Multi-channel spectral measurement optimized scanning method based on spatial light modulation

Abstract

The invention belongs to the technical field of multichannel spectral measurement, and particularly relates to a multichannel spectral measurement optimized scanning method based on spatial light modulation. According to the method, a greedy algorithm is used for solving a maximum independent set of a modulation region as a core, and the position of an initial point is selected in an auxiliary manner through a global search method based on a degree-of-freedom extreme value and a secondary optimization criterion based on a sparse set; the analysis of the actual spectrum projection geometrical shape is added as an algorithm basis, so that errors caused by crosstalk to a multi-channel spectrum measurement system are inhibited; the programmable and highly flexible control characteristics of the spatial light modulator are ingeniously combined, the modulation area is filled by combining the spatial light modulator with the geometric shape of the spectral projection, finally, an optimized scanning algorithm capable of achieving efficient detection is formed, and compared with a periodic scanning mode of a traditional multi-channel spectral measurement system, the method has the advantages that the detection efficiency is improved, and the detection cost is reduced. According to the invention, the scanning efficiency can be improved to the greatest extent while the vertical measurement resolution is ensured to be met.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-channel spectral measurement, and in particular to a multi-channel spectral measurement optimization scanning algorithm based on spatial light modulation. Background Art

[0002] Spectroscopic measurement technology is widely used in nondestructive testing, geological exploration, agricultural surveys, and many other fields. With the development of technology, related process devices have made significant progress in recent years. Spectroscopic measurement technology can be categorized into single-channel spectroscopy and multi-channel spectroscopy based on the data acquisition method. Multi-channel spectroscopy systems utilize detectors such as CCDs, CMOS, and photodiode arrays, enabling the system to simultaneously acquire spectral data at multiple wavelengths or multiple measurement locations. Compared with traditional single-point spectroscopy, multi-channel spectroscopy improves measurement efficiency, enhances the dimensionality of spectral information, and reduces scanning time. These advantages make it a key development direction in modern spectroscopy technology. Currently, the more commonly used multi-channel spectroscopy techniques include hyperspectral imaging, stacked scanning spectroscopy, and spectral confocal measurement.

[0003] Existing research has shown that crosstalk between spectra on the imaging plane can significantly impact the accuracy of multi-channel spectral measurement systems. In general spectral imaging systems, crosstalk can be reduced by controlling the scanning pitch, maximizing scanning efficiency while ensuring vertical measurement resolution. However, in current mainstream multi-channel spectral measurement systems, the spatial light modulator is typically scanned at a scanning pitch that minimizes crosstalk. In this method, the dot array of each scan is arranged in a regular grid, which prevents the spatial light modulator from fully utilizing its control capabilities. The number of points covered by a single scan is very limited, thereby reducing the utilization of the scanning area and overall scanning efficiency. Summary of the Invention

[0004] In response to the above-mentioned deficiencies in the prior art, the present invention proposes a multi-channel spectral measurement optimization scanning method based on a spatial light modulator; this method models the optimization scanning problem as a maximum independent set problem in graph theory, utilizes the programmability and highly flexible control characteristics of the spatial light modulator, and incorporates analysis of the actual spectral projection geometry as the algorithm basis. The maximum independent set of the modulation area is solved by a greedy algorithm as a scanning point matrix, thereby enabling efficient measurement of highly reflective samples; this method can overcome the efficiency bottleneck caused by the fixed scanning mode of the existing multi-channel spectral measurement system while ensuring that the vertical measurement resolution is met, thereby maximizing the scanning efficiency of the multi-channel spectral measurement system and suppressing the errors caused by crosstalk to the multi-channel spectral measurement system.

[0005] The technical solution of the present invention is specifically described as follows.

[0006] A multi-channel spectral measurement optimization scanning method based on spatial light modulation is proposed. The method determines the effective scanning range of the modulation area by analyzing the geometric shape of the actual spectral projection. The candidate scanning points are placed in the modulation area with the geometric shape as a constraint as the input set of the greedy algorithm. The greedy algorithm is then used to solve the maximum independent set in the modulation area as the scanning point array that satisfies the crosstalk constraint.

