Method for generating a reference spectrum combination, information storage and reading system

Abstract

The present disclosure relates to a method for generating a reference spectrum combination, an information storage and reading system, the method comprising: obtaining a plurality of candidate spectra; combining candidate spectra in the plurality of candidate spectra to obtain at least one candidate spectrum combination; synthesizing the plurality of candidate spectra in each candidate spectrum combination based on a plurality of synthesis modes to obtain a plurality of synthesized spectra corresponding to each candidate spectrum combination; determining a plurality of target synthesized spectra corresponding to each candidate spectrum combination according to the plurality of synthesized spectra corresponding to each candidate spectrum combination; and determining a reference spectrum combination in the at least one candidate spectrum combination according to a spectral difference degree between the plurality of target synthesized spectra corresponding to each candidate spectrum combination. The method for generating a reference spectrum combination provided by the present disclosure can improve the recognition accuracy of spectral information.

Description

Technical Field

[0001] This disclosure relates to the field of information processing, and in particular to a method for generating a reference spectral combination, and an information storage and retrieval system. Background Technology

[0002] In existing technologies, reference spectral information can be set, and the information of the spectrum to be identified can be compared with the reference spectral information to identify the spectrum. However, when the similarity between the reference spectral information is high, there is a possibility of misidentification, resulting in a low recognition accuracy. Summary of the Invention

[0003] In view of this, the present disclosure provides a method for generating a reference spectral combination to solve the technical problem of low recognition accuracy in the prior art.

[0004] According to one aspect of this disclosure, a method for generating a reference spectral combination is provided. The method includes: acquiring a plurality of candidate spectra; combining the candidate spectra from the plurality of candidate spectra to obtain at least one candidate spectral combination, wherein each candidate spectral combination contains the same number of candidate spectra; synthesizing the plurality of candidate spectra in each candidate spectral combination based on multiple synthesis methods to obtain a plurality of synthesized spectra corresponding to each candidate spectral combination; determining a plurality of target synthesized spectra corresponding to each candidate spectral combination based on the plurality of synthesized spectra corresponding to each candidate spectral combination; determining a reference spectral combination in the at least one candidate spectral combination based on the spectral difference between the plurality of target synthesized spectra corresponding to each candidate spectral combination; wherein the plurality of synthesized spectra corresponding to the reference spectral combination are used to compare with a spectrum to be identified to obtain an identification result corresponding to the spectrum to be identified.

[0005] In one possible implementation, combining the candidate spectra from the plurality of candidate spectra to obtain at least one candidate spectrum combination includes: selecting a first preset number of candidate spectra from the plurality of candidate spectra to form at least one initial candidate spectrum combination; determining a spectral difference value corresponding to each initial candidate spectrum combination based on the spectral difference degree between the plurality of candidate spectra in each initial candidate spectrum combination; and using the initial candidate spectrum combination with the spectral difference value greater than a preset spectral difference value as the candidate spectrum combination, or using the initial candidate spectrum combination with the largest spectral difference value and a second preset number as the candidate spectrum combination.

[0006] In one possible implementation, combining the candidate spectra from the plurality of candidate spectra to obtain at least one candidate spectrum combination includes: selecting a first preset number of candidate spectra from the plurality of candidate spectra to form at least one candidate spectrum combination.

[0007] In one possible implementation, the step of synthesizing multiple candidate spectra in each candidate spectral combination based on multiple synthesis methods to obtain multiple synthesized spectra corresponding to each candidate spectral combination includes: determining at least one set of values ​​corresponding to each candidate spectral combination according to at least one preset value, with each set of values ​​corresponding to a synthesis method; and linearly combining the multiple candidate spectra in each candidate spectral combination with each set of values ​​to obtain multiple synthesized spectra corresponding to each candidate spectral combination.

[0008] In one possible implementation, each candidate spectral combination corresponds to multiple numerical groups. Determining the reference spectral combination among the at least one candidate spectral combination based on the spectral difference between the multiple target synthetic spectra corresponding to each candidate spectral combination includes: for any candidate spectral combination corresponding to multiple numerical groups, determining a similar numerical group corresponding to each numerical group; determining the spectral difference value corresponding to the candidate spectral combination based on the spectral difference between the target synthetic spectrum corresponding to each numerical group and the target synthetic spectrum corresponding to each numerical group in the similar numerical groups; and selecting the candidate spectral combination with the largest corresponding spectral difference value among the at least one candidate spectral combinations as the reference spectral combination.

[0009] In one possible implementation, the absolute value of the difference between the sum of the values ​​of each value group and the sum of the values ​​of any value group in the similar value group corresponding to the value group is less than a preset difference, and / or the Hamming distance between each value group and any value group in the similar value group corresponding to the value group is less than a preset distance.

[0010] In one possible implementation, the dimension of the target synthetic spectrum is lower than that of the synthetic spectrum.

[0011] In one possible implementation, determining the multiple target synthetic spectra corresponding to each candidate spectral combination based on the multiple synthetic spectra corresponding to each candidate spectral combination includes: determining the multiple target synthetic spectra corresponding to each candidate spectral combination based on the multiple synthetic spectra corresponding to each candidate spectral combination and the preset weights corresponding to each band of the synthetic spectra; wherein, the target synthetic spectra include bands in the multiple synthetic spectra of the corresponding candidate spectral combination whose preset weights are greater than preset thresholds, and the preset weights are used to represent the influence of each band of the synthetic spectra on the calculation of the spectral difference between the multiple synthetic spectra.

[0012] According to another aspect of this disclosure, an information storage and retrieval system is also provided, the information storage and retrieval system comprising: a receiving spectrometer for storing multiple synthetic spectra corresponding to the reference spectral combination generated by the reference spectral combination generation method described above.

[0013] In one possible implementation, the information storage and retrieval system further includes: a transmitting spectrometer; the transmitting spectrometer is used to acquire the spectrum to be identified; the receiving spectrometer is used to compare the spectrum to be identified with multiple synthetic spectra corresponding to the combination of the reference spectrum to obtain the identification result corresponding to the spectrum to be identified.

[0014] The method for generating a reference spectral combination provided in this disclosure can determine at least one reference spectral combination among multiple target synthetic spectra corresponding to each candidate spectral combination, based on the spectral difference between them. This ensures that the determined reference spectral combination takes into account the difference between spectra, and thus allows the selection of candidate spectral combinations with a large spectral difference between the multiple target synthetic spectra as the reference spectral combination. This can prevent misidentification and improve the recognition accuracy.

[0015] Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0016] The accompanying drawings, which are included in and form part of this specification, illustrate exemplary embodiments, features, and aspects of this disclosure together with the specification and serve to explain the principles of this disclosure.

[0017] Figure 1 This is a schematic diagram illustrating the relationship between signal-to-noise ratio, accuracy, and spectral difference values ​​provided in this disclosure.

[0018] Figure 2 This is a schematic diagram illustrating the relationship between signal-to-noise ratio and spectral difference value provided in this disclosure.

[0019] Figure 3 This is a schematic flowchart illustrating a method for generating a reference spectral combination according to an embodiment of the present disclosure.

[0020] Figure 4 This is a schematic flowchart illustrating a method for generating a reference spectral combination according to an embodiment of the present disclosure.

