Hyperspectrum-based smoked mural pigment identification method
Patent Information
- Application Number
- CN202410990353.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-15
AI Technical Summary
然而,目前针对被烟熏污染的壁画颜料识别的研究仍较为匮乏
[0025] The embodiment of the present invention provides a method for identifying smoked mural pigments based on hyperspectral, which includes: collecting spectral curve data of a target pigment area on a smoked mural sample; using a spectral matching model to compare the spectral curve data of the target pigment area with The spectral curve data of multiple standard pigments of standard mural samples obtained in advance are matched one by one. The spectral matching model is established based on the following formula: SMSP=αSI nor +(1-α)(1-SAM nor ), where α is a weighted parameter ranging from 0 to 1, SI nor is the shape index between the spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment among the plurality of standard pigments, SAM nor is the spectral angle mapping index between the spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment; the standard pigment that matches the spectral curve data of the target pigment area is identified as the The type of paint for the target paint area. The method provided by the invention can realize non-destructive, efficient and real-time detection of smoke-contaminated pigments in smoke-contaminated murals, and has high accuracy. The hyperspectral detection equipment used is relatively portable and suitable for on-site use. In addition, hyperspectral technology is used to collect point spectra of murals, using its wide band response range and high sensitivity for spectral detection of surface materials to meet the need for identification of smoked pigments and avoid secondary damage to murals, which has significant Its practical significance and potential for promotion and application.
Smart Images

Figure CN118961609A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of software, and in particular to a method for identifying smoked mural pigments based on a hyperspectrum. Background Art
[0002] Ancient Chinese murals are mainly distributed in grottoes, temples and tombs. Due to the long-term influence of natural factors and human activities, they have suffered from a variety of diseases, which seriously threaten their preservation and value. In particular, smoke damage caused by burning incense, candles, and oil lamps during religious ceremonies in temples, or for heating and cooking, greatly obscures the content and visual effects of the murals. Many murals are covered with surface pigments due to smoke pollution, making it difficult to identify their types. Therefore, non-destructive, efficient and high-precision identification of pigments in smoked murals is of great and urgent practical significance.
[0003] At present, the identification of mural pigments is mainly focused on relatively intact murals, while the identification of pigments in murals after smoke pollution is still a research gap. As a precious historical and cultural heritage, murals have irreplaceable value. Traditional destructive detection methods may cause irreversible damage to murals, such as scraping samples and chemical reagent reactions. In contrast, non-invasive detection technology can avoid these potential risks, reduce physical and chemical damage to murals, and ensure that the integrity and original appearance of cultural relics are not affected.
[0004] Commonly used non-invasive detection techniques include Raman spectroscopy, X-ray fluorescence spectroscopy (XRF) and hyperspectroscopy. Raman spectroscopy is based on the Raman effect produced by the interaction between light and chemical bonds in materials. As an in-situ non-destructive detection technology, it can directly identify the molecular structure of organic and inorganic substances, and provide detailed information about the chemical structure, phase and morphology, crystallinity and molecular interactions of samples. Raman spectroscopy has good spatial resolution and can identify individual pigment particles, so it has become one of the most important analytical characterization methods in the field of cultural relics research. X-ray fluorescence spectroscopy (XRF) is a mature qualitative and quantitative analysis technology that is widely used in the analysis of mural pigments. By obtaining the element information in the pigment, the type of pigment used can be inferred. XRF has the advantages of fast analysis speed and accurate results. For murals dominated by inorganic pigments, the wide element detection range of XRF can detect the elements in most pigments. Compared with minimally invasive analysis methods such as Raman spectroscopy and portable X-ray fluorescence spectroscopy, hyperspectral imaging is a relatively safe non-destructive detection and restoration method. It can safely and quickly obtain large-scale images and reflectance spectra of cultural relics. It is used in the field of pigment analysis due to its advantages such as non-contact, no sampling, high spectral resolution and "graph-in-one". The spectral curve is obtained by using the hyperspectral spectrum to characterize the reflective characteristics of the pigment on the surface of the mural. Because different pigments have different reflective characteristics, this can be used to identify the pigment.