[0007] In the present invention, the spatial light modulator is a DMD, LCD or LCoS modulator. The modulation area refers to the area that the light modulator can modulate, that is, the scanning area.

[0008] In the present invention, the geometric shape of the spectrum projection is obtained by detecting and extracting the spectrum lines using an image processing method using connected region detection and Hough transform.

[0009] In the present invention, the method for solving the maximum independent set in the modulation area using the greedy algorithm is as follows:

[0010] The analysis area based on the spectral projection geometry is based on the unit pixel canvas of the spatial modulator. According to the pixel parameters of the spatial modulator, the spatial modulator is abstracted into a two-dimensional discrete point set (canvas). For example, when the spatial modulator is a Digtial Micromirror Device (DMD, charge-coupled device), each micromirror is abstracted into a point. The DMD with m×n micromirrors is abstracted into an m×n two-dimensional discrete point set.

[0011] Let G = (V, E) be an undirected graph, where the vertex set V represents all optional scanning points on the spatial light modulator, and the edge set E is determined by whether the Euclidean distance d(i, j) between two points i and j is less than a preset threshold F. If d(i, j) < F, then an edge is established in the vertex set V to connect them; let is a subset of vertices. If there is no edge between any two vertices, then S is called an independent set. The maximum independent set is the set with the largest number of vertices among all independent sets. The greedy algorithm is used to solve the maximum independent set:

[0012] (1) Initialize an empty set S;

[0013] (2) Select a vertex with the highest degree of freedom from V as the starting point. Each time, select the point closest to the current point from the remaining points and add it to S. Delete all vertices adjacent to this point to ensure crosstalk-free scanning. This process is repeated until all points are processed.

[0014] In the present invention, a global search method for the extreme value of freedom is used to find the point with the highest degree of freedom as the initial point to ensure the maximum scale of the initial solution and improve the calculation efficiency; the mathematical expression is as follows:

[0015]

[0016] Here, χ is an indicator function used to determine whether the constraints between points v and u in the vertex set V are satisfied, that is, whether the Euclidean distance d(v,u) between points v and u is less than the preset threshold F.

[0017] In the present invention, in step (2), the maximum independent set is solved through an iterative solution process: starting from the initial point, the closest point to the current point that meets the distance constraint conditions is selected until there are no points that meet the constraint conditions in the modulation area; specifically, first for a vertex with the highest degree of freedom in V, each time the point closest to the current point is selected from the remaining points and added to S, and then this point and the edges adjacent to it are removed from the graph until there are no remaining points, thereby achieving the solution of the first maximum independent set S1, and then the vertex with the highest degree of freedom is removed from the graph, and the remaining vertex with the highest degree of freedom is used as the initial point, and the maximum independent set solution is performed again; the operation is repeated until all points in V are taken, and several maximum independent sets are obtained as input dot matrices for scanning, at which time it is considered that the modulation area has been scanned.

[0018] In the present invention, when the maximum independent set sizes of multiple candidate initial points are the same, based on the sparse set secondary optimization criterion, the solution with the largest sum of inter-point distances is selected as the maximum independent set.

[0019] In the present invention, the selection of the initial point combines the global search method of the extreme value of freedom and the secondary optimization criterion of the sparse set. Specifically, first, the point with the highest degree of freedom in the vertex set V corresponding to the modulation area is solved, and then each point with the highest degree of freedom is used as the starting point to solve the maximum independent set (after the starting point is selected, in the iterative process, each time the point closest to the current point is selected from the remaining points to be added to the vertex subset S). Finally, among all the starting points, which starting points have the most elements in the maximum independent set? The distances of the maximum independent sets corresponding to these starting points are calculated, and the maximum independent set with the largest sum of distances between points is calculated. The starting point of this maximum independent set is used as the initial point.