[0021] Figure 5 This is a schematic flowchart illustrating a method for generating a reference spectral combination according to an embodiment of the present disclosure.

[0022] Figure 6 This is a schematic flowchart illustrating a method for generating a reference spectral combination according to an embodiment of the present disclosure.

[0023] Figure 7 This is a schematic flowchart illustrating a method for generating a reference spectral combination according to an embodiment of the present disclosure.

[0024] Figure 8This is a schematic flowchart illustrating a method for generating a reference spectral combination according to an embodiment of the present disclosure.

[0025] Figure 9 This is a schematic diagram illustrating the relationship between a first preset quantity, a number of combinations, a number of spectral data, a number of numerical groups, and the time cost of the traversal method, as provided in an embodiment of this disclosure.

[0026] Figure 10 This is a schematic diagram illustrating the relationship between the spectral difference values ​​corresponding to an initial candidate spectral combination and the spectral difference values ​​corresponding to candidate spectral combinations provided in this disclosure.

[0027] Figure 11 This is a schematic diagram illustrating the relationship between the spectral difference values ​​of a combination of reference spectra under various selection principles provided in this disclosure.

[0028] Figure 12 This is a schematic diagram illustrating the relationship between the influence of each band of the synthesized spectrum on the spectral difference judgment and the spectral wavelength when the first preset quantity is different.

[0029] Figure 13 This is a schematic diagram illustrating the relationship between Hamming distance and relative frequency between numerical groups provided in this disclosure.

[0030] Figure 14 This disclosure provides a schematic diagram illustrating the relationship between the difference between the sums of numerical values ​​in a set and their relative frequencies.

[0031] Figure 15 This is a schematic diagram of the structure of an information storage and retrieval system provided in an embodiment of this disclosure.

[0032] Figure 16 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0033] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0034] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0035] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0036] When using quantum dot spectroscopy for information storage, to optimize the selected quantum dot spectral channel combination (i.e., the baseline spectral combination mentioned below) and improve information storage accuracy, it is necessary to minimize the overall spectral similarity of the spectral library encoded by these spectral channels (i.e., high spectral differences between synthesized spectra in the baseline spectral combination). This avoids misidentification of spectra and improves accuracy. When the emission spectrum of the selected quantum dot is used as the candidate spectrum, because the emission spectrum of quantum dots is narrow-spectrum, it is required that the interference between the spectral channel combinations be minimized. This can be achieved by setting the peak heights of the spectral channels to be equal, the peak positions to be uniformly distributed within a given range (400-700 nm), and a narrower half-width, in order to achieve better information storage performance.

[0037] See Figure 1 as well as Figure 2 As shown, Figure 1 Curve 1 represents the relationship between the signal-to-noise ratio (SNR) and recognition accuracy of the generated composite spectrum when each candidate spectral combination includes 5 candidate spectra and the order is equal to 2. Curve 2 represents the relationship between the SNR and recognition accuracy of the generated composite spectrum when each candidate spectral combination includes 8 candidate spectra and the order is equal to 3. Figure 1 Curve 3 is used to represent the relationship between the signal-to-noise ratio of the generated synthesized spectrum and the spectral difference value of the candidate spectrum combination (i.e., the spectral difference between multiple synthesized spectra in the candidate spectrum combination) when each candidate spectrum combination includes 5 candidate spectra and the order number is equal to 2. Figure 1 Curve 4 is used to represent the relationship between the signal-to-noise ratio of the generated composite spectrum and the spectral difference value of the candidate spectrum combinations when each candidate spectral combination includes 8 candidate spectra and the order number is equal to 3. Here, the order number represents L (as used below). i The value of is used to determine the various composition methods mentioned below, i.e., if the order is equal to 2, then L i The value of can be either 0 or 1, and the order is 3, then L i The value of can be 3 (0, 1, 2), and when the order is 2, L i When the value of is 0 or 1, and each candidate spectral combination includes 3 candidate spectra, all synthesis methods can be found in numerical groups 0 to 7 below.

[0038] according to Figure 1 as well as Figure 2 It is known that the greater the spectral difference value among the candidate spectral combinations, the stronger the noise resistance of the generated composite spectrum and the higher the recognition accuracy. Therefore, to improve the recognition accuracy, it is necessary to select candidate spectral combinations with large spectral difference values.

[0039] See Figure 3 As shown, Figure 3 This is a schematic flowchart of a method for generating a reference spectral combination provided in this disclosure. Figure 3 The present disclosure provides a method for generating a reference spectral combination, including:

[0040] Step S100: Obtain multiple candidate spectra.

[0041] For example, the multiple candidate spectra can be the spectra of different quantum dots. For instance, in Example 1, the multiple candidate spectra include candidate spectrum A, candidate spectrum B, candidate spectrum C, candidate spectrum D, and candidate spectrum E. Taking Example 1 as an example, candidate spectrum A can be the spectrum of quantum dot A, candidate spectrum B can be the spectrum of quantum dot B, candidate spectrum C can be the spectrum of quantum dot C, candidate spectrum D can be the spectrum of quantum dot D, and candidate spectrum E can be the spectrum of quantum dot E.

[0042] For example, this disclosure does not limit the number of alternative spectra.

[0043] Step S200 involves combining the candidate spectra from the plurality of candidate spectra to obtain at least one candidate spectrum combination.

[0044] For example, the number of alternative spectra in each alternative spectral combination is the same.

[0045] There are several ways to form alternative spectral combinations, for example, see Figure 4 and Figure 5 In this process, each of the multiple candidate spectral combinations includes the same number of candidate spectra. For example, if the number is 3, each candidate spectral combination includes 3 candidate spectra. This number is also referred to as the first preset number below, and its specific value can be determined according to the actual situation.

[0046] In one possible implementation, see [reference] Figure 4 , Figure 4 The diagram shows that step S200 further includes:

[0047] Step S201: Select a first preset number of candidate spectra from multiple candidate spectra to form at least one candidate spectrum combination.

[0048] For example, at least one combination of candidate spectra can be formed by traversing and selecting 3 candidate spectra from candidate spectra A to candidate spectra E (i.e., the first preset number is 3). "Traversing and selecting 3" can be understood as selecting all possible combinations of the 3 candidate spectra. The alternative spectral combinations can include: Alternative spectral combination A (including alternative spectra A, B, and C), Alternative spectral combination B (including alternative spectra A, B, and D), Alternative spectral combination C (including alternative spectra A, B, and E), Alternative spectral combination D (including alternative spectra A, C, and D), Alternative spectral combination E (including alternative spectra A, C, and E), Alternative spectral combination F (including alternative spectra A, D, and E), Alternative spectral combination G (including alternative spectra B, C, and D), Alternative spectral combination H (including alternative spectra B, C, and E), Alternative spectral combination I (including alternative spectra B, D, and E), and Alternative spectral combination J (including alternative spectra C, D, and E). This is equivalent to constituting 10 alternative spectral combinations.

[0049] For example, see Figure 9 As shown, Figure 9 If the first preset quantity is 3, and the number of candidate spectra is 20, then selecting the first preset quantity of candidate spectra from the multiple candidate spectra can form 1140 (i.e., C). 3 20 (Number) alternative spectral combinations.

