KR102561905B1 - Monitoring system and monitoring method for surface contaminant treatment of stone cultural heritages using VNIR hyperspectral images - Google Patents

Abstract

The present invention relates to a monitoring system and method for monitoring the treatment of surface contaminants of stone cultural heritage using visible-near-infrared hyperspectral imaging, and more specifically, to the treatment of surface contaminants of stone cultural heritage using visible-near-infrared hyperspectral imaging. A monitoring system comprising: an image acquisition unit for obtaining visible-near-infrared hyperspectral images of stone cultural properties before and after surface contaminant treatment; an image pre-processing unit which converts the acquired image signals before and after the preservation process into spectral reflectance and performs geometric correction on the image; and an image analyzer for analyzing a change in spectral reflectance from the acquired image; wherein the method for analyzing the change in spectral reflectance in the image analyzer selects a specific wavelength from the acquired image, and Calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image; And calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area; will be.

Description

Monitoring system and monitoring method for surface contaminant treatment of stone cultural heritages using VNIR hyperspectral images}

The present invention relates to a monitoring system and monitoring method for surface contaminant treatment of stone cultural properties using visible-near-infrared hyperspectral imaging.

The visible-near-infrared (VNIR) hyperspectral system is a non-contact and non-destructive technology that can simultaneously acquire spatial and spectral information without preprocessing for analysis or contact with the measurement target. Due to the nature of cultural heritage, damage is irreversible, so the need for non-contact and non-destructive analysis devices is continuously increasing. In particular, in the preservation treatment process, the washing step is the point at which the surface state changes, but the physical properties are weakened, so a non-contact and non-destructive analysis method is required.

XRD, XRF, SEM, spectrophotometer, spectrophotometer, etc. have been used to analyze surface changes caused by conventional preservation treatment, but these methods have the disadvantage of requiring pretreatment for analysis or contact with the measurement target. there is. On the other hand, the visible-near-infrared (VNIR) hyperspectral system enables non-destructive analysis of situations involving color changes, such as before and after conservation treatment, or coated organisms that reflect near-infrared rays, without pretreatment.

Research using hyperspectral imaging for cultural properties is being actively conducted overseas. In particular, from the late 1990s to the early 2000s, non-destructive analysis of cultural assets was proposed using hyperspectral image analysis, and investigations of colored cultural assets such as painting cultural assets and murals, digitizing and character restoration of ancient documents, restoration of lost patterns or virtual restoration, etc. of research has been conducted. However, conventional studies using hyperspectral imaging either utilized a spectrophotometer or did not fully utilize the advantages of an imaging spectrometer and only compared spectral reflection curves. In addition, in Korea, there is a case in which hyperspectral images were used to compare before and after conservation treatment through vegetation index in the long-term monitoring process (Non-Patent Document 1), but the limitations of not being able to calculate the quantitative change area of spectral reflectance have

Therefore, the inventors of the present invention, as a system and method for monitoring surface contaminant treatment of stone cultural properties using visible-near-infrared hyperspectral images, visualize the quantitative change in spectral reflectance through hyperspectral images before and after surface contaminant removal treatment. We developed a system and method that can do this and completed the present invention.

Journal of Conservation Science 2021;37(6):659-669

The purpose of one aspect is

It is to provide a monitoring system and monitoring method for surface contaminant treatment of stone cultural properties.

In order to achieve the above purpose,

On one side,

A system for monitoring surface contaminant treatment of stone cultural properties using visible-near-infrared hyperspectral imaging,

An image acquisition unit for acquiring visible-near-infrared hyperspectral images of the stone cultural property before and after surface contaminant treatment;

an image pre-processing unit which converts the acquired image signals before and after the surface contaminant treatment into spectral reflectance and performs geometric correction on the image; and

An image analyzer for analyzing the amount of change in spectral reflectance from the acquired image;

The method for analyzing the spectral reflectance change in the image analyzer is,

selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength; and

A monitoring system for surface contaminant treatment of stone cultural assets is provided, comprising: calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area.

The surface contaminant is preferably a black inorganic contaminant, and the specific wavelength is 800 nm to 850 nm.

At this time, in the step of calculating the standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area, the standard deviation value for the positive change amount among the calculated spectral reflectance change amount is calculated, It may be a step of performing monochromation based on this and calculating the spectral reflectance change area.

