesis, storage, and display are performed according to the scanning radius and rotation angle until all spectral data are recorded. The data format is shown in Table 1:

[0044] Table 1:

[0045] .

[0046] After each circular acquisition, the control module sends a displacement command to the linear motion module 6, driving the imaging assembly and counterweight to move one step (equal to φ mm) toward the center and locking in the new radius R' = R - φ mm. The parameters for the new radius R' are calculated: circumference C' = 2 × π × R (mm), number of sampling points Pn = C' / φ, and the acquisition frame rate of the single-point spectrometer 2 is adjusted to Fn = F × Pn / P. The circular acquisition is repeated until the radius is reduced to zero, covering the entire area. The control module then simultaneously integrates the time-stamped datasets from each circle 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 size φ and the scanning range. ,in Indicates rounding up to ensure complete coverage of the circular area. The coordinates of the center of the canvas pixel are xc=W / 2, yc=H / 2.

[0049] Coordinate mapping conversion: For each sampling point, complete the conversion from polar coordinates to Cartesian coordinates and then to pixel coordinates. Polar coordinates to Cartesian coordinates: x = R * cos θ, y = R * sin θ. Cartesian coordinates to image pixel coordinates: xp = xc + X / φ, yp = yc + Y / φ.

[0050] Interpolation sampling to reconstruct the image: Since (xp, yp) are mostly non-integer values, interpolation sampling and other methods can be used to reconstruct the image. For example, the neighborhood grayscale weighted allocation method takes the integer part of the pixel coordinates i={xp}, j={yp} to obtain the coordinates of the surrounding pixels, such as Figure 5 As shown: A(i, j) on the upper left, B(i+1, j) on the upper right, D(i, j+1) on the lower left, and C(i+1, j+1) on the lower right. A weighting function can be calculated using horizontal and vertical distances. The spectral data DN of the current sampling point is distributed to the four neighboring pixels according to their weights. Finally, the weighted accumulation generates the spectral data for each pixel, outputting the 3D image data. This can then be used for conventional spectral data preprocessing, such as dark noise subtraction and noise reduction.

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

[0052] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A single-point, large-field-of-view, macro-distance hyperspectral system, characterized in that: include: An imaging component, a rotation module (5), a linear motion module (6), an object plane (7), and a control module, wherein the imaging component includes: a light source (1), a single-point spectrometer (2), a reflector module (3), and a lens module (4); The control module is electrically connected to the imaging component, the rotation module (5) and the linear movement module (6) to achieve data intercommunication; The light source (1) is arranged at the starting end of the system light path to provide stable and adaptive lighting for imaging; The single-point spectrometer (2) is connected to the lens module (4) to achieve spectral imaging; The reflector module (3) is arranged between the lens module (4) and the optical path of the object plane (7) carrying the desired imaging object; The lens module (4) is installed on the optical path of the system on a side close to the object plane (7), and the height of the lens module (4) is adjustable; The imaging assembly 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); The control module adjusts the rotation speed of the rotating module (5) and the linear movement distance of the imaging component based on the scanning range of the initial input to perform spectrum acquisition until the recorded spectrum acquisition data covers the entire area of ​​the required imaging object and then synthesizes the acquired data into three-dimensional image data.

2. The single-point, large-field-of-view, macro-distance hyperspectral system according to claim 1, characterized in that: A counterweight (8) is slidably connected to the side of the linear motion module (6) away from the imaging component. The counterweight (8) is electrically connected to the control module. The rotation module (5) is located between the imaging component and the counterweight (8).

3. The single-point, large-field-of-view, macro-distance hyperspectral system according to claim 2, characterized in that: The centers of the rotating module (5) and the linear moving module (6) coincide with each other.

4. The single-point, large-field-of-view, macro-distance hyperspectral system according to claim 3, characterized in that: The installation angle of the reflector module (3) is fixed. When the imaging assembly slides to the position closest to the rotating module (5), the reflector module (3) reflects the light source processed by the lens module (4) to the center point of the guide plane (7) located directly below the rotating module (5), and reflects the light signal reflected from the center point to the lens module (4).

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

6. The single-point, large-field-of-view, macro-distance hyperspectral system according to claim 1, characterized in that: The rotation module (5) is rotated by a motor with a built-in encoder.

7. A single-point, large-field-of-view, macro-range hyperspectral imaging method, implemented using the single-point, large-field-of-view, macro-range hyperspectral imaging system according to any one of claims 1 to 6, characterized in that: include: S1: placing the object to be imaged on the object plane (7), inputting the initial scanning range into the control module, the initial scanning range being set based on the size of the object to be imaged, and calculating 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 in the control module; S3: After completing one week of spectrum acquisition, the control module controls the imaging component to move and change the scanning radius, and continues to complete one week of spectrum 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.

8. The single-point, large-field-of-view, macro-distance hyperspectral system imaging method according to claim 7, 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 acquisitions per second of the single-point spectrometer (2), and the number of sampling points is the ratio of the circumference length of the initial scanning range to the field of view of the lens module (4).

9. The single-point, large-field-of-view, macro-distance hyperspectral system imaging method according to claim 8, characterized in that: S4: The control module synthesizes the collected data into three-dimensional image data including: S4.

1. Calculate the pixel size of the canvas and the pixel coordinates of the center of the canvas pixel; S4.2, converting the polar coordinates of each sampling point into pixel coordinates; S4.

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

10. The single-point, large-field-of-view, macro-distance hyperspectral system imaging method according to claim 9, 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 the 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 accumulation to obtain the spectral data of each sampling point.

Images