[0005] The analysis and identification of mural pigments is an important application of hyperspectral remote sensing technology in the field of cultural relics protection, and it is also one of the basic links in cultural relics protection and restoration. By calculating the similarity between the standard pigment spectrum in the spectral library and the measured spectrum, the pigment with the highest similarity in the spectral library is regarded as the type of unknown pigment. In this process, pigment spectrum matching is a key link. The spectral matching algorithm not only improves the efficiency and accuracy of pigment identification, but also provides a scientific basis for the protection and restoration of cultural relics, which has important practical significance and application value. However, there is still a lack of research on the identification of mural pigments polluted by smoke. For smoke-polluted murals, there is still a lack of effective spectral matching algorithms to accurately identify the type of polluted pigments. Therefore, conducting research on the identification of smoky mural pigments and constructing corresponding spectral matching algorithms are the key to solving this problem, which has important academic research value and application prospects. Summary of the invention
[0006] An object of the present invention is to solve at least the above problems and / or disadvantages and to provide at least the advantages which will be described hereinafter.
[0007] An object of the present invention is to provide a method for identifying smoky mural pigments based on hyperspectral, which can achieve non-destructive, efficient and real-time detection of smoky contaminated pigments in smoky murals with high accuracy.
[0008] In order to achieve these purposes and other advantages according to the present invention, a method for identifying smoked mural pigments based on hyperspectral is provided, comprising:
[0009] Collect spectral curve data of the target pigment area on the smoked mural sample;
[0010] The spectral curve data of the target pigment area is matched one by one with the spectral curve data of multiple standard pigments of the pre-acquired standard mural sample through a spectral matching model, and the spectral matching model is established based on the following formula:
[0011] SMSP=αSI nor +(1-α)(1-SAM nor ),
[0012] Among them, α is a weighting parameter ranging from 0 to 1, SI nor is a shape index between the spectral curve data of the target pigment region and the spectral curve data of the selected standard pigment among the plurality of standard pigments, SAM nor is a spectral angle mapping index between the spectral curve data of the target pigment region and the spectral curve data of the selected standard pigment;
[0013] The standard pigment matching the spectral curve data of the target pigment region is identified as the pigment type of the target pigment region.
[0014] Preferably, in the hyperspectral-based smoky mural pigment identification method, the spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment from the multiple standard pigments are both spectral curve data with a wavelength in the 1500-2500 nanometer band.
[0015] Preferably, in the hyperspectral-based smoky mural pigment identification method, the value of α is 0.5.
[0016] Preferably, in the method for identifying smoked mural pigment based on hyperspectral, in the formula, SI nor and SAM nor All have been homogenized.
[0017] Preferably, in the hyperspectral-based smoky mural pigment identification method, the collecting of spectral curve data of the target pigment area on the smoky mural sample comprises: collecting spectral curve data of the target pigment area on the smoky mural sample by a portable spectroradiometer.
[0018] Preferably, in the hyperspectral-based smoky mural pigment identification method, the standard mural sample is prepared by:
[0019] The standard mural sample is produced according to the mural production process, white lime paste is evenly applied on the surface of the substrate of the standard mural sample to form a white lime layer, and a plurality of standard pigments are applied on the white lime layer to form a plurality of standard color blocks.
[0020] Preferably, in the hyperspectral-based smoky mural pigment identification method, after preparing the standard mural sample, the method further comprises:
[0021] The standard mural sample is subjected to continuous smoke treatment until the standard mural sample is completely covered with smoke dust, thereby obtaining the smoked mural sample at the stage of constructing the spectral matching model.
[0022] Preferably, in the hyperspectral-based smoked mural pigment identification method, the continuous smoked mural sample is subjected to smoke treatment until the standard mural sample is completely covered with smoke, and the smoked mural sample is obtained in the stage of constructing the spectral matching model, comprising:
[0023] A combustible material is placed under the standard mural sample, and the combustible material is ignited to simulate a smoky environment, and the label mural sample is continuously smoked until the label mural sample is completely covered with smoke, wherein the combustible material includes wood chips, carbon, candle particles and incense.