The surface contaminant treatment may include laser cleaning.

The cultural heritage monitoring system

It may further include an image output unit that outputs at least one of a change portion, a change area, and a change amount of the spectral reflectance from the quantitative analysis result of the change amount of the spectral reflectance of the image analyzer.

In addition, the cultural heritage monitoring system

At least one of a removal site, a removal area, and a removal amount of surface contaminants may be calculated from the result of quantitative analysis of the change in spectral reflectance.

In addition, the cultural heritage monitoring system

The image acquisition unit obtains visible-near-infrared hyperspectral images of stone cultural properties at different points of time after surface contaminant treatment,

The image analysis unit analyzes the amount of change in spectral reflectance from the obtained visible-near-infrared hyperspectral images of stone cultural properties at different points in time,

It is possible to monitor the degree of surface damage caused by surface contaminant treatment of stone cultural properties.

In addition, the cultural heritage monitoring system

The image acquisition unit obtains visible-near-infrared hyperspectral images of stone cultural properties before and after treating surface contaminants under different conditions of a specific contaminant removal method,

The image analyzer analyzes the spectral reflectance change from the visible-near-infrared hyperspectral image of the stone cultural property before and after treating the surface contaminants under different conditions obtained above,

Critical conditions of the specific contaminant removal method can be calculated.

on the other side

As a monitoring method performed by a monitoring system for surface contaminant treatment of stone cultural properties,

Obtaining visible-near-infrared hyperspectral images of the stone cultural property before and after surface contaminant treatment by an image acquisition unit; and

converting the acquired image signals before and after the surface contaminant treatment into spectral reflectance by an image pre-processing unit and performing geometric correction on the image;

Analyzing the amount of change in spectral reflectance from the acquired image performed by an image analyzer; Including,

In the step of analyzing the spectral reflectance change,

selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength; and

A monitoring method for surface contaminant treatment of stone cultural assets is provided, including calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area.

The monitoring system and method for surface contaminant treatment of stone cultural heritage of the present invention has the advantage of being able to quantitatively analyze the surface contaminant removal treatment generated in stone cultural heritage in a non-contact and non-destructive method.

The monitoring system and method for treating surface contaminants of stone cultural properties of the present invention can set threshold values for treatment conditions in various treatment methods for removing surface contaminants, thereby providing a method for removing surface contaminants more effectively. can

The monitoring system and method for treating surface contaminants of stone cultural properties of the present invention can calculate the amount of contaminants removed, and can continuously observe changes by performing monitoring at regular intervals to determine the degree of damage that may occur in the future according to the removal of the contaminants. It has the advantage of being able to monitor.

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

1 is a photograph showing a national treasure, Wonju Beopcheonsa Temple Jigwangguksa Tower (left) and a VNIR hyperspectral system (right), which are stone cultural assets according to an embodiment.
2 is a photograph of a result of geometric correction of hyperspectral images obtained before and after surface contaminant treatment, according to an embodiment.
Figure 3 is a diagram showing the black contaminant removal effect according to the laser power ratio during laser cleaning according to an embodiment, (A) VNIR hyperspectral image before and after laser cleaning, (B) surface image using an optical microscope and (C) laser The spectral reflectance curve according to the power ratio is shown.
4 is a histogram of the number of pixels for spectral reflectance before and after laser cleaning through hyperspectral images before and after surface contaminant treatment obtained according to an embodiment.
5 is an image (left) and a result of image difference analysis (right) for a wavelength of 823 nm among hyperspectral images before and after surface contaminant treatment obtained according to an embodiment,
6 is a monochromatic image (left) and a variation analysis result (right) of hyperspectral images before and after surface contaminant treatment obtained according to an embodiment.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. However, embodiments of the present invention may be modified in various forms, and the scope of the present invention is not limited to the embodiments described below. In addition, the following examples are provided to more completely explain the present invention to those skilled in the art. In addition, "include" a component throughout the specification means that other components may be further included, rather than excluding other components unless otherwise stated.

on one side

A system for monitoring surface contaminant treatment of stone cultural properties using visible-near-infrared hyperspectral imaging,

An image acquisition unit for acquiring visible-near-infrared hyperspectral images of the stone cultural property before and after surface contaminant treatment;

an image pre-processing unit which converts the acquired image signals before and after the preservation process into spectral reflectance and performs geometric correction on the image; and

An image analyzer for analyzing the amount of change in spectral reflectance from the acquired image;

The method for analyzing the spectral reflectance change in the image analyzer is,

selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength; and

A monitoring system for surface contaminant treatment of stone cultural assets is provided, comprising: calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area.