[0050] In one possible implementation, see [reference] Figure 5 , Figure 5 The diagram shows that step S200 further includes:

[0051] Step S202: Select a first preset number of candidate spectra from multiple candidate spectra to form at least one initial candidate spectrum combination.

[0052] For example, taking Example 2 above as an example, three (i.e., a first preset number) candidate spectra are selected from candidate spectra A to candidate spectra E to form at least one initial candidate spectrum combination, which may include: initial candidate spectrum combination A to initial candidate spectrum combination J, the candidate spectra contained therein corresponding to the above candidate spectrum combination A to candidate spectrum combination J.

[0053] Step S203: Determine the spectral difference value corresponding to each initial candidate spectral combination based on the spectral difference degree between multiple candidate spectra in each initial candidate spectral combination.

[0054] For example, taking Example 2 as an example, the spectral difference between every two candidate spectra in each initial candidate spectral combination can be calculated using a preset algorithm to determine the most similar candidate spectrum (i.e., the candidate spectrum with the smallest spectral difference from the candidate spectrum) corresponding to each candidate spectrum. The spectral difference between each candidate spectrum in each initial candidate spectral combination and its most similar candidate spectrum is added together, and the average value (or root mean square error, etc.) of the spectral difference between multiple candidate spectra is determined based on the addition result. The obtained average value (or root mean square error, etc.) is used as the spectral difference value corresponding to the initial candidate spectral combination. For example, taking the calculation of the spectral difference value of the initial candidate spectral combination A as an example, the initial candidate spectral combination A includes candidate spectrum A, candidate spectrum B, and candidate spectrum C. The spectral difference S1 between candidate spectrum A and candidate spectrum B, the spectral difference S2 between candidate spectrum A and candidate spectrum C, and the spectral difference S3 between candidate spectrum B and candidate spectrum C can be determined using the Euclidean distance algorithm. When S1>S2>S3, the candidate spectrum most similar to candidate spectrum A is candidate spectrum C, the candidate spectrum most similar to candidate spectrum B is candidate spectrum C, and the candidate spectrum most similar to candidate spectrum C is candidate spectrum B. Therefore, the spectral difference value of the initial candidate spectrum combination is equal to (S2+S3+S3) / 3. Similarly, the spectral difference value corresponding to the initial candidate spectrum combination B to the initial candidate spectrum combination J can be obtained. The process of determining the spectral difference degree between candidate spectra using the Euclidean distance algorithm can be found in relevant technologies and will not be elaborated here. The spectral difference degree is used to represent the degree of difference between two spectra.

[0055] For example, the aforementioned preset algorithm can be one of the Euclidean distance algorithm, cosine angle algorithm, or correlation coefficient algorithm, and this disclosure does not limit it. The spectral difference value corresponding to the initial candidate spectral combination can be the average value, root mean square error, etc., of the spectral difference between multiple candidate spectra, and this disclosure does not limit it.

[0056] Step S204: Use the initial candidate spectral combination with a spectral difference value greater than the preset spectral difference value as the candidate spectral combination, or use the initial candidate spectral combination with the largest spectral difference value and the second preset number as the candidate spectral combination.

[0057] The spectral difference value of the initial candidate spectral combination is obtained based on the degree of spectral difference among multiple candidate spectra in the initial candidate spectral combination. It represents the degree of difference among the multiple candidate spectra in the initial candidate spectral combination. The larger the spectral difference value, the greater the degree of difference among the candidate spectra in the initial candidate spectral combination, that is, the lower the degree of similarity, and the more suitable it is as a reference spectral combination. (See also...) Figure 10 As shown, Figure 10shows the relationship between the spectral difference value corresponding to each initial candidate spectral combination and the spectral difference value corresponding to the candidate spectral combination, that is, the relationship between the spectral difference degree among multiple candidate spectra in each initial candidate spectral combination and the spectral difference degree among multiple target synthesized spectra corresponding to the corresponding candidate spectral combination. From Figure 10 , it can be seen that the overall trend of the spectral difference value corresponding to the initial candidate spectral combination and the spectral difference value corresponding to the candidate spectral combination is roughly the same, that is, the larger the spectral difference value corresponding to the initial candidate spectral combination, the greater the probability that the corresponding spectral difference value corresponding to the candidate spectral combination is larger. Therefore, those skilled in the art can select the initial candidate spectral combinations corresponding to large spectral difference values as the basis for determining the reference spectral combination, and screen out the initial candidate spectral combinations corresponding to small spectral difference values, thereby reducing the amount of calculation required for determining the reference spectral combination.

[0058] For example, also taking Example 2 as an example, under the condition that the initial candidate spectral combinations with spectral difference values greater than the preset spectral difference value are taken as candidate spectral combinations, the initial candidate spectral combination A to the initial candidate spectral combination J respectively correspond to spectral difference values m1 to m10. If the magnitude relationship of the spectral difference values of the above multiple initial candidate spectral combinations is m1<m7<m10<m2<m4<m6<m8<m9<m3<m5, and the value of the preset spectral difference value is between m10 and m2, then the initial candidate spectral combination B corresponding to m2, the initial candidate spectral combination D corresponding to m4, the initial candidate spectral combination F corresponding to m6, the initial candidate spectral combination H corresponding to m8, the initial candidate spectral combination I corresponding to m9, the initial candidate spectral combination C corresponding to m3, and the initial candidate spectral combination E corresponding to m5 are all candidate spectral combinations. Also taking the above example as an example, under the condition that the second preset number of initial candidate spectral combinations with the largest spectral difference values are taken as candidate spectral combinations, if the second preset number is 2, and the magnitude relationship of the spectral difference values of the above multiple initial candidate spectral combinations is still m1<m7<m10<m2<m4<m6<m8<m9<m3<m5, then the initial candidate spectral combination C corresponding to m3 is taken as candidate spectral combination C, and the initial candidate spectral combination E corresponding to m5 is taken as candidate spectral combination E. In other words, the initial candidate spectral combinations corresponding to the top 20% of the largest spectral difference values among the at least one initial candidate spectral combination can be taken as the above-mentioned candidate spectral combinations. Refer to Figure 11 , Figure 11 shows the difference of the spectral difference values corresponding to the finally obtained reference spectral combinations under different selection principles. Figure 11The calculation method for the spectral difference value corresponding to Method 1 is as follows: From the spectral difference values ​​corresponding to the initial reference spectral combination A to the spectral difference values ​​corresponding to the initial reference spectral combination J, the initial candidate spectral combination with the largest spectral difference value is selected as the reference spectral combination. The spectral difference value corresponding to the reference spectral combination is then used as the spectral difference value for Method 1. The calculation method for the spectral difference value corresponding to Method 2 is as follows: The initial reference spectral combinations with the largest spectral difference values ​​(top 1%) are selected as candidate spectral combinations, and the largest spectral difference value among the candidate spectral combinations is used as the spectral difference value for Method 2. The calculation method for the spectral difference value corresponding to Method 3 is as follows: The initial reference spectral combinations with the largest spectral difference values ​​(top 10%) are selected as candidate spectral combinations, and the largest spectral difference value among the candidate spectral combinations is used as the spectral difference value for Method 3. The calculation method for the spectral difference value corresponding to Method 4 is as follows: The initial reference spectral combinations are selected as candidate spectral combinations, and the largest spectral difference value among the candidate spectral combinations is used as the spectral difference value for Method 4. Figure 10 as well as Figure 11 It can be seen that although the spectral difference values ​​of the initial candidate spectra are somewhat similar to those of the candidate spectra, if we want to ensure both low computational load and high accuracy in identifying spectral information, we need to determine the spectral difference values ​​of the reference spectral combinations corresponding to the top 10% of the initial reference spectral combinations with the largest spectral difference values.