[0024] The present invention has at least the following beneficial effects:
[0025] The embodiment of the present invention provides a method for identifying pigments of smoked murals based on hyperspectral, comprising: collecting spectral curve data of a target pigment area on a smoked mural sample; matching the spectral curve data of the target pigment area with spectral curve data of a plurality of standard pigments of a pre-acquired standard mural sample one by one through a spectral matching model, wherein the spectral matching model is established based on the following formula: SMSP = αSI nor +(1-α)(1-SAM nor ), where α is a weighting parameter ranging from 0 to 1, SI nor is a shape index between the spectral curve data of the target pigment region and the spectral curve data of the selected standard pigment among the plurality of standard pigments, SAM nor It is the spectral angle mapping index between the spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment; the standard pigment that matches the spectral curve data of the target pigment area is identified as the pigment type of the target pigment area. The method provided by the present invention can realize non-destructive, efficient and real-time detection of smoke-contaminated pigments in smoked murals, and has high accuracy. The hyperspectral detection equipment used is relatively portable and suitable for on-site use. In addition, by using hyperspectral technology to collect point spectra of murals, using its wide band response range and high sensitivity to spectral detection of surface materials, the demand for identifying smoky pigments is met, and secondary damage to murals is avoided, which has significant practical significance and potential for promotion and application.
[0026] Other advantages, objectives and features of the present invention will be embodied in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1 It is a flow chart of a method for identifying smoked mural pigment based on hyperspectrum in an embodiment of the present invention;
[0028] Figure 2a The spectrum of malachite before and after smoking in the embodiment of the present invention;
[0029] Figure 2b The spectrum of Paris green before and after smoking in the embodiment of the present invention;
[0030] Figure 2c Spectra of green agate before and after smoking in an embodiment of the present invention;
[0031] Figure 2d Spectra of chlorotacamite before and after fumigation in the embodiment of the present invention. DETAILED DESCRIPTION
[0032] The present invention is further described in detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.
[0033] like Figure 1 As shown, an embodiment of the present invention provides a method for identifying pigments of smoked murals based on hyperspectral, comprising: collecting spectral curve data of a target pigment area on a smoked mural sample; matching the spectral curve data of the target pigment area with spectral curve data of a plurality of standard pigments of a pre-acquired standard mural sample one by one through a spectral matching model, wherein the spectral matching model is established based on the following formula: SMSP = αSI nor +(1-α)(1-SAM nor ), where α is a weighting parameter ranging from 0 to 1, SI nor is a shape index between the spectral curve data of the target pigment region and the spectral curve data of the selected standard pigment among the plurality of standard pigments, SAM nor It is a spectral angle mapping index between the spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment; the standard pigment matching the spectral curve data of the target pigment area is identified as the pigment type of the target pigment area.
[0034] In a preferred embodiment, in the hyperspectral-based smoky mural pigment identification method, the spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment from the multiple standard pigments are both spectral curve data with a wavelength in the 1500-2500 nanometer band.
[0035] In a preferred embodiment, in the hyperspectral-based smoky mural pigment identification method, the value of α is 0.5.
[0036] In a preferred embodiment, in the hyperspectral-based smoky mural pigment identification method, in the formula, SI nor and SAM nor All have been homogenized.
[0037] In a preferred embodiment, in the hyperspectral-based smoky mural pigment identification method, the collecting of spectral curve data of the target pigment area on the smoky mural sample includes: collecting spectral curve data of the target pigment area on the smoky mural sample by a portable spectroradiometer.
[0038] In a preferred embodiment, in the hyperspectral-based smoky mural pigment identification method, the standard mural sample is produced in the following manner: the standard mural sample is produced according to a mural production process, white lime paste is evenly applied on the substrate surface of the standard mural sample to form a white lime layer, and a plurality of standard pigments are applied on the white lime layer to form a plurality of standard color blocks.
[0039] In a preferred embodiment, in the hyperspectral-based smoky mural pigment identification method, after preparing the standard mural sample, the method further includes: continuously subjecting the standard mural sample to a smoke treatment until the standard mural sample is completely covered with smoke, thereby obtaining the smoky mural sample at the stage of constructing the spectral matching model.