Hereinafter, a monitoring system for treating surface contaminants of stone cultural properties according to an embodiment will be described in detail for each component.

A monitoring system for surface contaminant treatment of stone cultural heritage according to an embodiment is a system for monitoring surface contaminant treatment of stone cultural heritage using visible-near-infrared hyperspectral imaging.

In this case, the surface contaminant treatment is a contaminant generated on the surface of the stone cultural property, and may be an organic contaminant or an inorganic contaminant.

As the surface contaminant treatment, various treatment methods for removing the surface contaminant may be used. The surface contaminant treatment may be laser cleaning.

The visible-near-infrared hyperspectral image contains data in which spectral information and spatial information are fused, and thus, the spectral reflectance change value according to the treatment of surface contaminants of stone cultural properties can be obtained using the hyperspectral image, and through this, the surface The amount of contaminants removed can be quantitatively confirmed.

A monitoring system for treating surface contaminants of stone cultural properties according to an embodiment includes an image acquisition unit that acquires visible-near-infrared hyperspectral images of the stone cultural properties before and after surface contaminant treatment.

The image acquisition unit may acquire a hyperspectral image from an imaging system of a hyperspectral system.

The hyperspectral imaging system may include a light source, an image sensor, and a spectrometer.

In order to obtain the spectral information change value according to the surface contaminant treatment of the stone cultural property using the hyperspectral image acquired by the image acquisition unit, the hyperspectral imaging is preferably performed indoors where external light sources are blocked. Accordingly, as the light source, separate lighting such as a halogen lamp having a wavelength band similar to that of sunlight may be used.

The image sensor may include a hyperspectral camera for obtaining a hyperspectral image and capable of capturing a visible ray and a near infrared ray of 400 nm to 1000 nm and a control unit controlling the camera. For example, the sensor may include a rotating visible-near-infrared range (400 nm-1000 nm) hyperspectral camera (VNIR Spectral Camera PS V10E, SPECIM, Finland) and a computer controlling the camera.

At this time, the camera is rotatable so that it can take images of one or more sides of the stone cultural property where surface contaminants occur. For example, the image acquisition unit measures the east, west, and south surfaces of the upper stylobate stone. And in order to obtain the spectral information change value according to the laser cleaning of the black contaminants formed on the North side, a total of 8 images may be acquired by acquiring images before and after laser cleaning, respectively, for the four surfaces.

The spectrometer may use a grating, a prism, or a Prism-Grating-Prism (PGP) spectrometer in which a prism and a grating are combined.

The image acquisition unit may obtain an image of a specific wavelength in the wavelength band of 400 nm to 1000 nm through the spectrometer.

For example, the image acquisition unit may acquire images for 258 wavelengths in a wavelength band of 400 nm to 1000 nm.

On the other hand, the image acquisition unit then converts the spectral reflectance in the image pre-processing unit, under the same condition of acquiring the hyperspectral image, to obtain an image for a white reference having a reflection characteristic of 99% or more and a shutter of the hyperspectral sensor. Close and acquire an image of the acquired dark current together.

A monitoring system for treatment of surface contaminants of stone cultural assets according to an embodiment includes an image pre-processing unit that converts the obtained image signals before and after the treatment of surface contaminants into spectral reflectance and performs geometric correction on the images.

The image pre-processing unit converts the image signal acquired by the image acquisition unit into spectral reflectance and corrects distortion occurring in the image acquired through the hyperspectral system.

At this time, the reflectance conversion and geometric correction may be performed in the following manner.

Spectral reflectance conversion

Correction of radiation dose of optical images, including hyperspectral images, is a necessary process to ensure or maintain stable image quality or to extract quantitative signal characteristics from data acquired at various times and sensors. The signal obtained in this way needs to be converted into the spectral reflectance of the surface to be analyzed.

Spectral reflectance is defined as the ratio of the radiant flux reflected from the surface to the radiant flux incident on the sensor. For the spectral reflectance conversion, the image signal may be converted into the spectral reflectance by converting the spectral reflectance of the acquired white plate image as a standard.

geometric correction

In the hyperspectral system, geometric distortion occurs depending on the position of the camera or the object to be analyzed, and if this is not corrected, spatial information required for comparison before and after the contaminant removal process cannot be utilized.