[0059] For example, the aforementioned preset spectral difference value and the second preset quantity can be determined according to the actual situation, or by those skilled in the art based on experience.

[0060] Step S300: Based on multiple synthesis methods, multiple candidate spectra in each candidate spectral combination are synthesized to obtain multiple synthesized spectra corresponding to each candidate spectral combination.

[0061] "Multiple synthesis methods" can refer to selecting one or more candidate spectra from multiple candidate spectra in the candidate spectral combination, summing them, or performing weighted summation, etc., to obtain different synthetic spectra under different synthesis methods. That is, each candidate spectral combination can yield multiple synthetic spectra.

[0062] In one possible implementation, see [reference] Figure 6 , Figure 6 Step S300 further includes:

[0063] Step S301: Based on at least one preset value, determine at least one set of values ​​corresponding to each candidate spectral combination, and each set of values ​​corresponds to a synthesis method.

[0064] For example, taking candidate spectral combinations including candidate spectral combination C as an example, when at least one value includes the value 0 and the value 1, the value groups corresponding to candidate spectral combination C and candidate spectral combination E can include: value group 0 (corresponding to 000), value group 1 (corresponding to 001), value group 2 (corresponding to 010), value group 3 (corresponding to 011), value group 4 (corresponding to 100), value group 5 (corresponding to 101), value group 6 (corresponding to 110), and value group 7 (corresponding to 111). The synthesis method corresponding to value group 0 is the first candidate spectrum * 0 + the second candidate spectrum * 0 + the third candidate spectrum * 0, which is equivalent to synthesizing a synthesized spectrum without selecting any candidate spectra from the candidate spectral combination; that is, the spectral curve corresponding to the synthesized spectrum is 0. The synthesis method corresponding to numerical group 1 is: first candidate spectrum * 0 + second candidate spectrum * 0 + third candidate spectrum * 1. This is equivalent to selecting the third candidate spectrum from the candidate spectrum combination to synthesize the synthesized spectrum, meaning the spectral curve of the synthesized spectrum is consistent with the third candidate spectrum. Similarly, the synthesis methods corresponding to other numerical groups can be obtained, which will not be elaborated here.

[0065] For example, see Figure 9 As shown, taking the number of values ​​in at least one numerical value as X, and the number of candidate spectra in the candidate spectral combinations as Y (i.e., the first preset quantity is Y) as an example, the number of numerical value groups (i.e., the number of synthesized spectra corresponding to each candidate spectral combination) is equal to X. Y . Figure 9 If at least one of the values ​​contains 2 values, then if the first preset quantity is 3, the number of value groups is 2. 3 That is, 8. With the first preset quantity equal to 4, the number of value groups is equal to 2. 4 That is, 16. The number of values ​​in a value group can be the same as the first preset number mentioned above, which is equal to the number of candidate spectra in the candidate spectrum combination, that is, there are Y values ​​in each value group.

[0066] For example, the numerical value included in at least one of the above values ​​can be determined according to the actual situation, and this disclosure does not limit it.

[0067] Step S302: Linearly combine multiple candidate spectra in each candidate spectral combination with each numerical group to obtain multiple synthetic spectra corresponding to each candidate spectral combination.

[0068] For example, taking candidate spectral combinations including candidate spectral combination C as an example, when at least one value includes the aforementioned value 0 and value 1, the value group corresponding to candidate spectral combination C can include the aforementioned value group 0 to value group 7. The multiple synthetic spectra obtained from candidate spectral combination C include synthetic spectra c0 to synthetic spectra c7, that is, synthetic spectrum c0 (calculated from candidate spectrum A*0 + candidate spectrum B*0 + candidate spectrum E*0), synthetic spectrum c1 (calculated from candidate spectrum A*0 + candidate spectrum B*0 + candidate spectrum E*1), and synthetic spectrum c2 (calculated from candidate spectrum A*0 + candidate spectrum B*0 + candidate spectrum E*1). The composite spectra are calculated as follows: C1 (obtained by adding candidate spectra B*1 and candidate spectra E*0), C3 (obtained by adding candidate spectra A*0, candidate spectra B*1, and candidate spectra E*1), C4 (obtained by adding candidate spectra A*1, candidate spectra B*0, and candidate spectra E*0), C5 (obtained by adding candidate spectra A*1, candidate spectra B*0, and candidate spectra E*1), C6 (obtained by adding candidate spectra A*1, candidate spectra B*1, and candidate spectra E*0), and C7 (obtained by adding candidate spectra A*1, candidate spectra B*1, and candidate spectra E*1).

[0069] For example, the formula for calculating the synthetic spectrum is as follows:

[0070]

[0071] in, Used to represent synthetic spectra, I1 to I k This represents the spectral data corresponding to each sampling point in the synthesized spectrum. C represents the individual candidate spectra in the candidate spectral combination. 1 This represents the first candidate spectrum in the candidate spectral combination (for example, if the candidate spectral combination is candidate spectral combination C, then C...). 1 Indicates the alternative spectrum A), to C represents the spectral data corresponding to each sampling point in the first candidate spectrum among the candidate spectral combinations. i Let represent the i-th candidate spectrum in the candidate spectral combination (taking i=2 and the candidate spectral combination as candidate spectral combination C as an example, then C...). i Indicates candidate spectrum B), to C represents the spectral data corresponding to each sampling point in the i-th candidate spectrum among the candidate spectral combinations. m Let C represent the m-th candidate spectrum in the candidate spectral combination (where m = 2 and the candidate spectral combination is candidate spectral combination C, then C...). m Indicates the candidate spectrum (E). to This refers to the spectral data corresponding to each sampling point in the m-th candidate spectrum among the candidate spectral combinations. L represents a set of numerical values. 1 L represents the numerical value corresponding to the first candidate spectrum in the candidate spectral combination. i L represents the numerical value corresponding to the i-th candidate spectrum in the candidate spectral combination. m This represents the numerical value corresponding to the m-th candidate spectrum in the candidate spectral combinations. For example, if i equals 2 and m equals 3, Taking the above numerical group 2, with candidate spectral combination C as an example, This represents the synthetic spectrum c2 (i.e., candidate spectrum A*0 + candidate spectrum B*1 + candidate spectrum E*0) in the above candidate spectrum combination C.

[0072] Step S400: Based on the multiple synthetic spectra corresponding to each candidate spectral combination, determine the multiple target synthetic spectra corresponding to each candidate spectral combination.

[0073] Optionally, the dimension of the target synthesized spectrum is lower than that of the synthesized spectrum. The dimension of the spectrum is the number of sampling points in the spectrum. For example, when the spectrum is between 400nm and 750nm, if sampling is performed at a rate of one sampling point per 1nm, the total number of sampling points in the spectrum is 351, and the dimension of the spectrum is 351.

