• Journal of Conservation Science
• /
• v.35 no.3
• /
• pp.237-244
• /
• 2019

In this study, we applied terahertz imaging technology to three Korean modern oil paintings ('Boy,' 'Girl,' and 'Hyehwadong Landscape'); investigated the types of inner layer information in the pictures that can be extracted with terahertz imaging technology; and analyzed the conditions for extracting them. The biggest problem in the terahertz imaging analysis we encountered was the image distortion caused by the change of the distance between the target surface and the terahertz detector, depending on the surface curvature of the target paintings. We then developed a 'working distance maintaining device' to solve this problem. As a result, the terahertz imaging technique was used to identify the base material characteristics and any patterns of damage inside, and presented the optimal conditions for identifying each characteristic. In addition, it was useful to employ the terahertz frequency-division image to check the characteristics of the background materials. To confirm coloring techniques such as brush strokes, it is effective to compare the maximum reflection peak image with the cross-section image; and to detect damage information inside the paintings that cannot be observed on the surface, to compare the cross-section image with the frequency-division image. On the other hand, according to the terahertz imaging analysis of the oil paintings, the internal structural damage marks of 'Boy' and 'Girl' were confirmed, and the artist's painting style was confirmed in 'Hyehwadong Landscape.' The above results are expected to be useful for the analysis and diagnosis of Korean modern oil paintings for their preservation.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.37 no.1
• /
• pp.76-84
• /
• 2019

This study was conducted for the purpose of examining the Outstanding Universal Value, World Heritage Values, in Suzhou classical gardens and the operational method of the protection management of historic gardens, and the results of the study are as follows; The first, as a world heritage, Suzhou gardens proved OUV by showing the taste and lifestyle of the literati, the unique class of Chinese history, to the natural environment of the East Yangtze Delta region. In addition, it showed the exchange of international culture with unique Chinese garden techniques. Second, a authentic reconstruction of garden have original value because it exist records of the near past. The 20th-century archives described by the contemporary language and describing the heritage through images using such media as photography and drawing played a decisive role in restoring gardens. Third, the protection management of Suzhou gardens, which began in the 1950s, was carried out in detail, including the reconstruction of components such as buildings, the restoration of plant materials and horticulture technique, the creation of a list through the records of trees, the maintenance of props such as plaque, furniture, and the project to improve water quality and restore waterscape. The last, after the World Heritage inscribing, Garden's protection management was proceeded in two directions. One is that through the reorganization of the administrative management system and the enactment of laws related to Suzhou classical Gardens, the listing heritage is strictly protected and managed, and the other is that the historical gardens are used to establish urban masterplan and urban identity in Suzhou. The range of garden properties greatly expanded by distinguishing the principle of conservation from protection to rational use and the grade of protection, such as whole protection, maintenance protection and protection of historical sites.

• Conservation Science in Museum
• /
• v.26
• /
• pp.119-142
• /
• 2021

The gilt-bronze-and-silver shoes, excavated from the Tomb of King Muryeong, revealed many defective parts, which constantly raised in-depth re-investigation and reporting. In that point, scientific treatment and analysis were conducted to investigate the manufacturing techniques applied. The conservation treatment was carried out with an emphasis on the restoration of the original form and joining parts. As a result, the upper part of the right shoe and some part of the back heel were joined, confirming the joining method and the number of rivet used for instep bonding. From the component analysis, it is presumed that the outer plate was gilded with mercury amalgam on pure copper, and pure silver was used for the inner plate. The main pattern of the gilt-bronze-and-silver shoes includes designs of phoenix, lotus, and six-petal flower, while the lotus pattern is high in terms of ratio. Thus, youngnaksa(twisted gilt-bronze thread), geumdongsa(gilt-bronze thread), wondujeong (rivet), and samyeonjeong(quadrangular cleats) were used for bonding each plate. Twisted gilt-bronze thread was used to join the inner and outer plates, while the center and rear of the inner plate were combined with only gilt-bronze threads. From the fact, it is presumed that the joining method was different whether the function was decoration or just bonding. Regarding the manufacturing process, traces of burnishing were confirmed on the outer surface of the gilt-bronze and silver shoes, but no burnishing traces were identified on the inner surface of the silver plate, which is likely to prove more importance put on the finishing of the outer plates rather than that of the inner.