Accordingly, geometric correction is performed to minimize such geometric distortion.

At this time, the geometric correction may also be referred to as image rectification and image registration. This geometric correction can be performed using an image to image registration technique, which is a process of transforming and rotating two images of the same area and having similar geometry so that the same objects appear in the same position. there is. In particular, the above technique has advantages in that it is possible to correct between terrestrial VNIR hyperspectral images where map coordinates are absent, and since coordinates are obtained from a reference image, it is easy to superimpose two or more images or analyze changes.

A monitoring system for treatment of surface contaminants of stone cultural properties according to an embodiment includes an image analyzer that analyzes a change in spectral reflectance from the acquired image.

At this time, the method of analyzing the spectral reflectance change in the image analyzer is,

and selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength.

In the above step, among change detection techniques used in the remote sensing field, a change detection technique using Image Difference (ID) may be applied to calculate the change in spectral reflectance before and after surface contaminant treatment. The change detection technique can quantitatively analyze the difference in the state or change of an analysis target using two data at different times, and can perform calculations on multi-temporal images or obtain change information on attributes.

The change detection technique using the image difference means subtracting the second period image band from the first period image band when it is assumed that both bands have the same spectral resolution. If the two images have almost identical radiation characteristics, ideally, the image difference result has a positive or negative number in a pixel where the radiation amount is changed, and has a value of 0 in an area where there is no change. This method can focus on the change and can avoid complex errors in the image, so that the change in spectral reflectance before and after surface contaminant treatment can be more clearly identified.

Meanwhile, the above step may include selecting a specific wavelength to perform a video difference technique in an image. In this case, the specific wavelength may be a wavelength having the largest change in spectral reflectance due to surface contaminant treatment.

For example, an optimal wavelength for comparison before and after laser cleaning may be selected using a ratio of spectral reflectance after cleaning to spectral reflectance before cleaning, which may vary depending on contaminants. For example, in the case of black contaminants, the optimum wavelength may be 800 nm to 800 nm to 850 nm, more preferably 823 nm.

Thereafter, the step includes calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength.

At this time, the change in spectral reflectance before and after the surface contaminant removal treatment can be calculated through the following equation.

ΔBV _ijk (i: row number of pixel, j: column number of pixel, k: band): change in spectral reflectance before and after surface contaminant treatment in the pixel in row j column

BV _ijk (before): Spectral reflectance before surface contaminant treatment at the pixel in row i and column j

BV _ijk (after): Spectral reflectance after surface contaminant treatment at the pixel in row i and column j

VNIR hyperspectral imaging can acquire data that combines spectral and spatial information, and by using this, it is possible to analyze the area where surface contaminants are removed and the spectral reflectance change of the corresponding pixel together.

The change image before and after surface contaminant treatment before performing image difference is expressed in grayscale because each pixel has a change in spectral reflectance and consists of one wavelength of 823 nm, as well as to harmonize with the surrounding colors. Since the contaminants are not completely removed, the brightness change is limited, so it is relatively difficult to compare with the naked eye. has the advantage of being easier to observe.

However, as a result of this image difference, the trend of pixel change can be known, but quantitative comparison is difficult because the standard for the amount of change of each pixel is ambiguous.

Accordingly, the method of analyzing the spectral reflectance change in the image analyzer includes calculating a standard deviation value for the calculated spectral reflectance change, performing monochromation based on the standard deviation value, and calculating a spectral reflectance change area.

In this step, by calculating a standard deviation value for the calculated spectral reflectance change amount and performing modeling as a two-dimensional image by performing monochrome based on this, it is possible to more clearly confirm the spectral reflectance change due to the surface contaminant treatment.

In addition, by applying the monochromatic section and calculating the area of change therefrom, it is possible to analyze the change in spectral reflectance according to the surface contaminant treatment of the stone cultural property.

At this time, it may be more preferable to calculate a standard deviation for pixels excluding pixels that may be included in the error range among the calculated spectral reflectance change amounts.

For example, in the case of black contaminants, since the spectral reflectance increases after washing, it may be more preferable to calculate the standard deviation by excluding negative values or using positive values among the calculated spectral reflectance changes.