[0020] In the present invention, it is applicable to a large-array spectral confocal measurement system based on a spatial light modulator; the measurement area of ​​the large-array spectral confocal measurement of the spatial light modulator is within 8mm×8mm.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] The present invention uses a greedy algorithm to solve the maximum independent set of the modulation area as the scanning point matrix, maximizes the number of crosstalk-free scanning points in a single scan, avoids redundant scanning in the traditional periodic scanning method, significantly reduces the number of repeated scans, and effectively improves the scanning efficiency.

[0023] Compared with the traditional periodic scanning method, the present invention introduces analysis based on spectral projection geometry and fills the modulation area by fusing the actual spectral projection geometry, which can suppress the error caused by crosstalk to the multi-channel spectral measurement system.

[0024] The proposed maximum independent set solution dynamically adjusts canvas input parameters based on the spatial light modulator's pixel parameters (e.g., DMD size and pitch) to meet varying resolution requirements. This flexibility makes the algorithm compatible with a variety of spatial light modulators and applicable to diverse multi-channel spectral measurement applications, adapting to complex measurement requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The multi-channel spectral measurement optimization scanning method based on spatial light modulation is further described below with reference to the accompanying drawings and embodiments.

[0026] Figure 1 This is a flow chart of the multi-channel spectral measurement optimization scanning method based on spatial light modulation of the present invention.

[0027] Figure 2 Schematic diagram of the actual spectrum projection geometry analysis process.

[0028] Figure 3 Visualized conflict graph and partially generated lattice for solving the maximum independent set problem by filling the modulation area with rectangles.

[0029] Figure 4 Visualized conflict graph and partially generated lattice when solving for the maximum independent set to fill the modulation region with the spectral projection shape. DETAILED DESCRIPTION

[0030] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0031] Example 1

[0032] This paper abstracts a Digtial Micromirror Device (DMD) as an m×n two-dimensional discrete point set. The edge set is determined by whether the Euclidean distance between any two points in the set is less than a threshold. If so, the two points are connected in the graph to form an edge. The initial point selection combines a global search method for the degree of freedom extreme value with a sparse set secondary optimization criterion. A greedy algorithm is then used to determine the maximum independent set of the spatial modulation region.

[0033] Figure 1The flowchart of the multi-channel spectral measurement optimization scanning method based on spatial light modulation of the present invention is shown. This embodiment uses a planar array spectral confocal system based on Digtial Micromirror Device (DMD) to specifically introduce the optimization scanning algorithm. The Digtial Micromirror Device (DMD) optimization scanning problem can be modeled as a maximum independent set problem in graph theory. Let the Digtial Micromirror Device (DMD) be an m×n two-dimensional discrete canvas, and each micromirror corresponds to a pixel point. In order to ensure that crosstalk is minimized during the scanning process, the distance between adjacent scanning points must meet certain constraints, which is converted into solving the maximum independent set of points on the canvas.

[0034] Let G = (V, E) be an undirected graph, where the vertex set V represents all optional scanning points on the Digital Micromirror Device (DMD), and the edge set E is determined by whether the Euclidean distance d(i, j) between two points is less than the preset threshold F. If d(i, j) < F, then an edge is established in the vertex set V to connect them. is a subset of vertices. If there is no edge between any two vertices, then S is called an independent set, and the maximum independent set is the set with the largest number of vertices among all independent sets.

[0035] Because the maximum independent set problem is NP-hard (non-deterministic polynomial), a greedy algorithm combined with an optimization strategy is used to solve it. The basic framework of the greedy algorithm involves initializing an empty set S and then selecting a vertex with the highest degree of freedom from the vertex set V as the starting point. During each iteration, the point closest to the current point is selected from the remaining points and added to the vertex subset S. All adjacent vertices to this point are removed to ensure crosstalk-free scanning. This process is repeated until all points have been processed.