[0074] For example, taking candidate spectral combinations, including candidate spectral combination C, as an example, candidate spectral combination C includes synthetic spectra c0 to synthetic spectra c7. Multiple target synthetic spectra corresponding to candidate spectral combination C can be obtained by dimensionality reduction processing of each synthetic spectrum in candidate spectral combination C. For example, if the dimension of the synthetic spectrum is 351, that is, each synthetic spectrum from synthetic spectrum c0 to synthetic spectrum c7 corresponds to 351 sampling points, and each sampling point corresponds to one spectral data point, those skilled in the art can, based on experience, select spectral bands in the synthetic spectrum that have a smaller impact on the judgment of spectral difference (hereinafter referred to as invalid bands). By deleting the spectral data corresponding to invalid bands in each synthetic spectrum, dimensionality reduction of the spectral data is achieved, thereby generating the target synthetic spectrum. For example: See Figure 12 As shown, Figure 12 This demonstrates the relationship between the influence of each band of the generated synthetic spectrum on the judgment of spectral difference and the spectral wavelength, under different first preset quantities. Specifically, curve 1 corresponds to a first preset quantity of 3, curve 2 to 5, curve 3 to 8, and curve 4 to 10. Figure 12It can be seen that the influence of each curve on the spectral difference judgment after 700nm is 0 (that is, the spectral curves of different synthesized spectra are similar after 700nm). Therefore, the bands greater than 700nm in the synthesized spectrum can be regarded as the above invalid bands. When comparing the spectral differences between synthesized spectra, the target synthesized spectrum can be generated by deleting the data corresponding to the bands greater than 700nm in each synthesized spectrum, thereby realizing the dimensionality reduction of the synthesized spectrum and reducing the amount of computation required when comparing the spectral differences of the synthesized spectra.

[0075] In one possible implementation, see [reference] Figure 7 , Figure 7 Step S400 further includes:

[0076] Step S401: Determine multiple target synthetic spectra corresponding to each candidate spectral combination based on the multiple synthetic spectra corresponding to each candidate spectral combination and the preset weights corresponding to each band of the synthetic spectra.

[0077] For example, the preset weights corresponding to each band of the synthesized spectrum can be determined based on the experience of those skilled in the art or a preset weight calculation method. For instance, as mentioned earlier, when the candidate spectrum is a quantum dot spectrum, since the spectral curves of different quantum dot spectra differ significantly between 400nm and 700nm, while the spectral curves are similar after 700nm, the preset weight of the band between 400nm and 700nm in the synthesized spectrum can be set to 10, and the preset weight of the band greater than 700nm in the synthesized spectrum can be set to 0. Another example is that the contribution value of each band of the synthesized spectrum to the calculation of spectral difference can be determined using the PCA algorithm (i.e., principal component analysis algorithm), that is, the influence of the spectral data corresponding to each band of the synthesized spectrum on the calculation result of spectral difference, thereby determining the preset weights corresponding to different bands of the synthesized spectrum. In other words, the preset weights are used to represent the influence of each band of the synthesized spectrum on the calculation of the spectral difference between multiple synthesized spectra.

[0078] For example, the target synthesized spectrum includes bands with preset weights greater than a preset threshold from multiple synthesized spectra of corresponding candidate spectral combinations. For instance, if the preset weight for a band between 400nm and 500nm is 10, for a band between 500nm and 600nm it is 6, for a band between 600nm and 630nm it is 4, for a band between 630nm and 700nm it is 2, and for a band greater than 700nm it is 0, and the preset threshold is 5, then the target synthesized spectra c0 to c7 corresponding to synthesized spectra c0 in candidate spectral combination C do not contain spectral data corresponding to bands greater than 600nm. Similarly, the target synthesized spectra e0 to e7 corresponding to synthesized spectra e7 in candidate spectral combination E also do not contain spectral data corresponding to bands greater than 600nm. The method for determining the preset weights can be determined according to actual circumstances, and this disclosure does not limit it here.

[0079] The method for generating a reference spectral combination provided in this disclosure can filter out the spectra corresponding to bands with preset weights less than or equal to preset thresholds in the synthesized spectrum by setting preset weights for different bands, thereby achieving dimensionality reduction of the synthesized spectrum, reducing the amount of computational data, and reducing the time required to generate the reference spectral combination.

[0080] Step S500: Based on the degree of difference between the multiple target synthesized spectra corresponding to each candidate spectral combination, determine at least one reference spectral combination among the candidate spectral combinations.

[0081] For example, a preset algorithm can be used to calculate the spectral difference between at least two target synthetic spectra in each candidate spectral combination, determining the most similar target synthetic spectrum (i.e., the target synthetic spectrum with the smallest spectral difference) corresponding to each target synthetic spectrum. The spectral difference between each target synthetic spectrum in each candidate spectral combination and its most similar target synthetic spectrum is added together, and the average value (or root mean square error, etc.) of the spectral difference among multiple candidate spectra is determined based on the addition result. This average value (or root mean square error, etc.) is then used as the spectral difference value corresponding to the candidate spectral combination. For example, taking the calculation of the spectral difference value of candidate spectral combination C as an example, candidate spectral combination C includes target synthetic spectra c0 to c7. The spectral difference between every two target synthetic spectra from c0 to c7 in candidate spectral combination C can be determined using an Euclidean distance algorithm, generating spectral difference values ​​L1 to L28. Based on spectral difference degrees L1 to L28, the target synthetic spectrum most similar to the target synthetic spectrum c0 is determined as target synthetic spectrum c1; the target synthetic spectrum most similar to target synthetic spectrum c1 is determined as target synthetic spectrum c2; the target synthetic spectrum most similar to target synthetic spectrum c2 is determined as target synthetic spectrum c4; the target synthetic spectrum most similar to target synthetic spectrum c3 is determined as target synthetic spectrum c6; the target synthetic spectrum most similar to target synthetic spectrum c4 is determined as target synthetic spectrum c2; the target synthetic spectrum most similar to target synthetic spectrum c5 is determined as target synthetic spectrum c6; the target synthetic spectrum most similar to target synthetic spectrum c6 is determined as target synthetic spectrum c5; and the target synthetic spectrum most similar to target synthetic spectrum c7 is determined as target synthetic spectrum c6. Therefore, the spectral difference value of the candidate spectral combination C is equal to (L1+L8+L15+L21+L15+L26+L26+L28) / 8. Similarly, the spectral difference value corresponding to the candidate spectral combination E can be obtained.

[0082] For example, multiple synthetic spectra corresponding to the reference spectral combination are used to compare with the spectrum to be identified to obtain the identification result corresponding to the spectrum to be identified. The identification result is represented by the aforementioned numerical groups. For instance, if the reference spectral combination is a candidate spectral combination E, and candidate spectral combination E includes synthetic spectra e0 (corresponding to numerical group 0) to synthetic spectra e7 (corresponding to numerical group 7), and the spectrum to be identified corresponds to synthetic spectrum e0, then the spectrum to be identified corresponds to numerical group 0, i.e., the identification result is 000.

[0083] In one possible implementation, see [reference] Figure 8 , Figure 8 Step S500 further includes:

[0084] Step S501: For any candidate spectral combination, determine the similar numerical groups corresponding to each numerical group.

[0085] For example, the absolute value of the difference between the sum of values ​​in each value group and the sum of values ​​in any value group in the corresponding similar value group is less than a preset difference, and / or, the Hamming distance between each value group and any value group in the corresponding similar value group is less than a preset distance.