The monitoring system for the treatment of surface contaminants of stone cultural properties according to an embodiment further includes an image output unit for outputting at least one of a change part, a change area, and a change amount of the spectral reflectance from the quantitative analysis result of the change amount of the spectral reflectance of the image analysis unit. can include

The monitoring system for treatment of surface contaminants of stone cultural properties according to an embodiment may calculate at least one of a removal site, a removal area, and a removal amount of surface contaminants from a quantitative analysis result of the spectral reflectance change, and through the video output unit. The analysis result can be output.

In addition, the monitoring system for the surface contaminant treatment of stone cultural heritage according to an embodiment may monitor the degree of surface damage caused by the surface contaminant treatment of stone cultural heritage.

More specifically, the image acquisition unit acquires visible-near-infrared hyperspectral images of stone cultural properties at different points of time after surface contaminant treatment,

The image pre-processing unit performs reflectance conversion and geometric correction on the images of different viewpoints acquired by the image acquisition unit,

The image analysis unit may monitor the degree of surface damage caused by the surface contaminant treatment of the stone cultural property by analyzing the amount of change in spectral reflectance from the images of different viewpoints acquired by the image acquisition unit.

In addition, the monitoring system for treating surface contaminants of stone cultural properties according to an embodiment may be used to calculate critical conditions for a specific contaminant removal method.

More specifically, the image acquisition unit acquires visible-near-infrared hyperspectral images of stone cultural properties before and after treating surface contaminants under different conditions of a specific contaminant removal method,

The image pre-processing unit performs reflectance conversion and geometric correction on images before and after processing surface contaminants under different conditions acquired by the image acquisition unit,

By analyzing the spectral reflectance change amount of the image analyzer, it is possible to calculate a critical condition of the specific contaminant removal method.

The monitoring system for surface contaminant treatment of stone cultural heritage according to an embodiment monitors the amount of spectral reflectance change before and after surface contaminant treatment of stone cultural heritage using visible-near-infrared hyperspectral images in the above method, so that stone can be treated in a non-contact and non-destructive manner. It is possible to quantitatively analyze the removal treatment of surface contaminants generated in cultural properties.

In addition, it is possible to continuously observe the change by performing monitoring at a regular period, so that the degree of damage that may occur in the future according to the removal of the contaminants can be monitored.

on the other side

As a monitoring method performed by the monitoring system for the treatment of surface contaminants of the stone cultural heritage,

Obtaining visible-near-infrared hyperspectral images of the stone cultural property before and after surface contaminant treatment by an image acquisition unit; and

converting the acquired image signals before and after the surface contaminant treatment into spectral reflectance by an image pre-processor, and performing geometric correction on the image;

Including; quantitatively analyzing the amount of change in spectral reflectance from the acquired image performed by an image analyzer;

The step of quantitatively analyzing the spectral reflectance change,

selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength; and

A monitoring method for surface contaminant treatment of stone cultural assets is provided, including calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area.

Hereinafter, the present invention will be described in detail through examples and experimental examples.

However, the following examples and experimental examples are only to illustrate the present invention, and the contents of the present invention are not limited by the following examples.

<Example 1>

In the following method, monitoring of stone cultural assets was performed according to black contaminant removal treatment using visible-near-infrared hyperspectral imaging.

(Research target)

In order to monitor stone cultural properties according to the removal of black contaminants using visible-near-infrared hyperspectral imaging, the upper stylobate of the national treasure Wonju Beopcheonsaji Jigwangguksa Pagoda, which is being preserved in an indoor environment as a stone cultural property, was selected as the analysis target (Fig. 1 left see photo).

The upper stylobate is light gray fine-grained granite, and it is made of a single member, and black contaminants are observed on all sides. These black contaminants are very thin and have a high degree of discoloration, but the condition of the rock shows good characteristics. In addition, it was easy to compare before and after cleaning for each surface because it showed concentration at the bottom of the four surfaces except for the top and bottom surfaces. Therefore, VNIR hyperspectral images were acquired before and after laser cleaning of the upper basalt stone for comparison before and after the black contaminant removal treatment.

(removal of black contaminants)

Laser cleaning was performed using a ytterbium-based fiber laser (Yb Fiber Laser) in order to remove the black contaminants formed on the upper basalt stone.