[0036] In terms of the strategy for optimizing the initial point selection, a global search method for the extreme value of freedom is used to find the point with the highest degree of freedom in the entire graph as the initial point to ensure the maximum scale of the initial solution and improve the computational efficiency. The mathematical expression is as follows:

[0037]

[0038] Among them, χ is the indicator function, which is used to determine whether the constraints between points are satisfied.

[0039] In the iterative solution process, each time a point is selected from the remaining points with a distance v from the current point currentThe nearest point is added to S, and then this point and its adjacent edges are removed from the graph. First, the first maximum independent set S1 is calculated, and then its points are removed from the graph. The maximum independent set solution is performed again on the set of remaining points in the modulation area until all points in V are taken. Several maximum independent sets are obtained as the input point matrix of the scan. At this point, the modulation area is considered to be scanned. The mathematical expression is:

[0040]

[0041] In addition, a sparse set secondary optimization criterion is added to further optimize the selection of independent sets. When the maximum independent set sizes of multiple candidate initial points are the same, the distance d(v i ,v j ) and the largest solution, namely:

[0042]

[0043] This optimization can ensure that the scanning points are evenly distributed, reduce intensive scanning in local areas, and improve overall scanning uniformity.

[0044] In order to further optimize the arrangement of scanning points and make it match the spectral imaging characteristics, adjustments are made in combination with the analysis of the spectral projection geometry. Due to the dispersion characteristics of spectral imaging, such as the broadening of red light (long wavelength) and the focusing of blue light (short wavelength), the projection shape of the spectrum on the Charge-Coupled Device (CCD) is not a regular rectangle, but has a certain tilt and deformation. Therefore, a series of image processing methods such as connected area detection and Hough transform are used to detect spectral lines and extract the spectral projection geometry to make it more consistent with the spectral imaging characteristics. The range of the planar array spectral confocal system used in this embodiment is 3.35mm. The spectral projection shapes are superimposed within the effective range, and the results are as follows. Figure 2 shown.

[0045] Figure 3 and Figure 4 The visualization of the conflict graph and the first generated dot matrix for solving the maximum independent set problem when the modulation area is filled with a rectangle and a spectral projection shape (comet shape) are shown. Table 1 tests the optimization algorithm proposed in this invention using 8mm×8mm and 5mm×5mm scanning areas, respectively. The results show that compared with conventional periodic scanning, the multi-channel spectral measurement optimization scanning method based on spatial light modulation proposed in this invention can reduce the number of scans while eliminating lateral crosstalk and ensuring measurement accuracy. In addition, the analysis of the spectral projection geometry can significantly reduce the number of scans and improve scanning efficiency.

[0046] Table 1 Comparison of the number of scans between the traditional scanning method and the optimized scanning algorithm under different scanning area sizes

[0047]

[0048] In this invention, the spectral projection geometry depends on factors such as the uniformity of the light source, the sensitivity of the Charge-Coupled Device (CCD), and the reflective properties of the object being measured. Therefore, the analysis of the spectral projection geometry is adjusted to the spectral characteristics of different samples and different experimental environments. The formation of the scanning array depends on the degree of match between the total scanning array size and the spectral projection geometry combination. The lower the match, the more significant the algorithm's improvement in scanning efficiency. Therefore, the input parameters of the greedy algorithm are dynamically adjusted as the resolution and size parameters of the visible light modulator change, forming a corresponding relationship between the generated dot matrix and the input size.

[0049] This method is applicable to Digtial Micromirror Device (DMD) arrays of varying resolutions and can be extended to various types of spatial light modulators (e.g., LCD and LCoS). Furthermore, spectral parameters can be adjusted based on the actual multi-channel spectral measurement system under varying light source conditions and Charge-Coupled Device (CCD) sensitivity characteristics, making it suitable for a variety of multi-channel spectral measurement applications. The system is highly versatile and adaptable.