[0040] In a preferred embodiment, in the hyperspectral-based smoked mural pigment identification method, the standard mural sample is continuously smoked until the standard mural sample is completely covered with smoke, and the smoked mural sample is obtained in the construction stage of the spectral matching model, including: placing a combustible material under the standard mural sample, igniting the combustible material, simulating a smoky environment, and continuously smoked the labeled mural sample until the labeled mural sample is completely covered with smoke, wherein the combustible material includes sawdust, carbon, candle particles and incense.
[0041] In summary, the method provided by the present invention can realize non-destructive, efficient and real-time detection of smoke-contaminated pigments in smoked murals, and has high accuracy. The detection equipment used is relatively portable and suitable for on-site use. In addition, by using hyperspectral technology to collect point spectra of murals, the wide band response range and high sensitivity of surface material spectrum detection are utilized to meet the needs of identifying smoke pigments and avoid secondary damage to murals, which has significant practical significance and potential for promotion and application.
[0042] A specific embodiment is provided below to further illustrate the hyperspectral-based smoky mural pigment identification method provided by the embodiment of the present invention, and to illustrate the process of constructing the spectral matching model.
[0043] 1. Overall flow chart
[0044] The present invention first collects point spectral data of the standard mural sample, and then places the standard mural sample in a smoke simulation device to smoke the simulated sample. The smoke device is 73 cm high and 34 cm wide. Combustion materials are placed at the bottom of the device. Then, a mesh iron disc is placed in the middle and upper part of the mural, and the sample is placed on the disc. Wood chips, carbon, candle particles and incense are used as combustible materials to simulate the smoke environment, and the sample is continuously smoked until the sample is completely covered with smoke. Then, point spectral data is collected again for the smoked mural sample. By comparing the spectral data before and after, it is found that smoke has little effect on the spectrum after 1500nm. Based on this, the present invention limits the spectral range to 1500-2500nm, and constructs a spectral matching algorithm for smoked pigments through spectral shape index and spectral angle. The overall block diagram is shown as follows Figure 1 shown.
[0045] 2. Sample preparation and data collection
[0046] With reference to the production process of ancient murals, standard mural samples are made, and after the samples are completely dried according to the standards, white lime paste is evenly applied on their surfaces as the base color of the pigment layer. After investigating the relevant literature on ancient murals, the pigments are divided into four color systems: red, yellow, green and blue. The red system selects cinnabar, real silver beads (synthetic cinnabar), red lead and earth red, the yellow system selects orpiment, realgar, realgar, realgar and ochre, the green system selects malachite, Paris green, green agate and green cuprite, and the blue system selects azurite, lapis lazuli, ultramarine and indigo. These pigments are all commonly used in ancient Chinese murals.
[0047] In order to ensure the quality of the pigment, the glue-free powder pigment produced by Jiang Sixutang was selected, and the gelatin produced by this brand was used to prepare a certain concentration of glue for the preparation of the pigment sample. Subsequently, different pigments were used to make color blocks of 3.5×3.5 cm in size on the white lime layer. Thus, the standard mural sample was obtained. The portable spectroradiometer of FieldSpec4 was used to collect point spectra of the specified pigment area of the mural. Before the collection, the white board reflection correction was performed to convert the surface reflection DN value of the mural sample into reflectivity. Each point was collected once in four directions to ensure the average of the spectrum collection. Subsequently, the sample was smoked continuously for 60 minutes until the sample was completely covered with smoke. Then, the portable spectroradiometer of FieldSpec4 was used to collect the spectrum of the smoked sample again as described above.
[0048] 3. Spectral analysis and matching method
[0049] Taking green pigment as an example, an in-depth comparison is made between the original spectrum and the spectrum after smoking, and the changes in the pigment spectrum before and after smoking are studied, which provides a basis for constructing a pigment spectrum matching method.