[0086] For example, the Hamming distance is used to represent the number of different characters at corresponding positions in two strings of the same length. For instance, if string A is 101 and string B is 010, then the Hamming distance between string A and string B is 3. Here, we also take the candidate spectral combination C as an example, which includes target synthetic spectra c0 to target synthetic spectra c7. As mentioned earlier, target synthetic spectra c0 to target synthetic spectra c7 in candidate spectral combination C correspond to numerical groups 0 to 7, respectively. Therefore, the Hamming distances between the numerical group 0 corresponding to the target synthesized spectrum c0 and the numerical group 1 corresponding to the target synthesized spectrum c1 to the numerical group 7 corresponding to the target synthesized spectrum c7 can be determined as 1, 1, 2, 1, 2, 2, 3 respectively; the Hamming distances between the numerical group 1 corresponding to the target synthesized spectrum c1 and the numerical group 2 corresponding to the target synthesized spectrum c2 to the numerical group 7 corresponding to the target synthesized spectrum c7 can be determined as 2, 1, 2, 1, 3, 2 respectively; the Hamming distances between the numerical group 2 corresponding to the target synthesized spectrum c2 and the numerical group 3 corresponding to the target synthesized spectrum c3 to the numerical group 7 corresponding to the target synthesized spectrum c7 can be determined as 1, 2, 3, 1, 2 respectively; and the Hamming distances between the numerical group 2 corresponding to the target synthesized spectrum c2 and the numerical group 3 corresponding to the target synthesized spectrum c3 to the numerical group 7 corresponding to the target synthesized spectrum c7 can be determined as 1, 2, 3, 1, 2 respectively. The Hamming distances between the corresponding numerical group 3 and the numerical group 4 corresponding to the target synthetic spectrum c4 to the numerical group 7 corresponding to the target synthetic spectrum c7 are 3, 2, 2, and 1, respectively. The Hamming distances between the corresponding numerical group 4 and the numerical group 5 corresponding to the target synthetic spectrum c5 to the numerical group 7 corresponding to the target synthetic spectrum c7 are 1, 1, and 2, respectively. The Hamming distances between the corresponding numerical group 5 and the numerical group 6 corresponding to the target synthetic spectrum c6 to the numerical group 7 corresponding to the target synthetic spectrum c7 are 2 and 1, respectively. The Hamming distance between the corresponding numerical group 6 and the numerical group 7 corresponding to the target synthetic spectrum c7 is 1.

[0087] For example, if the Hamming distance between any value group and its corresponding similar value group is less than a preset distance, the similar value groups corresponding to each value group can be determined by setting a preset Hamming distance. For instance, the preset Hamming distance can be set to 2. Then, similar value groups A corresponding to value group 0 include: value groups 1 to 6; similar value groups B corresponding to value group 1 include: value groups 0, 2, 3, 4, 5, and 7; similar value groups C corresponding to value group 2 include: value groups 0, 1, 3, 4, 6, and 7; and similar value groups D corresponding to value group 3 include: value groups 0, 1, 2, 5, 6, and 7. The similar value groups E corresponding to value group 4 include: value group 0, value group 1, value group 2, value group 5, value group 6, and value group 7. The similar value groups F corresponding to value group 5 include: value group 0, value group 1, value group 3, value group 4, value group 6, and value group 7. The similar value groups G corresponding to value group 6 include: value group 0, value group 2, value group 3, value group 4, value group 5, and value group 7. The similar value groups H corresponding to value group 7 include: value group 1, value group 2, value group 3, value group 4, value group 5, and value group 6.

[0088] For example, see Figure 13 As shown, Figure 13 This demonstrates the relationship between Hamming distance and relative frequency among numerical groups. Among them, Figure 13 In the spectral designation, C represents the number of candidate spectra in each candidate spectral combination. Specifically, 3C indicates that each candidate spectral combination includes three candidate spectra (i.e., the first preset number is 3), 5C indicates that each candidate spectral combination includes five candidate spectra (i.e., the first preset number is 5), 8C indicates that each candidate spectral combination includes eight candidate spectra (i.e., the first preset number is 8), and 10C indicates that each candidate spectral combination includes ten candidate spectra (i.e., the first preset number is 10). Figure 13 The relative frequency in the formula represents the probability that the target synthetic spectrum most similar to the target synthetic spectrum corresponding to the numerical group will appear. For example, in the curve corresponding to 3C, a Hamming distance of 1 indicates the numerical groups with a Hamming distance of 1 from that numerical group. If the numerical group is numerical group 0, then according to the previous text, the numerical groups with a Hamming distance of 1 from numerical group 0 are numerical group 1, numerical group 2, and numerical group 4. Since the relative frequency corresponding to a Hamming distance of 1 is 0.55, the probability that the target synthetic spectrum corresponding to numerical group 1, the target synthetic spectrum corresponding to numerical group 2, and the target synthetic spectrum corresponding to numerical group 4 contains a target synthetic spectrum most similar to the target synthetic spectrum corresponding to numerical group 0 is 0.55.

[0089] For example, according to Figure 13The content can be set to a preset Hamming distance of 5, meaning that when the Hamming distance is greater than 5, the relative frequencies corresponding to each curve tend to be 0.

[0090] For example, the sum of the values ​​in each numerical group is the sum of the values ​​in the numerical group. For instance, in the above example, numerical group 5 is 101, so the sum of the values ​​in numerical group 5 is 1 + 0 + 1 = 2. Here, we also take the candidate spectral combination including candidate spectral combination C as an example. Candidate spectral combination C includes target synthetic spectra c0 to target synthetic spectra c7. As mentioned earlier, target synthetic spectra c0 to target synthetic spectra c7 in candidate spectral combination C correspond to numerical groups 0 to 7, respectively. Therefore, it can be determined that the sum of the values ​​in group 0 corresponding to the target synthetic spectrum c0 is 0, the sum of the values ​​in group 1 corresponding to the target synthetic spectrum c1 is 1, the sum of the values ​​in group 2 corresponding to the target synthetic spectrum c2 is 1, the sum of the values ​​in group 3 corresponding to the target synthetic spectrum c3 is 2, the sum of the values ​​in group 4 corresponding to the target synthetic spectrum c4 is 1, the sum of the values ​​in group 5 corresponding to the target synthetic spectrum c5 is 2, the sum of the values ​​in group 6 corresponding to the target synthetic spectrum c6 is 2, and the sum of the values ​​in group 7 corresponding to the target synthetic spectrum c7 is 3.

[0091] For example, if the absolute value of the difference between the sum of the values ​​in each value group and the sum of the values ​​in any value group in the similar value group corresponding to the value group is less than a preset difference, the similar value group corresponding to each value group can be determined by setting the preset difference. For example, if the preset difference is set to 2, then the similar value group A corresponding to value group 0 includes: value group 1 to value group 6; the similar value group B corresponding to value group 1 includes: value group 0, value group 2 to value group 7; the similar value group C corresponding to value group 2 includes: value group 0, value group 1, value group 3 to value group 7; the similar value group D corresponding to value group 3 includes: value group 0, value group 1, value group 2, value group 4 to value group 7; the similar value group E corresponding to value group 4 includes: value group 0 to value group 3, value group 5 to value group 7; the similar value group F corresponding to value group 5 includes: value group 0 to value group 4, value group 6, value group 7; the similar value group G corresponding to value group 6 includes: value group 0 to value group 5, value group 7; and the similar value group H corresponding to value group 7 includes: value group 1 to value group 6.