• Conservation Science in Museum
• /
• v.27
• /
• pp.39-56
• /
• 2022

The purpose of this project was to improve the stability of a pair of leather socks for use on snowy days in the collection of the National Museum of Korea (Namsan1567) by conducting conservation treatment and restoring the socks to their original form for use in research and exhibition. Leather socks are referred to in ancient documents with names combining the word "mal" for socks with a term indicating their material (e.g., pimal, meaning leather socks; nokpimal, meaning deer leather socks; jangpimal, meaning roe deer leather socks, or lipimal, meaning racoon dog leather socks) and are mentioned mostly in connection with Jeju Island. Related documents include the Seungjeongwon ilgi (Daily Record of the Grand Secretariat), Injaeilnok (Diary of Injae Jo Geukseon), and Hamel's Journal and a Description of the Kingdom of Joseon. Extant examples of ancient leather socks display the same form as beoseon (traditional Korean socks) and are made of either leather or a combination of leather and fabric. It is likely that such leather socks were worn on Jeju Island to protect the feet from the cold. A condition survey of the leather socks was first conducted to establish a plan for their conservation treatment. Since the socks were in rather poor condition, it was decided to identify their original form through an investigation of relics and pertinent previous studies. The socks were cleaned in consideration of results of the condition survey, and the missing parts around the necks of the socks were reinforced in a reversible manner using counting stitches with cloth dyed to match the original color. Since the bottoms of the socks had lost much of their original form due to deterioration and disintegration in the leather, supports were made and inserted inside the socks to help retain their shape. Through these processes, the structure and characteristics of the socks and the techniques used in their production could be analyzed, their condition was stabilized, and their original form was recovered.

• Conservation Science in Museum
• /
• v.26
• /
• pp.83-182
• /
• 2021

This study summarized the results of the conservation treatment and investigation on the production method of the golden crown (Treasure No. 339) excavated from Tomb No. 129 (also known as Seobongchong Tomb) in Noseo-dong, Gyeongju-si, Gyeongsangbuk-do Province. The golden crown from Seobongchong Tomb was discovered during the excavations conducted by the Museum of the Government-General of Korea in 1926 during the Japanese colonial era. It is currently in the collection of the National Museum of Korea. A total of six Silla golden crowns have survived in Korea, among which the crown from Seobongchong Tomb is the only example with a dome-shaped hemispherical decoration attached with a bird ornament while otherwise showing the typical features of Silla crowns. The crown had been repaired following its excavation using metallic materials and adhesives, but due to the partial deformation and damage brought about by the repair materials, it required further conservation treatment. This article describes in detail the overall process of the conservation treatment and the restoration of the original form of the golden crown from Seobongchong Tomb, particularly the method of reinforcing the joints to secure the stability of the crown. It presents the characteristics of the crown's production as revealed in the investigation during the conservation treatment, and further analyzes the relationship of this crown from Seobongchong Tomb with other Silla crowns through a comparison of their production techniques. The investigation revealed that the crown was primarily decorated with golden sequins at the time of its production. At a later point some of the sequins in the upright ornament were replaced with comma-shaped jade beads and additional comma-shaped jade beads were added to the headband. In order to determine if such modifications to the decoration had occurred with other Silla crowns, the decoration of the six extant Silla golden crowns were investigated. The crown from Cheonmachong Tomb features traces of this same modification to the decoration and possesses other similarities with the crown from Seobongchong Tomb.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.15 no.5
• /
• pp.360-366
• /
• 2022