[0050] Figure 2aand Figure 2b The spectra of malachite and Paris green before and after smoking are shown. The chemical formula of malachite is CuCO3·Cu(OH)2. Its reflectance spectrum has a reflection peak at 520nm and Cu in the range of 550-1000nm. 2+ In addition, there are strong broad absorption features of CO3 at 2270nm and 2350nm 2- After smoking, the reflection peak at 520nm in the visible light range disappears. However, in the short-wave infrared region, in addition to the decrease in reflectivity, CO3 2- and OH - The absorption characteristics of the product remain basically unchanged, showing strong resistance to smoke.
[0051] Paris green is chemically called copper acetate arsenate, which is a double salt of copper acetate and copper arsenate, with the chemical formula Cu(C2H3O2)2·3Cu(AsO2)2. It is a synthetic pigment. After smoking, the reflection peak at 500nm in the visible spectrum disappears. However, in the short-wave infrared region, except for the decrease in reflectivity, the reflection and absorption characteristics remain basically unchanged.
[0052] Figure 2c and Figure 2d The spectra of green agate and cuprate before and after smoking are shown. The main mineral component of green agate is quartz, with the chemical formula SiO2. Its reflectance spectrum gradually rises in the visible light range, reaches a maximum reflectance of about 0.9 at 550nm, then slowly decreases and stabilizes at about 0.85. After smoking, the reflectance peak in the visible light range disappears, and an obvious small absorption valley appears at 2312nm. There are also small reflection peaks and absorption valleys between 2315nm and 2370nm.
[0053] Chlorotachite, also known as basic copper chloride, is a halogen mineral with a chemical composition of Cu2(OH)3Cl. It has a reflection peak at 520nm and shows the same wavelength as Cu in the range of 550-1000nm. 2+ Broad absorption features with consistent spectral characteristics. After smoke exposure, the reflection peak in the visible light range disappears. In the short-wave infrared region, except for the decrease in reflectivity, the absorption characteristics of OH- have almost no change (around 2350nm and 2500nm), which also shows that OH- ions have strong resistance to smoke exposure.
[0054] By analyzing the spectral curves of 16 pigments before and after smoking, the following conclusions were drawn:
[0055] (1) Smoke mainly affects the visible spectrum and near-infrared spectrum, with the visible spectrum being the most seriously affected. After being affected by smoke, almost all reflection peaks and absorption valleys of the 16 pigments in the visible spectrum disappeared.
[0056] (2) In contrast, the effect of smoke on the short-wave infrared spectrum is relatively small (especially after 1500nm). Among the 16 pigments, most of the pigments have little change in the reflectance characteristics of the short-wave infrared spectrum after being affected by smoke, except for a decrease in reflectivity.
[0057] (3) It is worth noting that OH - and CO3 2- Ionic pigments, such as azurite, malachite and cuprate, show strong resistance to smoke pollution and still maintain obvious absorption and reflection characteristics in the short-wave infrared spectrum.
[0058] (4) In addition, lead-containing pigments such as red lead and myrrh are significantly affected by smoke. This is because the lead-containing pigment reacts with carbon at high temperatures, causing significant changes in its overall spectrum.
[0059] Considering that smoke has little effect on the spectrum of most pigments in the short-wave infrared spectrum, the spectral range of 1500-2500nm was selected to construct the spectral matching method. Considering that the spectral reflectance will decrease significantly after smoke pollution, the spectral shape indicators: shape index (SI) and spectral angle mapping (SAM) were compared to construct the spectral matching method of pigments after smoke pollution.
[0060] SI is an indicator used to quantify the shape similarity between two curves or two signals. It mainly considers the local geometric features of the curves, such as the change of gradient and curvature (here we choose gradient), to evaluate their shape similarity. The calculation of SI usually includes the following steps:
[0061] (1) Calculate the gradient. Calculate the gradient of each spectral curve, that is, the rate of change between adjacent points. For spectral vectors t and r, the gradients are grad t and grad r , the mathematical expression is shown in equation (1).
[0062]
[0063] Among them, i represents the index position of the current element in the vector.