[0092] For example, see Figure 14 As shown, Figure 14 This demonstrates the relationship between the differences in the sums of the numerical values ​​and their relative frequencies. Among them, Figure 14 In this context, C represents the number of candidate spectra in each candidate spectral combination (i.e., the first preset number), as detailed above. Figure 13 The description. Figure 14The relative frequency in the formula represents the probability that the target synthetic spectrum most similar to the target synthetic spectrum corresponding to the numerical group will appear. For example, in the curve corresponding to 3C, a difference of -1 between the sums of the numerical values ​​in a numerical group indicates the numerical group with a difference of -1 between its sum and the sum of its numerical values. If the numerical group is numerical group 0, then according to the previous text, the numerical groups with a difference of -1 between their sums of numerical values ​​and that of numerical group 0 are numerical group 1, numerical group 2, and numerical group 4. Since the relative frequency corresponding to a difference of -1 between the sums of the numerical values ​​in a numerical group is 0.25, the probability that the target synthetic spectrum corresponding to numerical group 1, the target synthetic spectrum corresponding to numerical group 2, and the target synthetic spectrum corresponding to numerical group 4 contains a target synthetic spectrum most similar to the target synthetic spectrum corresponding to numerical group 0 is 0.25.

[0093] For example, according to Figure 14 The content can be set to a preset difference of 3, that is, when the difference between the sums of the numerical values ​​in the numerical group is greater than 3 and less than -3, the relative frequencies corresponding to each curve tend to 0.

[0094] For example, the similar value group corresponding to each value group can be determined based on the above-mentioned preset Hamming distance and preset difference. That is, the absolute value of the difference between the sum of the values ​​of each value group and the sum of the values ​​of any value group in the similar value group corresponding to the value group is less than the preset difference, and the Hamming distance between each value group and any value group in the similar value group corresponding to the value group is less than the preset distance.

[0095] Step S502: Determine the spectral difference value corresponding to the candidate spectral combination based on the spectral difference degree between the target synthetic spectrum corresponding to each numerical group and the target synthetic spectrum corresponding to each numerical group in the similar numerical group.

[0096] For example, to determine the spectral difference value of the candidate spectral combination C, the following formula for determining the spectral difference value corresponding to the candidate spectral combination can be applied to improve the recognition accuracy.

[0097]

[0098] Among them, T true For the target synthesis spectrum, T testThe actual synthesized spectrum is defined as either the target synthesized spectrum with added noise or the spectrum actually prepared according to the theoretical generation process of the target synthesized spectrum. k represents the dimension of each target synthesized spectrum (i.e., the dimension of each actual synthesized spectrum). GapT represents the Euclidean distance between the target synthesized spectrum and its corresponding actual synthesized spectrum. From the above formula, it can be seen that the actual target synthesized spectrum corresponding to each target synthesized spectrum can be determined first based on each target synthesized spectrum in the candidate spectral combination C. The Euclidean distance GapT (i.e., Euclidean distances GapT0 to GapT7) between each actual target synthesized spectrum and its corresponding target synthesized spectrum is then calculated.

[0099]

[0100] Among them, T test For the actual synthesized spectrum, T interf The target synthetic spectrum is the most similar to the target synthetic spectrum corresponding to the actual synthetic spectrum in the same candidate spectral combination. k is the number of synthetic spectra in each candidate spectral combination. k is the dimension of each target synthetic spectrum (i.e., the total dimension of each actual synthetic spectrum). GapI represents the difference between GapT and the Euclidean distance between the actual target synthetic spectrum and its most similar target synthetic spectrum. From the above formula, it can be seen that the target synthetic spectrum most similar to the target synthetic spectrum corresponding to each numerical group (0 to 7) can be determined from the similar numerical groups A to H corresponding to each numerical group. Then, the Euclidean distance GapY (i.e., Euclidean distance GapY0 to Euclidean distance GapY7) between each actual target synthetic spectrum and its most similar target synthetic spectrum is determined, where Euclidean distance GapY represents the distance between the actual target synthetic spectrum and its most similar target synthetic spectrum in the above formula. The calculation results are as follows. The difference between the Euclidean distance GapT and the Euclidean distance GapY, GapI (i.e., GapI0 to GapI7), corresponding to the synthesized spectrum of each actual target can be calculated. Specifically, the Euclidean distance GapT is calculated as follows: n - Euclidean distance GapY n , where n takes the values ​​0, 1, 2, 3, 4, 5, 6, and 7.

[0101]

[0102] Where n represents the order number mentioned above, m is the first preset quantity (i.e., the number of candidate spectra in each candidate spectral combination), n m Used to represent the total number of synthesized spectra corresponding to the alternative spectral combinations. Gapl libThis is used to represent the spectral difference value corresponding to the candidate spectral combinations. From the above formula, we can see that the average value of Gapl is calculated as (difference Gapl0 + difference Gapl1 + difference Gapl2 + difference Gapl3 + difference Gapl4 + difference Gapl5 + difference Gapl6 + difference Gapl7) / 8. The average value (or root mean square deviation) of the difference values ​​Y is taken as the spectral difference value corresponding to each candidate spectral combination.

[0103] Step S503: Select the candidate spectral combination with the largest corresponding spectral difference value among at least one candidate spectral combination as the reference spectral combination.

[0104] For example, when the candidate spectral combination includes candidate spectral combination C and candidate spectral combination E, if the spectral difference value corresponding to candidate spectral combination E is the largest, then candidate spectral combination E is used as the reference spectral combination.

[0105] The method for generating reference spectral combinations provided in this disclosure can determine the spectral difference value corresponding to the candidate spectral combination based on the degree of difference between the two target synthetic spectra corresponding to adjacent numerical groups in each candidate spectral combination. This can reduce the amount of data that needs to be processed when calculating the spectral difference value corresponding to the candidate spectral combination and reduce the time consumed in calculating the spectral difference value corresponding to the candidate spectral combination.

[0106] Taking quantum dot spectroscopy as an example application scenario, the alternative spectra may include spectra obtained by quantum dot spectral sensors.

[0107] The spectra obtained by quantum dot spectral sensors are spectrally tunable over a wide spectral band, thus allowing the formation of a spectral candidate library containing multiple different quantum dot spectra (this spectral candidate library can correspond to at least one of the candidate spectra mentioned above). The method for generating reference spectral combinations provided in this disclosure can filter based on the spectral difference values ​​corresponding to each initial candidate spectral combination, identifying initial candidate spectral combinations with large spectral difference values ​​as candidate spectral combinations. This process then determines multiple synthetic spectra corresponding to each candidate spectral combination, reduces the dimensionality of the synthetic spectra to target synthetic spectra, and determines the quantum dot spectral combination used to characterize the numerical data (i.e., the reference spectral combination) based on the degree of difference between the target synthetic spectra. Therefore, the method for generating reference spectral combinations provided in this disclosure can improve the identification accuracy. Furthermore, compared to the "ergodic method" of traversing all possible candidate spectral combinations to determine the reference spectral combination, it reasonably reduces the number of candidate spectral combinations that need to be determined, thereby reducing the number of synthetic spectra that need to be calculated to determine the degree of difference between synthetic spectra. This achieves the goal of reducing the time spent determining the quantum dot spectral combination used to characterize the numerical data and reducing the amount of data that needs to be processed.