This paper presents a AI model structure for restoring Korean traditional palace photographs, which remain only black-and-white photographs, to color photographs using Pix2Pix, one of the adversarial generative neural network techniques. Pix2Pix consists of a combination of a synthetic image generator model and a discriminator model that determines whether a synthetic image is real or fake. This paper deals with an artificial intelligence model by adjusting a receptive field of the discriminator, and analyzes the results by considering the characteristics of the ancient palace photograph. The receptive field of Pix2Pix, which is used to restore black-and-white photographs, was commonly used in a fixed size, but a fixed size of receptive field is not suitable for a photograph which consisting with various change in an image. This paper observed the result of changing the size of the existing fixed a receptive field to identify the proper size of the discriminator that could reflect the characteristics of ancient palaces. In this experiment, the receptive field of the discriminator was adjusted based on the prepared ancient palace photos. This paper measure a loss of the model according to the change in a receptive field of the discriminator and check the results of restored photos using a well trained AI model from experiments.

• Journal of the Korean Academy of Esthetic Dentistry
• /
• v.30 no.2
• /
• pp.71-90
• /
• 2021

The virtual patient dataset is a collection of diagnostic data from various sources acquired from a single patient into a coordinate system of three-dimensional visualization. Virtual patient dataset makes it possible to establish a treatment plan, simulate various treatment procedures, and create a treatment planning delivery device. Clinicians can design and simulate a patient's smile on the virtual patient dataset and select the optimal result from the diagnostic process. The selected treatment plan can be delivered identically to the patient using manufacturing techniques such as 3D printing, milling, and injection molding. The delivery of this treatment plan can be linked to the final prosthesis through mockup confirmation through provisional restoration fabrication and delivery in the patient's mouth. In this way, if the diagnostic data superimposition and processing accuracy during the manufacturing process are guaranteed, 3D digital smile design simulated in 3D visualization can be accurately delivered to the real patient. As a clinical application method of the virtual patient dataset, we suggest a decision-making method that can exclude occlusal adjustment treatment from the treatment plan through the digital occlusal pressure analysis. A comparative analysis of whole-body scans before and after temporomandibular joint treatment was suggested for adolescent idiopathic scoliosis patients with temporomandibular joint disease. Occlusal plane and smile aesthetic analysis based on the virtual patient dataset was presented when treating patients with complete dentures.

• Journal of Conservation Science
• /
• v.38 no.3
• /
• pp.234-242
• /
• 2022

The study aimed to investigate the split casting method based on the recreation of the iron-seated Buddha (ISB) statue (number 1971) in the Chuncheon Museum. The statue was designed using three-dimensional scan data and reduced to half-size. Using the existing research results, the ISB statue was created by mold production and split casting. The mold was prepared by mixing sand and clay at a ratio of 3:4 and 1:3 on the outside and inside, respectively, and then casting was done. Various casting defects were observed in the ISB casting and similar shapes were seen. The casting defects included veining or finning, misrun, open or external shrinkage, surface or subsurface blowholes, surface pinholes, and shift. The microstructures were identified as branch-shaped dendrite and pearlite organizations, and black graphite was observed between the cementite organizations. The study findings may be relevant in exploring traditional casting and manufacturing techniques of ISB and may aid in the production of the original form of ISB.