[0064] (2) Calculate the dot product and norm of the gradient vector. The mathematical expression is as follows:
[0065]
[0066] (3) Calculate SI. SI is defined as the cosine similarity of two gradient vectors:
[0067]
[0068] Spectral Angle Mapping (SAM) is a method commonly used in spectral data analysis, especially in remote sensing and hyperspectral imaging. SAM is used to measure the similarity between two spectral vectors. Specifically, it determines their similarity by calculating the angle between the two spectral vectors. The smaller the angle, the more similar the two spectral vectors are. Given two spectral vectors t and r, their SAM is defined as follows:
[0069]
[0070] Where t·r is the dot product of two spectral vectors. ‖t‖ and ‖r‖ are the norms of the spectral vectors t and r, respectively. θ is the angle between the two spectral vectors, in radians.
[0071] Based on this, a smoke pigment spectral matching method (SMSP) was constructed. First, the shape index (SI) and spectral angle mapping (SAM) were normalized respectively, and then the normalized SI and SAM were combined using weights to obtain SMSP, whose mathematical expression is shown in equation (5).
[0072]
[0073] The wavelength is limited to the range of 1500-2500nm, where α is a weighting parameter ranging from 0 to 1 and is used to adjust the SI nor and SAM nor The ratio of SI nor and SAM nor The values of are all between [0,1], so the values of SMSP are also between [0,1]. A higher SMSP indicates a higher similarity between the two spectra.
[0074] 4. Dynamic parameter analysis and identification results
[0075] In order to determine the optimal value of parameter α, dynamic parameter analysis was performed. The parameter α was varied from 0.1 to 0.9, and the number of smoke pigments that could be correctly identified using the proposed SMSP method was observed. The results are shown in Table 1. As can be seen from Table 1, when the value of α is 0.5, the number of correctly identified pigments reaches the maximum value and no longer increases. This shows that the recognition effect is best when the ratio of SI and SAM is 50 / 50. Table 2 shows the pigments that can be correctly identified when α=0.5. The red font in the table indicates the correctly identified pigments. Among them, the highest score in the identification results of cinnabar after smoking is real silver beads, and the second highest is cinnabar. Since they have the same chemical formula, it is considered that the identification of real silver beads is also counted as a correct result. As shown in Table 2, the proposed SMSP method successfully identified 10 of the 16 smoke pigments.
[0076] Table 1 Number of pigments correctly identified under different parameters
[0077] α 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Correctly identify the number of pigments 7 7 8 9 10 10 10 10 10
[0078] Table 2 Pigments that can be correctly identified by SPSM (1500-2500nm)
[0079] pigment SMSP Identification Results Cinnabar Real Silver Beads (Second to Cinnabar) Real silver beads Real silver beads Red Lead Paris Green Earthy Red Earthy Red realgar Green Agate Orpiment Paris Green Myrrh Real silver beads ocher ocher Stone Green Stone Green Paris Green Paris Green Green Agate Cyanine Chlorotachite Chlorotachite Azurite Azurite lapis lazuli Cinnabar ultramarine ultramarine Cyanine Cyanine
[0080] SMSP showed the best recognition performance for pigments that still have obvious features in the short-wave infrared region after smoking. However, for pigments such as orpiment, realgar, lapis lazuli and green agate whose curves have lost obvious features after smoking, the recognition performance is poor. Pigments that have undergone chemical changes, such as red lead and realgar, cannot be recognized. Table 3 shows the recognition results of expanding the band range of the SMSP index to 350-2500nm. It can be seen that the number of correctly identified pigments is significantly reduced, which also proves that limiting the band range of SMSP to 1500-2500nm has the best recognition effect.
[0081] Table 3 Pigments that SPSM can correctly identify (350-2500nm)
[0082] pigment SMSP Identification Results Cinnabar Real Silver Beads (Second to Cinnabar) Real silver beads realgar Red Lead Paris Green Earthy Red Earthy Red realgar Paris Green Orpiment Paris Green Myrrh Chlorotachite ocher Paris Green Stone Green Stone Green Paris Green Cyanine Green Agate Paris Green Chlorotachite Chlorotachite Azurite Azurite lapis lazuli Paris Green ultramarine Paris Green Cyanine Paris Green
[0083] 5. Technical effects of the hyperspectral-based smoke mural pigment identification method provided by the present invention
[0084] (1) Non-destructive and efficient identification: This method uses hyperspectral technology to conduct SMSP analysis in the 1500-2500 nanometer band, which can quickly and efficiently identify the types of pigments contaminated by smoke without damaging the murals. This non-destructive testing method avoids the irreversible damage that traditional destructive testing may cause to the murals, ensuring the integrity and original appearance of the cultural relics.