[0050] As described above, the present invention provides a multi-channel spectral measurement optimization scanning method based on spatial light modulation; the method of the present invention is applicable to a multi-channel spectral measurement system based on a spatial light modulator; by using a greedy algorithm to solve the maximum independent set of the modulation area as the core, and using a global search method based on the extreme value of the degree of freedom and a secondary optimization criterion based on a sparse set to assist in selecting the position of the initial point, on this basis, the analysis of the actual spectral projection geometric shape is added as the algorithm basis to suppress the error caused by crosstalk to the multi-channel spectral measurement system; the present invention cleverly combines the programmability and highly flexible control characteristics of the spatial light modulator, and fills the modulation area by combining it with the spectral projection geometric shape, and finally forms an optimized scanning algorithm that can achieve efficient detection. Compared with the periodic scanning method of the traditional multi-channel spectral measurement system, the invention can minimize the number of scans while ensuring that the vertical measurement resolution is met, the scanning speed is faster, and the scanning efficiency is improved.

Claims

1. A multi-channel spectral measurement optimization scanning method based on spatial light modulation, characterized in that: It determines the effective scanning range of the modulation area by analyzing the geometric shape of the actual spectrum projection, and places candidate scanning points in the modulation area with the geometric shape as a constraint as the input set of the greedy algorithm. Then, the greedy algorithm is used to solve the maximum independent set in the modulation area as the scanning point matrix that meets the crosstalk constraint.

2. The multi-channel spectral measurement optimization scanning method according to claim 1, characterized in that: The spatial light modulator is a DMD, LCD or LCoS modulator.

3. The multi-channel spectral measurement optimization scanning method according to claim 1, characterized in that: The spectral projection geometry is obtained by detecting spectral lines through image processing methods such as connected region detection and Hough transform.

4. The multi-channel spectral measurement optimization scanning method according to claim 1, characterized in that: The method of using the greedy algorithm to solve the maximum independent set in the modulation area is as follows: The analysis area based on the spectral projection geometry is based on the unit pixel canvas of the spatial modulator. According to the pixel parameters of the spatial modulator, the spatial modulator is abstracted into a two-dimensional discrete point set as the basis for iterative solution. Let G = (V, E) be an undirected graph, where the vertex set V represents all optional scanning points on the spatial light modulator, and the edge set E is determined by whether the Euclidean distance d(i, j) between two points i and j is less than a preset threshold F. If d(i, j) < F, then an edge is established in the vertex set V to connect them; let is a subset of vertices. If there is no edge between any two vertices, then S is called an independent set. The maximum independent set is the set with the largest number of vertices among all independent sets. The greedy algorithm is used to solve the maximum independent set: (1) Initialize an empty set S; (2) Select a vertex with the highest degree of freedom from V corresponding to the modulation area as the starting point. Each time, select the point closest to the current point from the remaining points in V except the vertex subset S and add it to S. All vertices adjacent to the point are deleted to ensure crosstalk-free scanning. This process is repeated until all points have been processed.

5. The multi-channel spectral measurement optimization scanning method according to claim 4, characterized in that: The global search method of degree of freedom extreme value is used to find the point with the highest degree of freedom as the initial point to ensure the maximum scale of the initial solution and improve the calculation efficiency. The mathematical expression is as follows: Here, χ is an indicator function used to determine whether the constraints between points v and u in the vertex set V are satisfied, that is, whether the Euclidean distance d(v,u) between points v and u is less than the preset threshold F.

6. The multi-channel spectral measurement optimization scanning method according to claim 4, characterized in that: When the maximum independent set sizes of multiple candidate initial points are the same, the solution with the largest sum of inter-point distances is selected as the maximum independent set based on the sparse set secondary optimization criterion.

7. The multi-channel spectral measurement optimization scanning method according to claim 1, characterized in that: It is suitable for a large-array spectral confocal measurement system based on a spatial light modulator; the measurement area of ​​the large-array spectral confocal measurement of the spatial light modulator is within 8mm×8mm.

Images