At this time, the maximum power of the laser used is 30 W, the laser transmission length is 3 m, and the laser oscillation frequency is in the range of 50 to 200 kHz. The range of oscillation wavelength is around 1064±3nm, and a red guide laser module that checks the laser oscillation position is installed.

(VNIR hyperspectral system)

A terrestrial hyperspectral system was used to acquire hyperspectral images.

At this time, the sensor for acquiring the hyperspectral image consists of a rotating visible light-near infrared range (400nm-1000nm) hyperspectral camera (VNIR Spectral Camera PS V10E, SPECIM, Finland) and a computer controlling the camera ( See the photo on the right of Figure 1). Image is acquired by push-broom method through Prism Grating Psism (PGP) spectrometer. The camera can rotate 180° horizontally and acquired 258 bands in the available wavelength band. Hyperspectral photography was performed indoors, blocking external light sources as much as possible, and a halogen lamp was used as the light source. The shooting distance was fixed at 380 cm to minimize the distortion of the hyperspectral camera. VNIR hyperspectral images were obtained before and after laser cleaning on the 4 sides of the Upper Paleolithic: East, West, South and North, respectively, and a total of 8 images were taken. did

In addition, an image of a white plate (White Reference, Labsphere, U.S.A) having a Lambertian reflective surface having a reflection characteristic of 99% or more under the same conditions of acquiring the hyperspectral image for the spectral reflectance conversion of the image signal to be performed later, and The shutter of the hyperspectral sensor was closed and an image of the acquired dark current was acquired together.

(Image pre-processing)

In order to convert the obtained image signal into spectral reflectance, convert the image signal into spectral reflectance by converting 99.9%, which is the standard for spectral reflectance of a white plate photographed under the same conditions, into 10,000, and image to image registration The geometric correction was performed using the technique.

2 shows images obtained by performing geometric correction on hyperspectral images obtained before and after laser cleaning using the above method.

(video analysis)

Calculation of spectral reflectance change before and after black contaminant removal treatment

To calculate the spectral reflectance change before and after the black contaminant removal treatment, a change detection technique using image difference was applied.

In order to utilize the video difference technique for comparison before and after laser cleaning, it is desirable to perform calculation by extracting the wavelength most sensitive to the change of contaminants in the image. To this end, it was confirmed that the optimal wavelength for comparison before and after washing was 823 nm by using the ratio of spectral reflectance after washing to spectral reflectance before washing, and this was used for the video difference technique.

<Experimental Example 1> Examination of black contaminant removal effect according to laser power ratio

As a preliminary evaluation to evaluate the cleaning effect of black contaminants generated in the upper podium with VNIR hyperspectral images, in order to confirm the optimal laser power ratio, laser cleaning was performed on a member showing the same black contaminant pattern, and the power ratio under 30W output condition was set to 20% in 5% increments in a total of 4 steps to wash contaminants, and then extract spectral reflection curves from VNIR hyperspectral images and photograph them with an optical microscope. At this time, washing was performed as harmoniously as possible considering the color around the contaminant rather than focusing on the complete removal of the black contaminant of the upper plastid stone.

The black contaminant generated in the upper stylobate is an ecology that occurred on the surface of fine-neutral biotite granite. Inorganic contaminants generated on the surface of rocks, except for bleaching caused by the deposition of water-soluble salts, mostly have low spectral reflectance or show absorption bands at specific wavelengths. In particular, black visible to the human eye is actually a phenomenon caused by absorbing most wavelengths, so spectral reflectance is measured very low when photographed with a VNIR hyperspectral sensor including visible light. Therefore, in this step, the analysis was performed on the premise that the spectral reflectance increased after washing the black contaminant considering that the background layer was granite, and the hyperspectral image preprocessed by the method in Example 1 was performed at each output ratio step. The spectral reflection curves for the separated regions were extracted to compare the cleaning effects, and the results are shown in FIG. 3 .

As shown in FIG. 3 , it can be seen that the effect of cleaning contaminants increases as the laser power ratio increases. In addition, as a result of analyzing the surface state through an optical microscope, it can be confirmed that contaminants remain at 5% and 10% of the laser power ratio, but most of the contaminants are washed at 20%.