[0085] (2) Portability and field application: The hyperspectral detection equipment relied on by this method is highly portable and suitable for field use. Through point spectral data collection and wide-band response range, it can meet the needs of identifying smoke-contaminated pigments and avoid secondary damage to the murals. This provides a new scientific basis and technical means for the protection of smoke-contaminated murals, which has significant practical significance and potential for promotion and application.
[0086] (3) Accurate identification and wide applicability: The present invention constructs a smoke pigment matching method (SMSP) and determines the optimal parameter configuration through dynamic parameter analysis, thereby improving the recognition accuracy and efficiency. This method successfully identifies a variety of smoke pigments, filling the gap in the current research on smoke mural pigment identification and has important academic research value and application prospects.
[0087] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and the embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily realized. Therefore, without departing from the general concept defined by the claims and equivalent scope, the present invention is not limited to the specific details and the illustrations shown and described here.
Claims
1. A method for identifying smoked mural pigments based on hyperspectral, characterized in that: include: Collect spectral curve data of the target pigment area on the smoked mural sample; The spectral curve data of the target pigment area is matched one by one with the spectral curve data of multiple standard pigments of the pre-acquired standard mural sample through a spectral matching model, and the spectral matching model is established based on the following formula: SMSP=αSI nor +(1-α)(1-SAM nor ), Among them, α is a weighting parameter ranging from 0 to 1, SI nor is a shape index between the spectral curve data of the target pigment region and the spectral curve data of the selected standard pigment among the plurality of standard pigments, SAM nor is a spectral angle mapping index between the spectral curve data of the target pigment region and the spectral curve data of the selected standard pigment; The standard pigment matching the spectral curve data of the target pigment region is identified as the pigment type of the target pigment region.
2. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 1, characterized in that: The spectral curve data of the target pigment area and the spectral curve data of the selected standard pigment among the multiple standard pigments are both spectral curve data with a wavelength in the 1500-2500 nanometer band.
3. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 2, characterized in that: The value of α is 0.
5.
4. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 2, characterized in that: In the above formula, SI nor and SAM nor All have been homogenized.
5. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 1, characterized in that: The collecting of the spectral curve data of the target pigment area on the smoky mural sample includes: collecting the spectral curve data of the target pigment area on the smoky mural sample by a portable spectroradiometer.
6. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 1, characterized in that: The standard mural sample is produced in the following way: The standard mural sample is produced according to the mural production process, white lime paste is evenly applied on the surface of the substrate of the standard mural sample to form a white lime layer, and a plurality of standard pigments are applied on the white lime layer to form a plurality of standard color blocks.
7. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 6, characterized in that: After preparing the standard mural sample, the method further comprises: The standard mural sample is subjected to continuous smoke treatment until the standard mural sample is completely covered with smoke dust, thereby obtaining the smoked mural sample at the stage of constructing the spectral matching model.
8. The method for identifying smoked mural pigment based on hyperspectrum as claimed in claim 7, characterized in that: The method of continuously performing smoke treatment on the standard mural sample until the standard mural sample is completely covered with smoke dust, thereby obtaining the smoked mural sample in the stage of constructing the spectrum matching model, comprises: A combustible material is placed under the standard mural sample, and the combustible material is ignited to simulate a smoky environment, and the label mural sample is continuously smoked until the label mural sample is completely covered with smoke, wherein the combustible material includes wood chips, carbon, candle particles and incense.
Citation Information
Cited By
Mural smudging degree evaluation method based on hyperspectrum
CN120084739A
Smoked mural pigment spectrum unmixing method and system
CN120253707A
Method and system for spectral unmixing of smoked fresco pigments
CN120253707B
Smoked mural pigment identification method
CN121877766A
Method for identifying a smoke painting
CN121877766B