• Journal of Conservation Science
• /
• v.25 no.1
• /
• pp.101-114
• /
• 2009

Potsherds and kilns of the AD 3th century excavated from Gundong and Majeon sites, Korea were studied to investigate the production techniques and provenance of potsherds and kilns on the pastes. For potsherds, kilns blocks and in-situ paleo-soils, provenance of raw materials were estimated through mineralogy and geochemistry, while production technique and thermal feature of kilns were investigated through observation of textures and compositions as well as firing experiment on paleo-soils. As a result of study, potsherds and kilns were found to have similar mineralogical compositions as the neighboring paleo-soils and to have same evolution path with that of geochemistry. The potsherds were divided into 3 groups according to firing temperature and production technique. Group 1 consists of reddish stamped pattern pottery with loose textures, which has many pores and contains many iron oxides. Its temper is less than about 0.5mm, and was probably fired between 700 to $800^{\circ}C$. Group 2 contains ash to grayish blue stamped pattern pottery, which has vitrified texture and few pores. Its temper is less than about 0.5mm, and was probably fired from 900 to $1,000^{\circ}C$. However, some potsherd belongs to the group 2 in terms of features for temper and pastes, but it was probably fired over $1,100^{\circ}C$. Group 3 contains reddish and grayish stamped pattern pottery. It has vitrified matrix, few pores and temper consists of polycrystalline quartz and feldspar over 2mm, and it was probably fired around $1,000^{\circ}C$. The kiln had experienced temperature from 600 to $700^{\circ}C$ on the wall, from 900 to $1,000^{\circ}C$ on the bottom, suggesting the function of high temperature firing.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.175-197
• /
• 2021

Recently, with the development of deep learning technology, research on unstructured data analysis is being actively conducted, and it is showing remarkable results in various fields such as classification, summary, and generation. Among various text analysis fields, text classification is the most widely used technology in academia and industry. Text classification includes binary class classification with one label among two classes, multi-class classification with one label among several classes, and multi-label classification with multiple labels among several classes. In particular, multi-label classification requires a different training method from binary class classification and multi-class classification because of the characteristic of having multiple labels. In addition, since the number of labels to be predicted increases as the number of labels and classes increases, there is a limitation in that performance improvement is difficult due to an increase in prediction difficulty. To overcome these limitations, (i) compressing the initially given high-dimensional label space into a low-dimensional latent label space, (ii) after performing training to predict the compressed label, (iii) restoring the predicted label to the high-dimensional original label space, research on label embedding is being actively conducted. Typical label embedding techniques include Principal Label Space Transformation (PLST), Multi-Label Classification via Boolean Matrix Decomposition (MLC-BMaD), and Bayesian Multi-Label Compressed Sensing (BML-CS). However, since these techniques consider only the linear relationship between labels or compress the labels by random transformation, it is difficult to understand the non-linear relationship between labels, so there is a limitation in that it is not possible to create a latent label space sufficiently containing the information of the original label. Recently, there have been increasing attempts to improve performance by applying deep learning technology to label embedding. Label embedding using an autoencoder, a deep learning model that is effective for data compression and restoration, is representative. However, the traditional autoencoder-based label embedding has a limitation in that a large amount of information loss occurs when compressing a high-dimensional label space having a myriad of classes into a low-dimensional latent label space. This can be found in the gradient loss problem that occurs in the backpropagation process of learning. To solve this problem, skip connection was devised, and by adding the input of the layer to the output to prevent gradient loss during backpropagation, efficient learning is possible even when the layer is deep. Skip connection is mainly used for image feature extraction in convolutional neural networks, but studies using skip connection in autoencoder or label embedding process are still lacking. Therefore, in this study, we propose an autoencoder-based label embedding methodology in which skip connections are added to each of the encoder and decoder to form a low-dimensional latent label space that reflects the information of the high-dimensional label space well. In addition, the proposed methodology was applied to actual paper keywords to derive the high-dimensional keyword label space and the low-dimensional latent label space. Using this, we conducted an experiment to predict the compressed keyword vector existing in the latent label space from the paper abstract and to evaluate the multi-label classification by restoring the predicted keyword vector back to the original label space. As a result, the accuracy, precision, recall, and F1 score used as performance indicators showed far superior performance in multi-label classification based on the proposed methodology compared to traditional multi-label classification methods. This can be seen that the low-dimensional latent label space derived through the proposed methodology well reflected the information of the high-dimensional label space, which ultimately led to the improvement of the performance of the multi-label classification itself. In addition, the utility of the proposed methodology was identified by comparing the performance of the proposed methodology according to the domain characteristics and the number of dimensions of the latent label space.