[0108] In one possible implementation, see [reference] Figure 15As shown, Figure 15 The present disclosure provides an information storage and retrieval system 100, which includes a receiving spectrometer 110.

[0109] For example, the receiving spectrometer is used to store multiple synthesized spectra obtained by the reference spectrum combination generated by the reference spectrum combination generation method described above.

[0110] In one possible implementation, the information storage and retrieval system further includes a transmitting spectrometer 120.

[0111] For example, the transmitting spectrometer 120 is used to acquire the spectrum to be identified.

[0112] For example, the receiving spectrometer 110 is used to compare the spectrum to be identified with multiple synthetic spectra corresponding to the combination of the reference spectrum to obtain the identification result corresponding to the spectrum to be identified.

[0113] In some embodiments, the functions or modules of the apparatus provided in this disclosure can be used to perform the methods described in the above method embodiments. The specific implementation can be referred to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0114] This disclosure also proposes a computer-readable storage medium storing computer program instructions that, when executed by a processor, implement the above-described method. The computer-readable storage medium can be volatile or non-volatile.

[0115] This disclosure also proposes an electronic device, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0116] This disclosure also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in a processor of an electronic device, the processor in the electronic device performs the above-described method.

[0117] Figure 16 This is a block diagram of an electronic device 1900 for a method of generating a reference spectrum combination according to an exemplary embodiment. For example, the device 1900 may be provided as a server or terminal device. (Refer to...) Figure 16The apparatus 1900 includes a processing component 1922, which further includes one or more processors, and memory resources represented by memory 1932 for storing instructions, such as application programs, that can be executed by the processing component 1922. The application programs stored in memory 1932 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 1922 is configured to execute instructions to perform the methods described above.

[0118] Device 1900 may also include a power supply component 1926 configured to perform power management of device 1900, a wired or wireless network interface 1950 configured to connect device 1900 to a network, and an input / output interface 1958. Device 1900 can operate on an operating system, such as Windows Server, stored in memory 1932. TM Mac OS X TM Unix TM Linux TM FreeBSD TM Or similar.

[0119] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions that can be executed by a processing component 1922 of the device 1900 to perform the above-described method.

[0120] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0121] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0122] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0123] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0124] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0125] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0126] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0127] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0128] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for generating a reference spectral combination, characterized in that, The method includes: Acquire multiple alternative spectra; The candidate spectra from the plurality of candidate spectra are combined to obtain at least one candidate spectrum combination, wherein each candidate spectrum combination contains the same number of candidate spectra; Multiple candidate spectra in each candidate spectral combination are synthesized using various synthesis methods to obtain multiple synthesized spectra corresponding to each candidate spectral combination; Based on the multiple synthetic spectra corresponding to each candidate spectral combination, determine the multiple target synthetic spectra corresponding to each candidate spectral combination; Based on the spectral difference between the multiple target synthetic spectra corresponding to each candidate spectral combination, a reference spectral combination is determined among the at least one candidate spectral combination; wherein, the multiple synthetic spectra corresponding to the reference spectral combination are used to compare with the spectrum to be identified to obtain the identification result corresponding to the spectrum to be identified; The step of combining the candidate spectra from the plurality of candidate spectra to obtain at least one candidate spectrum combination includes: A first preset number of candidate spectra are selected from the plurality of candidate spectra to form at least one initial candidate spectrum combination; Based on the spectral difference degree between the multiple candidate spectra in each initial candidate spectral combination, determine the spectral difference value corresponding to each initial candidate spectral combination; The initial candidate spectral combination with the spectral difference value greater than the preset spectral difference value is used as the candidate spectral combination, or the initial candidate spectral combination with the largest spectral difference value and the second preset number is used as the candidate spectral combination. The step of determining the multiple target synthetic spectra corresponding to each candidate spectral combination based on the multiple synthetic spectra corresponding to each candidate spectral combination includes: Based on the multiple synthetic spectra corresponding to each candidate spectral combination and the preset weights corresponding to each band of the synthetic spectra, multiple target synthetic spectra corresponding to each candidate spectral combination are determined; wherein, the target synthetic spectra include bands in the multiple synthetic spectra of the corresponding candidate spectral combination whose preset weights are greater than preset thresholds, and the preset weights are used to represent the influence of each band of the synthetic spectra on the calculation of the spectral difference between multiple synthetic spectra.

2. The method for generating a reference spectral combination according to claim 1, characterized in that, The step of combining the candidate spectra from the plurality of candidate spectra to obtain at least one candidate spectrum combination includes: A first preset number of candidate spectra are selected from the plurality of candidate spectra to form at least one candidate spectrum combination.

3. The method for generating a reference spectral combination according to claim 1, characterized in that, The process of synthesizing multiple candidate spectra in each candidate spectral combination based on multiple synthesis methods to obtain multiple synthesized spectra corresponding to each candidate spectral combination includes: Based on at least one preset value, determine at least one set of values ​​corresponding to each candidate spectral combination, and each set of values ​​corresponds to a synthesis method; The multiple candidate spectra in each candidate spectral combination are linearly combined with each of the numerical groups to obtain multiple synthetic spectra corresponding to each candidate spectral combination.

4. The method for generating a reference spectral combination according to claim 3, characterized in that, Each candidate spectral combination corresponds to multiple numerical groups. Determining the reference spectral combination among the at least one candidate spectral combination based on the spectral difference between the multiple target synthesized spectra corresponding to each candidate spectral combination includes: For any candidate spectral combination, determine the similar numerical groups corresponding to each of the multiple numerical groups; Based on the spectral difference between the target synthetic spectrum corresponding to each numerical group and the target synthetic spectrum corresponding to each numerical group in the similar numerical group, the spectral difference value corresponding to the candidate spectral combination is determined. The candidate spectral combination with the largest corresponding spectral difference value among the at least one candidate spectral combination shall be used as the reference spectral combination.

5. The method for generating a reference spectral combination according to claim 3, characterized in that, The absolute value of the difference between the sum of the values ​​in each value group and the sum of the values ​​in any value group in the corresponding similar value group is less than a preset difference, and / or The Hamming distance between each value group and any value group in the corresponding similar value group is less than a preset distance.

6. The method for generating a reference spectral combination according to claim 1, characterized in that, The dimension of the target synthetic spectrum is lower than that of the synthetic spectrum.

7. An information storage and retrieval system, characterized in that, The information storage and retrieval system includes: A receiving spectrometer is used to store multiple synthesized spectra corresponding to the reference spectral combination generated by the method for generating the reference spectral combination according to any one of claims 1 to 6.

8. The information storage and retrieval system according to claim 7, characterized in that, The information storage and retrieval system also includes: a transmitting spectrometer; The transmitting spectrometer is used to acquire the spectrum to be identified; The receiving spectrometer is used to compare the spectrum to be identified with multiple synthetic spectra corresponding to the combination of the reference spectrum to obtain the identification result corresponding to the spectrum to be identified.

Images