In addition, when comparing the spectral reflectance before and after cleaning of each black contaminant, the spectral reflectance at 5% laser power ratio increased by an average of 13.3% compared to before cleaning, increased by an average of 34.9% at 10%, and increased by an average of 54.6% at 15%. In 20%, an average increase of 77.3% was observed. Taken together, L-15% and L-20% can efficiently work when cleaning contaminants, but the contaminants can be completely removed. It can be seen that it is preferable to perform washing in the range of

Through the above results, it can be confirmed that the black contaminant removal effect can be evaluated using VNIR hyperspectral image analysis and the output range for stably removing the contaminant can be confirmed. In addition, since the amount of change in the removal rate of contaminants can be measured, it can be confirmed that the threshold value of the laser output having no or insignificant change in the removal rate of contaminants can be set.

<Experimental Example 2> Comparison of spectral reflectance change before and after black contaminant removal treatment

In the method of Example 1, the spectral reflectance change amount for each pixel distributed in the image before and after washing is calculated using the image difference technique, and the number of pixels for the spectral reflectance value is calculated to confirm the spectral reflectance change trend. The results are shown in Figure 4.

Figure 4 is a result of comparing the changes before and after washing black contaminants on the four sides of the upper stylobate stone. It can be seen that the winter surface has three peak points corresponding to 815, 2548, and 4350 in the spectral reflectance change range before treatment, and after washing, the peak corresponding to 815 shifts to a peak in the 2695 to 4120 band. Seomyeon has three peak points in the range of 740, 2,407, and 3,961, and it is observed that the 740 peak shifts to the range of 2,648 to 3,961 after washing. On the north side, peaks appear at 760, 2,044, and 4,171, and pixels distributed in the range of 760 to 2,044 tend to move to the 2252 to 4080 band after washing. However, in the case of the southern surface, the peak observed between 740 and 815 before washing was not observed, and the peak points appeared as 2,234 and 4,108. appears.

Through the comparison of before and after washing through the above results, it can be confirmed that the pixels showing low spectral reflectance before washing tend to move to higher spectral reflectance while decreasing. In addition, it was found that the pixels with the spectral reflectance value in the range of 0 to 2,000 changed intensively, and the number of pixels with the spectral reflectance of 4,000 or more did not clearly change before and after washing. This suggests that the washing was concentrated on the highly contaminated part due to the washing method for harmony with the surrounding color.

<Experimental Example 3> Spectral reflectance change analysis

The VNIR hyperspectral image can acquire data in which spectral information and spatial information are fused, and by using this, it is possible to analyze the change in the spectral reflectance of the washed area and the corresponding pixel together. The change image before and after washing calculated through the image difference is output in grayscale because each pixel has a change in spectral reflectance and consists of one wavelength of 823 nm.

The drawing on the left of FIG. 5 is an image resulting from visual observation of the 823 nm band and the image difference calculation result of each surface before and after washing, and it can be seen that it is difficult to compare the image with the naked eye. Each surface is expressed in gray scale, and the change in brightness is limited because contaminants are not completely removed to harmonize with the surrounding colors. On the other hand, the right figure of FIG. 5 is an image obtained after performing the video difference technique. The video difference technique improves the brightness change for the part where the change occurs by improving this problem, and it can be confirmed that the erection change is relatively large. there is.

However, as a result of this image difference, the trend of pixel change can be known, but quantitative comparison is difficult because the standard for the amount of change of each pixel is ambiguous.

On the other hand, the larger the change, the brighter the pixel is output. At this time, if monochrome is performed based on the change amount and modeling is performed as a two-dimensional image, the change amount and the washed area can be read through the color of the pixel. In addition, since the spectral reflectance increases when black contaminants are removed, the standard deviation of the spectral reflectance change amount of pixels excluding pixels that may be included in the error range, such as the change amount of negative numbers (-), is calculated and applied to the monochrome section, and finally The change area was calculated and the results are shown in FIG. 6 .

6 is a result of monochrome processing, and the pixels whose spectral reflectance increased to a positive number (+) compared to all pixels were calculated as 9.3% in the east, 20.6% in the west, 21.6% in the south, and 20.6% in the north. In addition, looking at the part where the spectral reflectance changed, the change of 1800 or more tended to be concentrated on the lower part of the winter surface, and there was almost no change in the island part. In the writing, pixels with a high variation appear in black contaminants at the bottom, and changes also appear in the middle part where the icon is engraved. On the southern side, changes were seen in the lower and left-side parts of the road, but the areas where the amount of change was concentrated were not confirmed. In the north face, the change is concentrated in the top and bottom of the road.

Taken together, Dong-myeon, Seo-myeon, and Buk-myeon show that changes are concentrated in the black contaminants at the bottom, but the south-myeon shows a characteristic that changes appear on the island rather than black contaminants at the bottom. In particular, it can be seen that the washing of the black contaminant part concentrated on the lower part was focused, and it can be seen that quantitative comparison of the change in spectral reflectance and the number of pixels for the part where the washing was performed is easy.

Claims (9)

A system for monitoring surface contaminant treatment of stone cultural properties using visible-near-infrared hyperspectral imaging,
An image acquisition unit for acquiring visible-near-infrared hyperspectral images of the stone cultural property before and after surface contaminant treatment;
an image pre-processing unit converting the acquired images before and after the surface contaminant treatment into spectral reflectance and performing geometric correction on the images; and
An image analyzer for analyzing the amount of change in spectral reflectance from the acquired image;
The method for analyzing the spectral reflectance change in the image analyzer is,
selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength; and
Calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating a spectral reflectance change area; monitoring system for surface contaminant treatment of stone cultural assets, including.
According to claim 1,
The surface contaminant is a black inorganic contaminant,
The specific wavelength is 800 nm to 850 nm, a monitoring system for surface contaminant treatment of stone cultural assets.
According to claim 2,
The step of calculating the standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area,
A monitoring system for surface contaminant treatment of stone cultural assets, which calculates a standard deviation value for the positive change in the calculated spectral reflectance change, performs monochromation based on this, and calculates the spectral reflectance change area.
According to claim 1,
The surface contaminant treatment includes laser cleaning, a monitoring system for surface contaminant treatment of stone cultural assets.
According to claim 1,
The monitoring system for the treatment of surface contaminants of the stone cultural heritage
Further comprising: an image output unit for outputting at least one of a change part, a change area, and a change amount of the spectral reflectance from the quantitative analysis result of the change in the spectral reflectance of the image analysis unit; monitoring system for treatment of surface contaminants of stone cultural heritage.
According to claim 1,
The monitoring system for the treatment of surface contaminants of the stone cultural heritage
A monitoring system for surface contaminant treatment of stone cultural assets, which calculates at least one of a removal area, a removal area, and a removal amount of surface contaminants from the quantitative analysis result for the change in spectral reflectance.
According to claim 1,
The monitoring system for the treatment of surface contaminants of the stone cultural heritage
The image acquisition unit acquires visible-near-infrared hyperspectral images of stone cultural properties at different points of time after surface contaminant treatment,
The image analysis unit analyzes the amount of change in spectral reflectance from the obtained visible-near-infrared hyperspectral images of stone cultural properties at different points in time,
A monitoring system for surface contaminant treatment of stone cultural heritage, which monitors the degree of surface damage caused by surface contaminant treatment of stone cultural heritage.
According to claim 1,
The monitoring system for the treatment of surface contaminants of the stone cultural heritage
The image acquisition unit obtains visible-near-infrared hyperspectral images of stone cultural properties before and after treating surface contaminants under different conditions of a specific contaminant removal method,
The image analyzer analyzes the spectral reflectance change from the visible-near-infrared hyperspectral image of the stone cultural property before and after treating the surface contaminants under different conditions obtained above,
A monitoring system for surface contaminant treatment of stone cultural assets, which calculates the critical condition of the specific contaminant removal method.
A monitoring method performed by the monitoring system for surface contaminant treatment of stone cultural heritage of claim 1,
Acquiring visible-near-infrared hyperspectral images of the stone cultural property before and after surface contaminant treatment by an image acquisition unit; and
converting the acquired images before and after the surface contaminant treatment into spectral reflectance by an image pre-processing unit, and performing geometric correction on the images;
Analyzing the amount of change in spectral reflectance from the acquired image performed by an image analyzer; Including,
The step of analyzing the spectral reflectance change,
selecting a specific wavelength from the acquired image, and calculating a change in spectral reflectance before and after surface contaminant treatment in each of the pixels constituting the image at the specific wavelength; and
Calculating a standard deviation value for the calculated spectral reflectance change amount, performing monochromation based on this, and calculating the spectral reflectance change area; monitoring method for surface contaminant treatment of stone cultural assets, including.